Rwanda - Banque Rwandaise de Developpement Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 4921 PROJECT COMPLETION REPORT RWANDA - BANQUE RWANDAISE DE DEVELOPPENENT (BRD) (CREDIT 655-RW) February 10, 1984 Eastern Africa Regional Office This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency unit Rwanda Franc (RwF) US$1.0 = RwF 92.84 RwF 1.0 = US$0.10771 RwF 1.0 million = US$10,771 GLOSSARY OF ABBREVIATIONS BNR Banque Nationale du Rwanda - Central Bank BRD Banque Rwandaise de Dgveloppement CCCE Caisse Centrale de Cooperation Economique CIDA Canadian International Development Agency CDI Centre de Dgveloppement Industriel DEG Deutsche EntwickLungsgesellschaft EDF European Development Fund KfW Kreditanstalt fUr Wiederaufbrau MINECOM Ministare de l'Economie et du Commerce MAGERWA Magasins Ggn6raux du Rwanda SOMIRWA Socifte des Mines du Rwanda FISCAL YEAR January 1 - December 31  PROJECT COMPLETION REPORT FOR OFFICIAL USE ONLY RWANDA - BANQUE RWANDAISE DE DEVELOPPEMENT (BRD) (CREDIT 655-RW) TABLE OF CONTENTS Page Preface ........................ Basic Data Sheet .................... ii Highlights ......... iii I. INTRODUCTION Background and Project History. 1 Credit 655-Rw 1.. Credit 896-Rw.. 2 II. MACROECONOMIC, INDUSTRIAL AND FINANCIAL OBJECTIVES Macroeconomic objectives.. 2 - Economic Structure and Recent Developments . .2 - BRD's Role .... 4 Main Issues .....4 - Interest Rates.... 4 - SSE Financing . .. 6 III. INSTITUTIONAL OBJECTIVES Main Issues... 7 - Technical Assistance to Enterprises. .. 7 - Single Project over-Exposure. . . 8 - BRD's Appraisal Process .. 9 IV. OPERATIONAL AND FINANCIAL PERFORMANCE Operations .. 9 Portfolio ..................................................... 11 Financial Results and Condition ............................... 11 V. ALLOCATION OF IDA CREDITS Credit 655-Rw ....... ................................. 11 Credit 896-Rw .................................. 13 VI. IDA PERFORMANCE ......................................... 14 VII. CONCLUSION ......................................... 15 Annex No. List of Annexes 1. Projected and Actual Lending Operations, 1975-82 .16 2. Projected and Audited Balance Sheets, 1975-82 .17 3. Projected and Audited Income Statements, 1975-82 .18 4. Projected and Actual Financial Ratios, 1975-82 .19 5. Analysis of all Loans and Equity Investments Approved as of December 31, 1982 ...... ................. ............... 20-21 6. Subprojects refinanced under credit 655-Rw ..................... 22 7. Financial Characteristics of Subprojects Refinanced Under Credit 655-Rw ................................................ 23 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.  Table of Contents (Cont'd) Page 8. Economic Characteristics of Subprojects Refinanced Under Credit 655-Rw ................................................ 24 9. Summary Description and Present Status of subprojects Refinanced under Credit 655-Rw ......................... 25-27 10. Arrears Situation of Subprojects Refinanced under Credit 655-Rw 28 11. Sample of Subprojects not Refinanced by IDA ... ................. 29 12. Financial Characteristics of Subprojects not Refinanced by IDA . 30 13. Economic Characteristics of Subprojects not Refinanced by IDA 31 14. Arrears Situation on Subprojects not Refinanced by IDA ......... 32  PROJECT COMPLETION REPORT RWANDA - BANQUE RWANDAISE DE DEVELOPPEMENT (BRD) (CREDIT 655-RW) PREFACE This is a completion report on the First IDA credit (Cr. 655-Rw) to the Banque Rwandaise de Developpement (BRD). The credit of US$4.0 million was approved in July 1976, signed in August 1976 and became effective in March 1977. At the request of BRD US$0.1 was cancelled on January 7, 1982 and the credit was closed at that date. The Eastern Africa Regional Office has prepared this report on the basis of information gathered during a mission to Rwanda in November 1982. The report reviews BRD's institutional developments since 1975 and its utilization of the proceeds of the credit. No comments were received from the Borrower. This project has not been audited by the Operations Evaluation Department.   - ii - PROJECT COMPLETION REPORT RWANDA - BANQUE RWANDAISE DE DEVELOPPEMENT (BRD) (CREDIT 655-RW) BASIC DATA SHEET (Amounts in US$ million) As of April 30, 1983 Original Disbursed Cancelled Repaid Outstanding Cr. 655-Rw 4.0 3.9 0.1 - 3.9 Cumulative Credit Disbursement FY78 FY79 FY80 FY81 (i) Planned 1.0 2.4 3.6 4.0 (ii) Actual 0.8 2.5 3.5 3.9 (iii) (ii) as % (i) 80 104 97 98 Project Data Original Credit Dates Actual or Re-Estimated Board approval 7/15/76 7/15/76 Signing 8/20/76 8/20/76 Effectiveness 3/2/77 3/2/77 Commitment 6/78 12/78 Credit Closing 6/30/81 1/7/82 Mission Data Date No. of No. of Dates of month/year weeks persons Manweeks report Identification 5/75 0.4 2 0.8 7/14/75 Appraisal 10/75 1.4 2 2.8 6/24/76 Supervision I 11/76 0.7 2 1.4 3/25/77 Supervision II 2/78 1.7 2 3.4 12/13/78* Supervision III 7/79 0.9 2 1.8 9/25/79 Supervision IV 5/80 1.1 1 1.1 6/20/80 Supervision V 6/81 0.9 2 1.8 7/20/81 Supervision VI 5/82 1.4 2 2.8 8/11/82* Completion 11/82 1.0 1 1.0 5/20/83 * Combined with appraisal. Follow-up Projects Second BRD Project, Credit 896-Rw approved on April 24, 1979 in the amount of US$5.2 million;and Third BRD Project, Credit 1344-Rw approved on July 7, 1983 in the amount of US$7.0 million.   - iii - PROJECT COMPLETION REPORT RWANDA - BANQUE RWANDAISE DE DEVELOPPEMENT (BRD) (CREDIT 655-RW) HIGHLIGHTS Credit 655-Rw was the first IDA operation with the Banque Rwandaise de Dgveloppement (BRD) and represented a diversification of IDA's assistance to Rwanda which, until then, was concentrated on improving the road network and strengthening agriculture production. Since appraisal in 1975, IDA has enjoyed a close relationship with BRD's management and its senior staff who appreciated IDA missions's comments and guidance. Today, after 3 credits, BRD remains IDA's only channel for assisting the small but fast growing industrial sector in Rwanda. BRD's success in establishing itself as a strong finance company and the main support for the development of the private sector in Rwanda was due to three factors: (i) a stable political environment and a good investment climate which prevailed all along the period 1975-82 when the economy grew at a fast rate while a strong foreign exchange reserve position was maintained by the central bank; (ii) a competent and dynamic management at BRD who was fast to seize good investment opportunities but pursued very prudent financial policies; and (iii) excellent support and advice from a small but dedicated team of expatriate experts with a good knowledge of the country and the institution. In the future, BRD's operations are expected to continue increasing at a high rate, but with the Rwandese economy facing a more difficult situation, BRD should, more than in the past, focus attention on the impact of its operations on the country's foreign exchange reserves and employment situation. In this respect, it should somewhat move away from secured lending and emphasis on collateral to lending based on project appraisal. The flexible position taken by IDA on many occasions over the last 7 years, including the exemption from payment of the commitment fee granted to BRD under the first two projects and the decision not to impose a limitation on the use of IDA funds despite BRD's tendency to finance large projects with IDA lines of credit (paras. 