Rwanda - Fourth Banque Rwandaise de Developpement Project

DoCument Of The World Bank FO0 OMCIAL USK ONLY cR Reupw No. P-4204-RW REPORT AND RECONMENDAIION OF THE PRESIDENT OF THE INTERNATIONAL DEVELOPMENT ASSOCIATION TO THE EXECUTIVE DIRECTORS ON A PROPOSED DEVELOPMENT CREDIT IN AN AMOUNT EQUIVALENT TO USSt9. 0 MILLION TO THE RWANDESE REPUELIC FOR A FOURTH RWANDESE DEVELOPMENT BANK (BRD) PROJECT November 26, 1985 This ooain hu a dbmtb. d m May be _d by rIsp0 only is th peufomaue of ohk oehl dd. Its aulautg me a dw be dtdue withot W.di Bak mdhmwdloi.  CURRENCY EQUIVALENTS (End May 1985) USS1.00 - Rwandese franc (RwF) 103.7 RwF 100 - US$0.964 (Project data are based on the exchange rate of US$1.00 - RwF 105) FISCAL YEAR January 1 - December 31 GLOSSARY OF ABBREVIATIONS ADF African Development Fund BACAR Banque Contineatale Africaine au Rwanda BNR Banque Nationale du Rmanda BPI Bureau de Promotion Industrielle BRD Banque Rwandaise de Dsveloppement CCCE Caisse Centrale de Cooperation Economique CER Caisse d'Epargae du Rwanda EDF European Development Fund EIB European Investment Bank FAD African Development Fund KfW Kreditastalt fMr Viederaufbau HINIFiNECO Ministare des Finances et de l'Economie HINDMART MinistAre des Mines, de l'Industrie et de l'Artisanat OCIR-Th6 Office des Cultures Industrielles du Rwanda-The OCIR-Caff Office des Cultures Industrielles du Rwanda-Caff SSE Small-scale Enterprises SSI Small-scale Industries USAID US Agency for Iaternational Development WEIGHTS AND MEASURES Metric British/US Equivalents 1 meter tm) = 3.3 feet I hectare = 2.47 acres 1 are (100 m2) 0.02 acre 1 kilometer (km) 0.62 mile 1 are (100 m2) 0.01 hectare 1 square kilometer (km2) 0.39 square miles (sq. mi.) 1 kilogram (kg) = 2.2 pounds (lb) 1 liter (1) = 0.26 uS gallons (gal) 1 metric ton (m ton) 2,204 pounds (lb)  FOR OFFICUIL USE ONLY RWANDA FOURTH RWANDESE DEVELOPMENT BANK (BRD) PROJECT Credit and Project Summary Borrower: Republic of Rwanda Beneficiary: Rwandese Development Bank (Banque Rwandaise de D6veloppenent - BRD). Amount: SDR 8.5 million (USS9.0 million equivalent) Terms: Standard IDA terms Relending Terms: The borrower would onlend the credit proceeds to BRD at the following terms: (a) IBRD interest rate in effect at the time of distribution of credit docu- ments to the Executive Directors, with a flexible amortization schedule that would substantially con- form with the aggregate of the amortization schedules of subloans made by BRD and financed tinder this pro- ject (subject to a maximum term of 15 years, includ- ing a grace period not to exceed 5 years) on the portion of the line of credit allocated for medium- sized and larger projects CSDR 4.7 million); and (b) 7 percent p.a. with a fixed amortizatina schedule of In years. including 5 years of grace on the remaining SDR 3.8 million to be used for relending to small- scale enterprises (SSEs)I/. The weighted average cost of IDA funds to BRIi would he about 8 percent per annum. The Government would hear the foreign ex- change risk. Project Objectives and Description: (i) Objectives: The project would continue with IDA's support of BRD as an effective financial intermediary - and the main instrument in Rwanda for developing the private sector - to provide term finance for well-conceived development pro- jects, including SSEs. The project would assist the Government in improving the policy framework for SSEs, through the establishment of a special regime for them in the investment code and impro- vement of the functioning of the Special Guarantee Fund. The project would also assist the Govern- ment in its review of Rwanda's import tariffs sys- tem for necessary changes. I/ Small-scale enterprises (SSEs), as used in this report, are defined as enterprises with net total assets, hefore the project, of less than RwF 30 million (US$285,000). of their official duties. Its contents may not otherwise be disclosed without World Bank authorizaion.  (ii) (ii) Description: The project woud provide for a line of credit of SDR 8.5 million equivalent to BRD, of which about SDR 4.7 million equivalent would be used for financing medium-sized and larger enter- prises, and about SDR 3.8 millioa equivalent would be for SSE projects. Benefits and Risks: (i) Benefits. The project would provide investment funds to assist in developing Rwanda's private sector, including SSEs. It would help create about 1,500 jobs at an estimated average cost of US$10,000 per job. The Project would enable IDA to continue to support BRD in its efforts to improve its efficiency, financial strength and profitability. IDA would also help the Government in improving the policy framework for SSEs. Finally, IDA would initiate a dialogue with the Government on a reform of Rwanda's import tariffs system. (ii) Risks. The main risk in this project relates to possible deterioration of BRD's financial position due to an increase in its arrears. This risk, however, should not be overrated: (a) because BIRD has developed a sound program for assisting SSEs, the first phase of which has yielded encouraging results; and (b) because