1.05, 3.04 and 5.05), has substantially contributed to the fast growth and present sound financial position of the institution. This understanding attitude has also tightened IDA's relationship with BRD and helped maintain a meaningful dialogue with BRD's management on important policy issues. One result has, however, been that IDA funds only financed a few large projects (paras. 5.01 - 5.04).  - iv - The lesson to be learned from this experience is that, while IDA should continue to be flexible in justifiable circumstances, we should also maintain a close scrutiny over BRD's operations, particularly when reviewing subprojects to ensure that IDA funds are used to the maximum benefit for the country. The following points may be of particular interest: - IDA's efforts to simplify BRD's interest rate structure (paras. 2.07-2.09). - BRD and SSE financing (para. 2.10). - Technical assistance to enterprises (para. 3.04). - Single project over-exposure (para. 3.05). - BRD's appraisal process (para. 3.06).  PROJECT COMPLETION REPORT RWANDA - BANQUE RWANDAISE DE DEVELOPPEMENT (BRD) (CREDIT 655-RW) I. INTRODUCTION Background and Project History 1.01 The Banque Rwandaise de Dgveloppement (BRD) was established in 1967 to provide financing for the creation or development of Rwandese enterprises through equity investments and/or granting of term loans. At least 55% of its share capital are reserved for the Government and public Rwandese agencies, the remainder is open to subscription by private Rwandese and foreign institutions. There are, however, no differences between the rights and privileges of either category of shareholders. BRD started operations in 1968 and since then, has been able to attract a large number of new shareholders and foreign creditors. It now holds lines of credit from three different aid donors and its capital, besides the public sector, is subscribed by four foreign institutions and more than ten private banks and enterprises in Rwanda. 1.02 IDA relationship with BRD started in 1975 when, at the request of the Government, a mission visited Rwanda to examine with BRD's management the situation and prospects of the company and to consider the possibility of IDA assistance. The mission concluded that in view of BRD's important role in the financing of development in Rwanda, IDA assistance was warranted. In July 1976, IDA approved the first credit to BRD. A second credit was granted in April 1979 and a third in April 1983. Since 1975, IDA has sent full supervision or appraisal missions to BRD at a rate of one mission a year. Credit 655-RW 1.03 The main objectives of the first IDA credit, as stated in the Appraisal Report were to: (i) strengthen the project appraisal capabilities, internal organization and accounting procedures of BRD; (ii) provide term financing to well conceived investment projects; and (iii) provide technical assistance to enterprises through BRD. In assisting BRD, IDA also hoped to improve the Government's own project evaluation procedures and promote the development of private enterprises. The following conditions were attached to this first credit: (a) BRD should not charge interest rates below 7% per annum on subloans to small enterprises and 9% per annum on all other sub-loans; (b) BRD should calculate the EROR for all import substitution projects above the free limit of US$100,000; (c) BRD should change its Policy Statement in order not to bear any foreign exchange risk on its borrowing which would be assumed by the Government at no cost to BRD.  - 2 - (d) BRD's accounts should be audited by independent auditors acceptable to IDA. The cost of the first three audits would be financed by the Credit; and (e) BRD would provide technical assistance to local enterprises, the cost of which would be financed by a one-point extra spread given to BRD under this credit (BRD needs a spread of 4% on its borrowed funds to cover expenses and makes a normal profit, but was granted a 5% spread on its IDA borrowing). Appraised in May 1975, the credit was approved in July 1976 and became effective in March 1977. It was fully committed on time and disbursed at a relatively fast pace. The credit was closed on January 7, 1982. Credit 896-Rw 1.04 The second credit was essentially the continuation of the institution building efforts undertaken under the first credit. It also had an additional objective of assisting the Government in undertaking a study to assess the need for an audit firm and to design a program for its establishment. Special conditions on this credit were that: (a) BRD would raise its interest rates to 8.5% on medium term sub-loans to small enterprises and to 9% on medium term sub-loans to all other enterprises, and to 10% on long term sub-loans to small enterprises and to 11% on long term sub-loans to all other enterprises; and (b) BRD would obtain on a regular basis adequate guarantees of loan repayment on at least that part of the loan that exceeds its normal exposure limit of 20% of net worth, either from Government for public sector projects, or from other creditworthy sources for private sector projects. 1.05 As under the first credit, BRD was exempted from paying the commitment fee and received IDA funds from Government at a rate (6%) below the full IBRD rate (at that time 7%). Finally, the Government continued to assume the foreign exchange risk at no cost to BRD or the subborrowers. The credit was appraised in February 1978, approved in July 1979 and became effective in January 1980. At the end of May 1983, US$4.1 million (78%) was committed and US$1.1 million (21%) disbursed. II. MACROECONOMIC, INDUSTRIAL AND FINANCIAL OBJECTIVES Macroeconomic objectives 2.01 Economic Structure and Recent Developments. Rwanda, a small landlocked country, which depends on Kenya and Tanzania for its access to the sea, has a population estimated at 5.5 million, most of whom deriving their livelihood from agriculture. Despite the predominantly rural characteristics of its economy, Rwanda has a relatively active modern sector. Services have grown fast over the last 6-7 years and now account  - 3 - for 37% of GDP. This was mainly due to the rapid growth of foreign aid, which directly resulted in higher public consumption and investment. These expenditures, in turn stimulated the growth of commerce and other service activities. The industrial sector is still small but is growing at a fast rate. Production is typically of the first generation of import substitution: brewing, food processing, textiles, plastic utensils etc. Exports are limited to processed coffee and tea. Agroindustrial enterprises dominate the sector, accounting for about two-thirds of its value added. Manufacturing activity is mainly centered around Kigali, the capital, although some towns in the interior are becoming more active and important. Ownership is mostly local private with foreign control limited to a few large enterprises. With the exception of food processing, most industrial enterprises are highly dependent on imported inputs and, although there have been so far no strict import controls, transport difficulties have often resulted in interruption of supplies, forcing firms to hold large stocks of imports and spare parts, thus increasing their costs. 