of the action BRD has already initiated to deal with its two major pro- blem projects and the measures proposed in the context of this project to improve BRD's collec- tion performance. - US$ millioa Local Foreiga Total Estimated Cost BRD's operations (1985-88) 9.6 18.9 28.5 Total Financing Required 9.6 18.9 28.5 Financing Plan BRD resources availahle for commitments 2.9 3.4 6.3 BRD - rolled over donor funds 6.7 - 6.7 Proposed fourth IDA credit - 9.0 9.0 ADF - 5.0 5.0 EDF - 0.4 0.4 Contribution by other donors/lenders - 1.1 1.1 Total 9.6 18.9 28.5  (iii) USs million (IDA FY) 1987 1988 1989 1990 1991 1992 1993 Estimated Disbursements Annual 0.4 0.9 1.9 2.0 1.6 1.6 0.6 Cumulative 0.4 1.3 3.2 5.2 6.8 8.4 9.0 Staff Appraisal Report No: 5791-RW  INTERNATIONAL DEVELOPMENT ASSOCIATION REPORT AND RECOMMENDATION OF THE PRESIDENT TO THE EXECUTIVE DIPECTORS ON A PROPOSED CREDIT TO THE RWANDESE REPUBLIC FOR THE FOURTH RWANDESE DEVELOPMENT BANK (BRD) PROJECT 1. I submit the following report and recommendation on a proposed development credit to the Rwandese Republic for the equivalent of SDR 8.5 million (US$9.0 million equivalent) on standard IDA terms for a fourth Rwandese Development Bank (BRD) Project. PART I - THE ECONOMY 2. A Country Economic Memorandum (Report No. 4059-RW) was distri- buted to the Executive Directors on May 20, 1983. The findings of this memorandum, updated as much as possible, are summarized below. Country data are provided in Annex I. 3. Rwanda's salient characteristics include its small size, an annu- al population growth rate of 3.7 percent (ranking among the highest in Africa), a population density (in terms of agricultural land) of about 390 per square km (slightly higher than that of India but about four times that of Zaire and about eight times that of Tanzania), hilly terrain and high average altitude, a landlocked position, lack of natural resources (includ- ing shortage of arable land), underdeveloped physical and institutional in- frastructure, and a very low level of development as measured by a variety of social as well as economic indicators. These indicators include a per capita income of about USS270 (1983), among the lowest in the world; an average life expectancy of 44 years; and an adult literacy rate of 40 per- cent. 4. The country is heavily dependent on agricultural exports (coffee, tea, pyrethrum, cinchona) which provide 75 percent of its foreign exchange earnings. Rwanda's manufacturing base is narrow, and the growth of modern manufacturing is limited by the small size of the market and the lack of raw materials, marketing facilities, entrepreneurial skills, and skilled manpower. The country, consequently, imports capital goods, steel, petro- leum products, cement and other construction materials, and virtually every modern consumer product. Its merchandise trade is hampered by high trans- portation costs and dependence on neighboring countries for access to the sea ports of Mombasa and Dar-es-Salaam. 5. Despite these constraints, Rwanda has made a creditable effort toward economic and social development. During the 1976-80 period, the country managed to satisfy its subsistence needs and to make important advances not only in agriculture but in other fields such as education, health, water supply, and small-scale industry. It succeeded in building  -2- basic transport and communication infrastructure and set in place a state administrative apparatus serious about development. These achievements reflect the Government's commitment to economic and social progress, as well as its prudent fiscal, balance of payments and debt management poli- cies. At the same time, Rwanda has been able to attract substantial ex- ternal aid from a great diversity of sources, confirming donor perceptions that Government is indeed development-oriented and is pursuing generally appropriate objectives. The Government's efforts during this period were significantly helped by favorable weather conditions and the coffee boom of the late 1970's. 6. Achievement and good fortune notwithstanding, development efforts in Rwanda have not alleviated the fundamental problems which continue to shadow development prospects: population growth, increasingly at odds with land availability, an undiversified economic structure, Government's li- mited capacity to provide social services, and the economy's inability to generate jobs either outside or within the rural sector. Though fooderop output has increased sufficiently to feed the growing population, it has apparently done so at a cost to the land's long-term fertility. Agricul- tural land use in Rwanda has reached a point where there is little fallow or grazing area left. The population pressure has resulted in serious deforestation, accompanying erosion, and has required the cultivation of marginal lands with a fragile ecology. Concomitant with these adverse trends has been a deterioration of nutritional standards induced bv the shift to traditional high-yield, high-calorie, but low-protein crops. The limited success of Government's efforts has been largely due to