2.02 Rwanda's financial system consists of the following institutions: the central bank, two commercial banks, the Savings Association, a mortgage bank (Caisse Hypothgcaire), a network of cooperative banks similar to credit unions (Banques Populaires) and BRD, the development bank. The two commercial banks are mostly engaged in short term financing of exports, particularly coffee, and imports. Term lending is provided by the other financial institutions, with the Savings Association making loans only to its members and the Banques Populaires concentrating in rural areas. The Caisse Hypothecaire, which is supposed to finance all housing and building of a value of RwF 0.5 million or more, is still relatively inactive. BRD is entrusted with financing development projects although the other financial institutions, particularly the commercial banks, can also provide term credit for productive investments with central bank's approval. 2.03 During the period 1975-80, the Rwandese economy experienced a rapid growth of over 5% per year, as a result of a combination of favorable external factors (weather, terms of trade and capital inflows) and a prudent economic and financial management. Per capita income increased from US$76 in 1974 to US$200 in 1980, but is still among the lowest in the world. In 1981 and 1982, GDP continued to grow at a high rate, but, in contrast to previous years, this growth was accompanied by a weakening of the budget and the balance of payments, reflecting unfavorable world market conditions for coffee (Rwanda's main export) and some relaxation of the Government traditional policy of expenditure restraint. At the end of 1982, however, Rwanda still enjoyed a strong reserve position (US$128 million, representing 5.6 months of 1982 imports). 2.04 Rwanda's Third Development Plan (1982-86) retained most of the objectives of the Second Plan. With regard to industry, these include: (i) increasing the production of basic goods either for consumption by the population, or for use in agriculture; (ii) promotion of small scale enterprises; and (iii) increasing the processing of agricultural and mineral products to provide the maximum possible employment to compensate for the limited oppportunities in the rural sector. To achieve these objectives, the Government uses a number of instruments, the most important of which are import licensing, price control, industrial licences and  - 4 - import tariffs. Efforts are being made to streamline the procedures of the investment code and improve its applicability. Because Rwanda's landlocked location and the high transport costs provide significant advantages to local industry, the Government does not have to resort to a deliberate protection policy. Rwanda maintains a relatively liberal exchange and trade system. Quantitative restrictions are not normally used. Existing enterprises enjoy on average a moderate level of tariff protection in nominal terms but relatively high effective protection as few of them pay duty on imported inputs. Protection to industry has, however, been raised significantly by the recent introduction of an import deposit scheme. 1/ IDA recently sent a mission to review the manufacturing sector in Rwanda. The report, which is being prepared, will serve as a basis for discussion with the Government on policy and other issues affecting the sector. 2.05 IDA has generally supported Rwanda's development objectives and strategies, particularly the Government's efforts to increase employment and to diversify the economy. This has been the major reason for IDA assistance to BRD which was established to help achieve these objectives. 2.06 BRD's Role. Working within the Government's framework of economic policies, BRD has succeeded in establishing itself as the major source of term lending in Rwanda and has substantially contributed to the development of the economy. Since its creation in 1967, BRD has provided financing totalling RwF 3.5 billion (US$37.5 million) for 259 projects with an estimated total investment cost of about RwF 8.5 million (US$90.2 million). More than 75% of its interventions went to privately owned enterprises. BRD has thus become the Government's main instrument to promote the development of the private sector in Rwanda. Main Issues 2.07 Interest Rates. Because of different views regarding the role of interest rates in the development process, this issue has been the most discussed topic between IDA and BRD over the last 7 years. In Rwanda where the level of private investment is low, the Government feels that promotors should not be discouraged by high interest rates. Thus, in setting rates charged to clients, 2/ BRD should not only take into account its financial situation and the expected inflation rate, but also the Government's policy of facilitating private investment in a difficult environment where project opportunities are limited. 1/ Effective March 1, 1983, imports of certain goods - those with an import duty of 80% or more, as well as those manufactured locally - require a 100% advance deposit, which cannot be financed through bank credit. 2/ BRD's interest rates are not subject to central bank's regulation but must be approved by its Board in which the Government has a majority vote.  -5- 2.08 During the period 1975-79, rates set for short-term loans by commercial banks ranged from 4% for tea export to 11% for import credits and working capital. The Savings Association, the other institution mainly engaged in term lending, was allowed a maximum rate of 9% on its medium-term credits to the private sector. Over that period, BRD charged rates ranging from 6.5% to 9.0% (including commissions), depending on the maturity of the loan and the sector. In November 1979, the central bank put into effect an upward revision of the interest rate structure in Rwanda and introduced specific guidelines regarding rates charged by commercial banks for productive investments which, if approved by the central bank, should not exceed 11%. Other term loans (mostly housing) carry rates of about 12% for medium-term loans and 13% for long-term ones. Following that readjustment, BRD raised the overall level of its interest rates to 8-13% (excluding commissions which included a flat fee of 0.75% of the loan amount to be paid at signature and a commitment fee of 1%) but introduced another differentiation, this time, by project profitability. In relation to inflation in Rwanda, BRD's interest rates were significantly negative in real terms in 1978 and 1979 when local prices increased at about 14% per year as a result of frequent disruptions of supply routes through neighboring countries and increased import prices. After the 1979 adjustment and as inflation abated, BRD's interest rates became again positive in real terms, except for those charged to small enterprises and agricultural projects. 