the country's structural constraints (among them the critical population pro- blem), institutional weaknesses (stemming largely from shortage of skilled personnel), and insufficiency of domestic financial resources. At the same time, fragmentation of institutional responsihilities, and poor interagency coordination have limited the country's capacity to absorb external resour- ces. 7. The Rwandese authorities have taken steps to address these pro- blems. Of particalar importance have been actions concerned with the ex- treme'y rapid population growth, and the lack of education and training. In 1981, the Government established the National Population Office (ONAPO) to plan, coordinate, and monitor all population activities. ONAPO has started to sensitize the population about the implications of excessive demographic pressure on the country's limited resources; a few pilot family planning programs have been set up; and a nationwide fertility survey, now completed, is expected to provide valuable information on the potential demand for family planning. Obviously, the impact of these measures will be felt only in the long run. To address the problem of lack of education and training, the Government introduced an education reform in 1979 which - after subsequent adjustment - was responsive to the country's needs, but whose implementation has been hampered by financial constraints and a lack of teachers. 8. A quantitative assessment of Rwanda's recent economic performance can only be tentative, as the aational account estimates have serious shortcomings. Nevertheless, these estimates suggest that. compared to the period of 1978-1980, during which the GDP grew by about 6 percent annually,  - 3 - economic growth in Rwanda slowed down during 1983-1984 - GDP grew on aver- age by 3.3 percent p.a. - mainly on account of a downswing in tertiary sector activities. Balance of payments and budgetary constraints impinged heavily upon trade and transport activities whose rapid growth in the past had been facilitated by greatly enlarged availabilities of foreign exchange and budgetary revenues. Budgetary austerity measures introduced since 1983 resulted in stagnation of Government's expenditures in real terms during the past two years. Notwithstanding this domestic recession, the manufac- turing sector expanded its productive capacity and improved its capacity utilization due, in part, to the import restrictions which limited foreign competition. Mining continued the decline that started in 1980 due to per- sistent financial and management problems of the major mixed-ownership com- pany (SOMIRWA) which is now being liquidated. In agriculture, the sluggish performance of the food crop production has beea partially offset by the rise in the production of export crops. 9. Rwanda's external position came under strong pressure, as the volume of imports continued to expand at a time when export earnings were declining. Since 1981, the total value of exports has been consistently below their 1977-80 levels, reflecting the decline in world coffee prices rather than a shortfall in export volume. The terms of trade deteriorated by about 24 percent between 1979 and 1984. In an attempt to arrest the de- terioration of the balance of payments, the Government adopted - beginning in 1983 - what they expected would be temporary measures, including licensing, prior deposit requirements and increased import tariffs. The restrictions aimed mainly at curtailing imports of non-esseatial consump- tion goods as well as those of locally produced commodities. The Rwandese authorities also shifted the peg of the Rwandese franc (in September 1983) from the US dollar to the SDR, entailing a 5.2 percent depreciation vis-a- vis the US dollar. These measures, coupled with some improvement in the terms of trade in 1984 (with respect to the 1981-1983 period) helped to reduce the current account deficit to about 9 percent of GDP in 1984, as compared to 12 and 11 percent, respectively, in 1982 and 1983. 10. On the budgetary side, the authorities failed to realize that the high level of tax revenues recorded in 1979 and 1980 was a temporary pheno- menon not justifying a permanent increase in expenditures. This failure reflected a major weakness in the assessment of resource availability and prospects. As a result, the sharp decline in revenues from coffee export taxes beginning in 1981, together with increases in current outlays (in order to achieve the social objectives of the Plan), resulted in an overall budgetary deficit equivalent to about 2 percent of GDP in both 1982 and 1983 (as compared to a surplus of 0.9 percent in 1980). Concern over the worsening budgetary situation led Government to introduce corrective fiscal measures. In 1984, the growth of current expenditures (in nominal terms) was limited to 2.3 percent (compared to 36 and 10 percent in 1981 and 1982, respectively). The growth of public sector employment was held under 0.5 percent - the education sector was exempted in order to achieve the re- vised targets of the 1979 Education Reform. This represented a major break from Government's past policy of practically guaranteeing employment to every secondary school graduate. Large reductions were also decided in net transfers to parastatals. Henceforth, state enterprises of a commercial or industrial nature will not be granted budgetary subsidies and will have to  -4- resort to commercial bank financing. On the revenue side, despite an up- ward revision of import duties on luxury goods and a more progressive tax structure on wages, the tax ratio to GDP remained relatively low at 9 per- cent. The same austerity measures have been included in the 1985 budget. 