2.09 IDA's efforts in this area have been to (i) simplify BRD's interest rate structure which is characterized by differentiation by sector, term and project profitability as measured by the financial rate of return estimated at appraisal and (ii) bring all rates to a positive level in real terms so as to improve resource allocation and maintain BRD's profitability. Under the first credit, after long and strenuous discussions, it was agreed with BRD that IDA funds be passed on to clients at a minimum rate of 9% for medium and large enterprises and 7% for SSEs with the foreign exchange risk borne by the Government. These rates were further raised under the second credit to 8.5%-10% for SSEs and 9-11% for other enterprises. At negotiations of the Third Credit, BRD accepted to (i) eliminate the financial rate of return as a criterion for determining interest rates; (ii) not to give any interest rate preferential to SSEs, with the exception of very small loans of less than US$11,000 for which no detailed evaluation is made; (iii) to set and maintain all interest rates at positive levels in real terms and to review the level and structure of interest rates with IDA at least once a year and (iv) to carry out a study on interest rate structure in Rwanda to see how the system can further be rationalized and better contribute to the achievement of the Government's development objectives. As a concession, we agreed to allow BRD to maintain a small differentiation by sector. IDA's dialogue with BRD and the Government on this issue has thus been quite fruitful. All BRD's interest rates are now positive in real terms as can be seen in the table below:  - 6 - BRD's Present Interest Rate Structure (in percent per year) Medium-term loans Long-term loans Agriculture and Livestock 10 11 Industry and Mining 11 12 Services 12 13 Note: In addition to the interest rate, a flat rate of 0.75% of the loan amount is paid at signature and a commitment fee of 1% is charged on undisbursed funds, effectively increasing the rates charged to clients to about 10.5% and 11.5% for agriculture, 11.5% and 12.5% for industry and mining, and 12.5% and 13.5% for services. These rates compare with inflation projected at 9.5-10.0% over the period 1983-87. 2.10 SSE Financing. Perhaps the most difficult challenge facing Rwanda is to provide sufficient job opportunities for the country's expanding labor force. Unemployment and underemployment are acute in urban areas, particularly in Kigali (the capital), where the population has been growing at a fast rate of 9% per year over the last ten years. Although the Third Development Plan (1982-86) accords a high priority to the promotion of small enterprises, a labor intensive sector, no coherent program has been developed to assist them. Until recently BRD had not placed strong emphasis on financing SSEs because of the risks and costs attached to this type of lending. In the first two credits, IDA did not press BRD into this activity as the first priority was to build up and strengthen the institution. Now that BRD is financially strong and has a relatively experienced staff, it has decided to launch a new intensive effort to promote and finance SSEs. Under the third IDA credit, we will assist BRD in reorienting its operations towards SSEs and have allocated at least US$2.5 million for that purpose. This third credit also includes a technical assistance component to help the Government carry out a study of the issues related to the development of an SSE program in Rwanda which could lead to a future SSE project. III. INSTITUTIONAL OBJECTIVES 3.01 BRD's management is highly motivated, competent and stable. Since IDA's first appraisal in 1975, there was only one change at the top in 1982 when the last Managing Director was replaced at the end of his third three-year term by the Advisor to the President of the Republic. The new Managing Director is assisted by two competent expatriate experts who have been with BRD since the mid-1970's. BRD's staff now number 68 (of whom 23 professionals), i.e., a threefold increase since 1975. Largely due  - 7 - to a good compensation policy, 3/ BRD has been able to retain staff of good quality. Some have been with BRD for more than five years. Most of the professional staff have university degree and are generally dedicated. Staff development relies mainly upon practical on-the-job training which is performed by the two expatriate experts who have the confidence of the staff. 3.02 Since 1975, BRD has substantially improved the quality of its appraisal work and has developed good operating policies and procedures. Appraisal reports prepared by the staff are first thoroughly reviewed and discussed in the Investment Department, then presented to the Studies Committee, which is chaired by the Director General and composed of all the Bank's senior staff, before being submitted to the Board. BRD's staff is particularly strong in financial analysis. In reviewing subprojects, IDA seldom questioned BRD's financial evaluation. In contrast, economic analysis of subprojects is still weak. BRD has been advised to pay more attention to the economic merits of its operations and has been requested to calculate the economic rate of return for all non-service sector projects above US$200,000. Over the last 7 years, BRD has significantly strengthened its supervision capability. The emphasis put by IDA on assistance to enterprises and the provision under the first credit of an extra one point spread to help BRD finance a technical assistance program played a part in this improvement (para. 3.05). 3.03 Since 1976, in compliance with the provisions of the first Credit Agreement, BRD accounts have been audited by an independent auditing firm. The audits are generally satisfactory and in line with IDA requirements. BRD has sound accounting and financial management systems and uses appropriate disbursement procedures. 3.04 Under the first credit, funds were allocated to cover the cost of the first three audits. Although BRD did not use these funds and financed the audits out of its operating profit, this measure, together with the exemption for BRD from paying the commitment fee to the Government under the first two credits (para. 1.05) are indicative of the flexible position IDA is prepared to take in justifiable circumstances. Main Issues 3.05 Technical Assistance to Enterprises. Under the first credit, BRD was granted an extra point in the spread to help it provide technical assistance to enterprises. It was expected that BRD would liaise with the Bureau of Industrial Promotion (Bureau de Promotion Industrielle, to be established by UNIDO within the Ministry of Economy and Finance) to select quarterly a number of small- and medium-size firms and establish with their management an assistance program to be implemented over an agreed period. BRD recognized that it has not provided assistance to enterprises as it should have done, but pointed out that (i) the Bureau of Industrial 3/ BRD must comply with Government salary directives but always grants staff a bonus at the end of the year. The importance of the bonus depends on staff performance and BRD's financial results and is decided by the Director General. Some senior staff also have housing privileges.  -8- Promotion, which was supposed to be established in 1977, only became effective in 1982; and (ii) alone, BRD could not do much until it was better established. At that time, BRD was still a young and inexperienced development institution whose own procedures and capabilities needed strengthening. The funds made available under the first IDA credit for assistance to enterprises actually enabled BRD to recruit staff and improve the supervision of its own projects which is at present adequate. BRD's staff now visits each project under supervision twice a year. Those under construction are systematically visited after the first disbursement and at completion. Projects encountering difficulties receive additional visits. During supervision missions, BRD's staff often assists small entrepreneurs in estimating their working capital needs and suggests ways to improve the quality of their products. Without IDA financing it was unlikely that BRD could have developed that needed supervision capacity. Thus, IDA's objective of providing assistance to Rwandese enterprises was not achieved but BRD made good use of the funds granted to it to reinforce its procedures and evolve into a strong financial intermediary. Now that BRD is embarking on SSE financing, and intends to assume a more active role in the sector, assistance and training of entrepreneurs will by necessity be an integral and important part of its activity. To carry out this task, BRD intends to reinforce the Supervision and Assistance Division by recruiting two more professionals in the next three years. 3.06 Single Project over-Exposure. As the major term lender in Rwanda, BRD has financed projects exceeding the normal exposure limit of 20% of net worth. This was recognized as an issue at the time of appraisal of the second IDA credit and was discussed with BRD's management. Several alternatives were explored in the Appraisal Report of BRD II (No. 2142-Rw), but were not pursued because they were either premature or posed problems to BRD. Co-financing with other local financial institutions in Rwanda was not feasible as they were not ready to participate in term lending, except for real estate. The managed fund approach would have reduced BRD's autonomy and independence in deciding upon the projects which, however, it would have to administer. Thus, BRD might be faced with the problem of having to supervise difficult projects which could cost it much more in staff time and other administrative charges than the management fees normally granted for this kind of work. At a time when BRD's financial situation was still fragile, it would have been imprudent to push BRD into that direction. It was therefore decided to include in the IDA Credit Agreement a covenant requiring BRD to obtain adequate guarantees on the part of the loan that exceeded its normal exposure limit. BRD has always been in compliance with this covenant. In late 1980, BRD obtained from its Board a new increase in its authorized capital from RwF 416 million to RwF 1 billion. Most of it has been paid-in, which together with reserves and retained earnings, have increased BRD's exposure limit to nearly RwF 220 million (US$2.3 million), as compared to about RwF 90 million (US$1.0 million) at the time of BRD II. In view of this increase and considering BRD's greater emphasis on smaller projects, at negotiations of the third IDA credit, the over-exposure guarantee condition was eliminated and BRD was requested to stay within its exposure limit. A covenant to that effect was included in the Credit Agreement.  -9- 3.07 BRD's Appraisal Process. As discussed in para. 3.02, the quality of BRD's appraisal work is good. Projects are thoroughly analyzed and, in some cases, redesigned during appraisal to make them more suitable to the Rwandese economic and financial conditions. However, BRD has been somewhat slow in processing projects which could take more than a year from the time of submission by the promotor to Board presentation. This was the case for 10 out of a sample of 21 SSE projects financed between 1976 and 1980. The average processing time for medium and large size projects is about 8 months, with some exceeding a year. Considering that most BRD's projects are refinanced under external lines of credit and must be approved by foreign institutions, funds are sometimes made available to the promotors only 1 year to 1-1/2 year after submission of their project to BRD. Over the period 1979-82, with 7 project officers, BRD had approved on average only 13 projects per year, of which a third were very small operations of less than RwF 3 million for which there was no Board presentation. Promotors have often complained to IDA missions about BRD's delays in processing their request for finance, some pointing out that BRD was regarded as a lender of last resort, an institution to go to only if there was no other alternative. However, it must be noted that many projects submitted to BRD needed substantial preparatory work before they could be appraised. Until 1981-82 when the UNDP/UNIDO supported Industrial Promotion Bureau started to provide assistance to entrepreneurs, 4/ BRD was the only body in Rwanda with some expertise in project preparation and had to do the work, which in other countries is done by specialized institutions. It is also true that BRD's project officers do a very comprehensive work when appraising a project, reviewing and checking all the information provided by the promotors and this takes time in Rwanda where things move slowly. On the other hand, BRD's priviledged position as the major term lender in Rwanda with a strong project pipeline, does not encourage it to be flexible to promotors. IDA discussed this issue with BRD's new Director General who agreed that there is room for increasing staff productivity (and thus decreasing administrative costs) and that efforts would be made to accelerate the processing of projects and improve the institution's image with the business community. IV. OPERATIONAL AND FINANCIAL PERFORMANCE Operations 4.01 As of December 31, 1982, BRD has approved 246 loans totalling RwF 3,268 million and 13 equity investments amounting to RwF 221 million. While most of the equity investments were made before the mid-1970's, loan approvals started to become important only in 1974 when they amounted to RwF 163 million. Since then, they have expanded rapidly to reach a peak of RwF 624 million in 1980, but fell in 1982 to RwF 267 million as a result of slippage of a very large project. BRD's overall level of activity is indeed affected by a few large projects that, as the major term lender in Rwanda, BRD is called upon to finance. Compared to forecasts made at the time of appraisal of the first credit, BRD's approvals increased at about 4/ However, due to UNDP's shortage of funds the Bureau may have to cease its activities at the end of 1983.  - 10 - the rate forecast (16% per year) in 1975-78, fell in 1979, but were 42% higher than the level forecast in 1980 because of three large projects approved that year for a total amount of RwF 471 million. Commitments and disbursements lagged considerably behind approvals and represented respectively 57% and 54% of Bank forecasts during the period 1975-80. The main reasons for this rather poor performance included transport difficulties, particularly during the Uganda war in 1978-79, delays of clients in meeting BRD's disbursement conditions, and difficulties in the execution of a few large projects. 5/ In 1981, disbursements increased at a high rate of 76% and preliminary dates indicate that the increase was also strong in 1982. 4.02 Characteristics of BRD's loans and equity investments approved as of December 31, 1982 are shown in annex 5. Approvals of less than RwF 10 million accounted for 81% in number but only 10% in value. The total cost of projects financed ranged from RwF 1.2 million to RwF 577.0 million. On average, BRD financed about a third of the total project cost. Eighty five percent of the loans carried interest rates of up to 9% and maturities ranging from 5 to 12 years. Most of these loans were granted before 1979 when BRD's interest rates were maintained at a relatively low level. Over the last three years, most loans approved had interest rates in the range of 10-13%. At the end of 1982, these loans represented 15% of total approvals but 40% in value. The breakdown of BRD's operations by sector is as follows: Approvals (Amounts in RwF million) Number % Amount % Agriculture and Livestock 29 11 52 1 Agro-industry 15 6 708 20 Manufacturing Industry 23 9 580 17 Other Industries, Mining and Artisanat 39 15 791 23 Hotels and Tourism 11 4 165 5 Transport Equipment 126 49 621 18 Other 16 6 572 16 Total 259 100 3,489 100 Thirty six percent of BRD's operations, representing 79% of approval value, were concentrated in Kigali. Other locations of importance included Gisenyi, Ruhangeri and Cyanguru. 5/ The Air Rwanda Cargo plane project (US$2.5 million), for example, was approved in late 1977, but only committed and disbursed in 1979 because of difficulties in finding financing. (BRD requested IDA financing for this project under the second credit but the request was turned down on marketing and economic grounds, and BRD had to use local resources to finance it). Similarly, the SOMIRWA subproject refinanced under the first IDA credit was approved by BRD's Board in late 1976 but only committed in late 1977 because of delays in IDA credit effectiveness and problems associated with the subproject when it was initially submitted to IDA for review (see Annex 9).  Portfolio 4.03 At the end of December 1982, BRD's portfolio amounted to RwF 1,709 million, of which 87% in loans and 13% in equity investments. Loan arrears of more than 3 months amounted to RwF 17 million, i.e., less than 1% of the portfolio, and affected 10 projects accounting for 6% of the overall outstanding portfolio. All of the projects affected by arrears were small and medium enterprises, with the exception of SOMIRWA, the mining company, which is partly owned by the Government. Debts considered doubtful or bad represented RwF 49 million, or 3% of the portfolio. To cover potential losses, BRD has made specific provisions amounting to RwF 17 million, which is adequate as some of these loans are guaranteed by the Government. In addition, there is a provision for general risk amounting to RwF 50 million (3% of loans outstanding). The quality of BRD's loan portfolio is good, and generous provisions have been set aside for possible losses. Financial Results and Condition 4.04 During the period 1975-80, BRD's total assets expanded at a high rate of 20% per year, albeit slower than forecasted by the Bank appraisal mission (26%), because of disbursement difficulties discussed in para. 4.02. In 1981 and 1982, however, the growth of total assets increased to 26% per year as disbursements picked up. All along the period 1975-82, BRD has maintained a strong financial position. Cash expenses (administrative costs and financial charges) were kept at a level constantly lower than that forecasted at appraisal, while income was, most of the time, higher, mainly as a result of the very profitable equity investment BRD made in the MAGERWA project (Magasins Ggngraux du Rwanda) in 1969. Net profit as a percentage of average net worth averaged 7.2%, by far exceeding the appraisal forecasts of 5.8%, despite constitution of generous provisons for possible losses. At the end of 1982, BRD's debt equity ratio was 0.7:1.0, well within the 3:1 limit imposed by IDA. V. ALLOCATION OF IDA CREDITS Credit 655-Rw 5.01 As shown in the table below, BRD was very fast in committing IDA funds. The credit became effective in March 1977, but at the end of the year nearly 60% of the line of credit were already committed. Disbursements followed closely the rate estimated at appraisal, which is remarkable considering the much longer disbursement profile for IDF projects in East Africa.  - 12 - Table 1: Commitments and Disbursements of Credit 655-Rw Cumulative Commitments Cumulative Disbursements US$ '000 Percentage US$ '000 Actual (%) Appraisal End December 1977 2,360 59 323 8 10 1978 2,360 59 1,668 42 40 1979 4,000 100 3,207 80 75 1980 4,000 100 3,545 89 100 1981 4,000 100 3,875 97 100 5.02 The main reason for this remarkable performance is the fact that the first IDA credit financed only 6 subloans, all above the free limit, for 5 medium and large enterprises: SOMIRWA, a mining company in which the Government has a 49% participation; SULFO-RWANDA, a sweet manufacturing firm; MIRONKO (2 subloans), a plastic goods manufacturing project; BATAVIA, a photo processing laboratory; and TOLIRWA, a plant processing corrugated iron sheets. One subloan used 40% of the line of credit. Indeed, BRD has the tendency of using IDA funds for large projects as all its other external lines of credit carry special conditions with regard to sector or size of projects which can be financed. This has created the problem of over-exposure discussed in para. 3.05. Under credit 655-Rw, 2 out of the 6 subloans exceeded BRD's exposure limit of 20% of net worth. In one case, BRD's over exposure was so large that before approving the subloan, IDA requested that adequate supplementary guarantee be obtained from Government (Annex 9). 