11. Government has been traditionally conservative in its monetary and credit policies. Inflationary pressures have arisen mainly from supply shortages caused by frequent disruptions of supply routes through neighbor- ing countries, high international transport costs, and increased prices of imports and seasonal fluctuations in the price of domestic foodstuffs. Inflation averaged 10.6 percent per annum during 1977-82, it peaked at 12.6 percent in 1982, reflecting mainly the wubstantial increases in electricity and water tariffs and educational fees. In 1983 and 1984, inflation was down to 6.6 and about 5.4 percent, respectively. 12. Rwanda has been one of the most favored beneficiaries of foreign aid ia receat years. Per capita disbursements of net official development assistance (ODA) have been above US$30, much larger than the average for Africa. During the period 1981-84, grants comprised about 80 percent of total external aid flows, and were provided mainly by Belgium, the Federal Republic of Germany, and France. In view of its level of development, Rwaada will need external assistance for a long time to come. Even main- taining per capita income at the current low level -- a very modest objec- tive indeed - will require a continuation of assistance at least at the current level, given population growth and export constraiats. The donor community should make efforts to increase its assistance, provided Rwanda continues to pursue sound macro-economic, sectoral (particularly in agri- culture), and population policies. 13. Rwanda's medium- and long-term external public debt is relatively Small, amounting to US$220 million at end-1983, equivalent to about 15 per- cent of GDP. Most of Rwanda's external debt was contracted on highly con- cessionary terms; the grant element was over 70 percent on average during the 1972-84 period. Due to the concessionary nature of these loans, the debt service payments on medium- and long-term debt are relatively low: 7.0 percent of exports of goods and nonfactor services in 1984. Hence, there remains scope for further borrowing. However, given the poverty of the country, its overwhelming constraints and vulnerability, and its unfavor- able terms-of-trade prospects, external funds should continue to he pro- vided in the form of grants or loans at highly concessionary terms, and ia- clude a high proportion of local cost financing and non-project assistance. PART II - THE BANK GROUP ASSISTANCE 14. The Bank Group assistance, started in 1970, was initially focused on the improvement of the road network and the strengthening of agricul- tural production. As of September 31, 1985, Rwanda has received twenty- IDA credits totalling US$247.2 million: eight for agriculture (37 per- cent); five for roads (28 percent); three for DFC's (7 percent); two for education (7 percent); one for telecommunications (3 percent); two for technical assistance (4 percent); one for water supply (5 percent) and two  -5- for power (10 percent). There have been no Bank loans. Three IFC invest- ments (one of US$535,000 for a tea factory; a second of US$226,000 with contingent equity commitment of up to US$60,000 to expand the tea factory; a third of USS249,000 also to expand the tea factory) were signed in 1976, 1980 and 1985, respectively. Annex II contains a summary statement of IDA credits and IFC investments. 15. In fiscal years 1982-84, disbursements for Rwanda totalled US$49.1 million, compared to new commitments of US$86.2 million. The aunual disbursement rate increased steadily over this period, to reach about 25 percent in FY85 which is above the average for countries of the Eastern and Southern Africa Region. There are no problem projects in Rwanda. 16. The Bank Group lending has been based on a country strategy which has emphasized: (i) agriculture and rural development, the main objective heing to increase food production as well as export crops, while maintain- ing soil fertility; (ii) human resources development, focusing on support to basic education and skills training to improve agricultural producti- vity, provide skilled manpower, and influence atritudes on the population issue, and, more recently, on family planning programs per se; (iii) infrastructure development, particularly roads, to reduce the country's isolation and to provide incentives to further intensification of agriculture as well as increased specialization and diversification through better marketing; (iv) energy, to lessen the country's demand for fuel imports; and (v) development of small and medium-size enterprises in manufacturing and other sectors. 