5.03 Summary description of the 6 subprojects, their financial and economic characteristics, as well as their status of implementation are given in Annexes 6-10. The subloans carried interests rates of 7%-12%, excluding commissions, and maturities between 6 and 12 years. All recipients of subloans were private enterprises, with the Government holding a minority participation in one of them. Four subprojects were new ventures. As of December 31, 1981, the financial performance of two subprojects was satisfactory. The three other subprojects were in deficit, but for two of them the situation is expected to improve as production and sales have recently increased at a fast rate. Compared to appraisal's schedule, implementation of most of the subprojects was delayed and all of them incurred sizeable cost overruns. In general, transport difficulties and, in some cases, changes in the investment program were the main causes of implementation delays. It should also be noted that an usually long period of time elapsed between subproject approval by BRD's Board and by IDA. In 4 out of the 6 subprojects, this period ranged from 9 to 13  - 13 - months. Generally, Bank staff acted promptly and reviewed subprojects as soon as they received them, but the need for additional information, clarification (and, in one case, obtaining supplementary guarantees from Government) lengthened the review period. In 2 cases, it was also due to delays by BRD in submitting subprojects for review (4 and 8 months after approval by BRD's Board). These implementation delays may have adversely affected the final cost of projects, but the main cause of cost overruns appeared to be an underestimation of costs at appraisal and changes in the investment plan. In 3 cases, purchases of additional equipment and/or machines not foreseen at appraisal were made. As shown in Annex 10 only one subproject has arrears of more than 3 months. 5.04 As requested by the first IDA credit, BRD only calculated economic rates of return for import substitution projects. They were estimated at 11% for MIRONKO, and 15.5% for TOLIRWA. As both have been in operation only since 1980, no actual economic rate of return is therefore available. However, from available data, it appears that the economic contribution of TOLIRWA has been, so far, marginal. The project needs a ban on imports to remain in operation (TOLIRWA management and the Government maintain that the company suffers from unfair competition. See Annex 9) and, although it was not its main objective, job creation by TOLIRWA was limited and very costly: 75 permanent jobs at an average cost per job of about US$47,400. The MIRONKO project, on the other hand, operates without special protection. The company is at present in deficit,but its products appear competitive with imports, particularly the beer cases which are bulky and very expensive to import. Of the 3 other subprojects, 2 used mainly local raw materials and one exported most of its production. As in large operations generally, subprojects financed under the first IDA credit have a relatively high cost per job created (US$19,500-47,400), with the possible exception of SOMIRWA where the number of jobs created directly was limited (175), but by rehabilitating mining operations, the project helped sustain the employment of about 7,000 wage earners as well as several thousands artisan miners who sell their ores to SOMIRWA. Credit 896-Rw 5.05 Since April 1980 when this credit became effective, 86% of the line of credit of US$5.0 million have been committed for 8 subprojects, of which 3 below the free limit. Two subloans were over US$1 million and one, US$21,000 short of one million. These three subloans used about two-thirds of the line of credit. BRD's tendency to use IDA funds for very large projects has thus continued under the second IDA credit. This was flagged as an issue in the Pre-mission Issues Paper of BRD III which raised the possibility of imposing a limitation on the use of IDA funds. However, it was recognized that BRD as the major financial institution in Rwanda cannot but finance large projects and that IDA is at present the only institution providing BRD with untied funds. We have nevertheless indicated to BRD our desire to see IDA funds spread over a larger number of projects. 5.06 Unlike the first line of credit, subprojects financed under credit 896-Rw were spread over a larger number of sectors. Only 3 subprojects were for industry and assisted firm in expanding their  - 14 - production line (corrugated iron sheets and blankets), or start a new operation (mattresses). Four were for tourism development in the interior, purchase of printing materials and financing of equipment for automobile and truck repair. The last one was to assist MAGERWA (Magasins Generaux du Rwanda, a company in which BRD is the main shareholder) in renewing and expanding its storage facilities which had become insufficient to meet the country's needs. BRD's loans to these subprojects amounted to RWF 400 million (35% of total project costs) and carried interest rates of 11-13%, excluding commissions and maturities of 5-15 years. Most of them are completed and have started operation. They are all concentrated in Kigali and its vicinity, with the exception of the two hotels in the interior. VI. IDA PERFORMANCE 6.01 In 1975 when credit 655-Rw was appraised, BRD was still a young and inexperienced development bank whose procedures and systems needed considerable strengthening. Thus, the Credit included a number of covenants aimed at improving BRD's efficiency and protecting its financial position, including the obligation for BRD to (i) change its policy statement in order not to bear any foreign exchange risk; (ii) have its accounts audited by independent auditors; (iii) charge higher interest rates to its subborrowers; and (iv) calculate the economic rate of return for all import substitution projects above the free limit of US$100,000. 6.02 BRD has always been in compliance with these covenants. Its accounts are regularly audited by a reputable and independent auditing firm and with IDA's support, BRD has considerably improved its appraisal and supervision capacity, which was almost non-existent in 1975. 6.03 The flexible position taken by IDA on many occasions over the last 7 years, including the exemption from payment of the commitment fee granted to BRD under the first two projects and the decision not to impose a limit on the use of IDA funds per subloan, has substantially contributed to the fast growth and present sound financial position of BRD. This understanding attitude has also improved IDA's relationship with BRD and helped maintain a fruitful dialogue with BRD's management on important policy issues, such as BRD's level and structure of interest rates. IDA succeeded in convincing BRD to simplify its interest rate structure which was characterized by differentiation by sector, term and profitability as measured by the financial rate of return estimated at appraisal and to increase all its rates to a positive level so as to improve resource allocation and maintain BRD's profitability. One adverse result has, however, been that IDA funds only financed a few large projects, some of which appear marginal from an economic point of view. Thus, while IDA should continue to be flexible in justifiable circumstances, it should also maintain a close scrutiny over BRD's operations, particularly when reviewing subprojects to ensure that IDA funds are used to the maximum benefit for the country.  - 15 - VII. CONCLUSION 7.01 The major objective that IDA set out to accomplish in this first credit, which was to strengthen project appraisal capabilities, internal organization and accounting procedures of BRD, has been substantially met. Over the last seven years, BRD has achieved a rapid growth in operations and has been very effective in mobilizing local and foreign reserves to support its expanding activities. It has now a strong financial position, a sound portfolio and its staff has a good understanding of project appraisal methodology. BRD has substantially improved its accounting systems and procedures and since 1976, its accounts are regurlarly audited by an independent auditing firm acceptable to IDA. 7.02 BRD has not achieved the specific objective set forth in this credit of providing technical assistance to Rwandese enterprises (para. 1.03). It has, however, made good use of the extra point spread given to it to develop and strengthen its supervision capability which was almost non-existent at that time. As discussed in para. 3.05, BRD did not assist enterprises as agreed with IDA because the Bureau of Industrial Promotion, which was supposed to be established in 1977 with UNIDO assistance, became operational only in 1982 and, alone, BRD could not do much until its own procedures and capabilities were reinforced. Now that BRD has gained more experience and is embarking on an intensive effort to finance SSEs, assistance to enterprises will become an important part of its activity. 7.03 With regard to the last objective of providing term financing to well conceived investment project, the performance of enterprises financed under the credit has been so far below expectations. At present, only 2 out of the 5 enterprises supported by IDA funds are financially profitable, while the economic impact of some of them is, at best, marginal. Admittedly, it was impossible for BRD to foresee the drastic changes in the mineral prices over the last 2-3 years which are one of the main causes of SOMIRWA's present difficult financial position. Nor could it foresee the marketing difficulties encountered by the MIRONKO operation. Both subprojects make sense from an economic point of view. This does not seem to be the case for the TOLIRWA subproject whose products have little local value added and could not save much on transportation costs. 7.04 All in all, BRD as a project has been generally successful. BRD is well known in Rwanda as an independent and efficient financial institution with developmental objectives that it tries to carry out seriously. With a competent management, a sound financial situation and a relatively experienced staff, it is bound to increase its impact and role in the development of Rwanda and deserves continued IDA support.  BANQUE RWANDAISE DE DEVELOPPEMENT (BRD) Projected 1/ and Actual Lending Operatlons, 1975-82 (RwF mililon) 1975 1976 1977 1978 1979 1980 1981 1982 Actual Forecast Actual Forecast Actual Forecast Actual Forecast Actual Forecast Actual Actual Actual Approvals Loans 201 260 229 295 319 330 367 375 298 420 624 427 267 EquIty Investments 34 35 13 35 - 40 - 40 5 40 30 - - Total 235 295 242 330 319 370 367 415 303 460 654 427 267 Commitments Loans 148 245 79 286 78 321 106 363 539 420 193 968 n.a. O Equity Investments 13 35 13 35 - 39 - 40 5 40 30 - - Total 161 280 92 321 78 360 106 403 544 460 223 968 n.a. Disbursements Loans 97 185 41 252 51 296 89 335 484 409 179 313 n.a. Equlty Investments 13 35 26 35 9 39 - 40 4 40 10 20 n.a. Total 110 220 67 287 60 335 89 375 488 449 189 333 n.a. 1/ As per Bank appralsal In 1975. (D X  BANQUE RWANDAISE DE DEVELOPPEMENT (BRD) Projected 1/ and Audited Balance Sheets, 1975-82 (RwF million) 1975 1976 1977 1978 1979 1980 1981 1982 Actual Forecast Actual Forecast Actual Forecast Actual Forecast Actual Forecast Actual Actual Actual Assets Current Assets 92 40 65 68 46 94 68 93 65 112 84 270 288 Portfolio 446 614 531 852 564 1,111 693 1,371 1,069 1,654 1,203 375 1,683 Loans (341) (474) (400) (677) (424) (897) (553) (1,117) (926) (1,360) (1,012) 1,163 (1,470) Equity Investments (105) (140) (131) (175) (140) (214) (140) (254) (143) (294) (191) (212) (213) Fixed Assets (net) 3 2 11 1 26 1 34 1 37 1 41 101 126 TOTAL ASSETS 541 656 607 921 636 1,206 795 1,465 1,171 1,767 1,328 1,746 2,097 ' Currrent Liabilities 20 20 64 20 33 20 7 20 86 20 81 17 82 Term Borrowings 280 382 230 629 270 785 400 1,013 594 1,280 510 678 871 Provislons 16 20 20 25 19 31 23 37 34 43 37 41 49 For General Rlsks (13) (16) (16) (21) (19) (27) (23) (33) (34) (39) (37) (41) (49) For Rents (3) (4) (4) (4) (-) (4) (-) (4) (-) (4) (-) (-) (-) KfW Investment Fund - - - - 1 - 3 - 4 - 6 10 14 Speclal Investment Fund - - - - - - - - - - - 2 7 Equlty 25 234 293 247 313 370 362 395 453 424 694 998 1,074 Funds for Share Capital Increase 2/ - - (50) - (50) - _ - - - - - Pald-in Share Capital (208) (208) (208) (208) (208) (312) (342) (312) (416) (312) (604) (875) (892) Reserves and Retained Earnings (17) (26) (35) (39) (55) (58) (20) (83) (37) (112) (90) (123) (182) s Total 541 656 607 921 636 1,206 795 1,465 1,171 1,767 1,328 1,746 2,097 1/ As per Bank appralsal in 1975. 2/ Transformatlon of a loan from the Central Bank Into a participatlon to permit the Government to subscrlbe to the next capital Increase.  BANQUE RWANDAISE DE DEVELOPPEMENT (BRD) Projected 1/ and Audited Income Statements, 1975-82 (RwF mililon) 1975 1976 1977 1978 1979 1980 1981 1982 Actual Forecast Actual Forecast Actual Forecast Actual Forecast Actual Forecast Actual Actual Actual Income Interests and Commlsslons 15.0 28.2 30.1 41.9 32.2 57.2 40.5 75.0 67.6 90.3 100.7 118.7 157.7 Dividends 12.0 5.0 24.0 5.0 23.2 5.0 22.4 5.0 23.2 5.0 22.8 21.2 21.2 Other Income 4.0 18.0 2.4 2.7 1.2 4.0 1.7 4.0 2.3 4.8 6.0 4.9 32.5 2/ Total Income 31.0 35.0 56.5 49.6 56.6 66.2 64.6 84.0 93.1 100.1 129.5 144.8 211.4 Expenses