17. One of the major constraints to Rwanda's development is the shortage of technical/managerial capacity. This affects all sectors and inhibits project preparation and implementation. Intensive technical assistance and on-the-job training of Rwandese staff have, therefore, been a salient feature of the Bank Group program for Rwanda, both under indivi- dual projects in various sectors, and through a free-standing technical assistance project (Credit 1217-RW which became effective in November 1982). This project has helped strengthen inter-ministerial coordination, provided training to staff in the studies units of concerned agencies, and contributed to improve the preparation of the next five-year plan. 18. The Rank Group program has been pursued in a cjLaate of good relations with the Government. In all sectors in which we are active, there exist a general receptivity to our advice and willingness to take action where required. In agriculture, improved sector management as well as more rigorous project planning and financial controls have been attained, with a shift away from integrated rural development schemes to- wards more directly productive projects. In the highway sector, our active dialogue with the Ministry of Public Works a-.d Energy is now focusing on the ways to reduce reliance upon expatriate expertise. In industrial deve- lopment, we have gained Government's commitment to promote small-scale enterprises. In the education sector, we had difficulties due to two misprocurements (some six years ago) and an overly ambitious education reform proposal at the primary level. However, a positive dialogue with the Ministry of Primary and Secondary Education (MINIPRISEC) has emerged in  - 6 - the past three years; the Ministry has since adopted a less costly and lengthy primary cycle. On population, our initially cautious approach to Government has developed into cooperative efforts to promote suitable family health and planning programs. 19. The Bank Group strategy continues to center on the five sectors mentioned above (para. 16), with special emphasis on agriculture and popu- lation. For the former, we give priority to: (i) promote intensification by developing and strengthening Rwanda's agricultural research and exten- sion capabilities; and (ii) reinforce the key sector institutions - the Rwandese Research Institute and the Ministry of Agriculture. In the popu- lation sector, our program involves close collaboration with the National Population Office (the agency responsible for designing and promoting educational programs and service delivery systems for family planning), support to maternal/child health initiatives through the Ministry of Health, recently reorganized, and to participate more effectively in other efforts to deal with Rwanda's bourgeoning population problem. 20. Since the mid-1970's, the Bank Group has been active in the industrial sector in an effort to assist Rwanda's small private sector which has been developing rapidly under a stable environment and a growing economy. BRD has been the main instrument for IDA's intervention in this field. Under our first and second BRD credits (approved in FY77 and FY79, respectively), the main objectives were to help improve BED's project appraisal capability, internal organization and accounting procedures. These objectives have now been largely achieved (para. 44). With a third BRD credit (approved in FY83), we began to help BRD and Government in starting a concerted effort to assist SSEs, including improvements to BRD's SSE appraisal and loan processing procedures, and funds to support a SSE pilot program by BRD. The credit also provided technical assistance to help Government to carry out a study of issues facing SSE development, to prepare the way for a more substantial IDA contribution at a later stage. We also sent a mission to Rwanda to review the prospects for manufacturing development in the country, as a basis for fruitful dialogue with the Government on development issues in this increasingly important sector in the economy of Rwanda (para. 45). 21. Our dialogue with the Government has also had a macroeconomic dimension. Rwandese authorities have recently indicated readiness to pro- ceed with the preparation of an economic recovery program (-Programme de Relance'), and asked the Bank's assistance in its formulation. A Bank eco- nomic mission recently visited Rwanda in response to this request. The mission's work, which extended in particular over agriculture and mining sectors and public enterprises, will help further strengthen the coopera- tion between Rwanda and the Bank at the macroeconomic level.  - 7 - Part III. THE MANUFACTURING AND FINANCIAL SECTORS A. The Manufacturing Sector Structure and Performance 22. Despite the predominantly rural nature of its economy, Ibranda has a relatively significant manufacturing sector, representing about 17 per- cent of GDP and employing some 40,600 people. During the period 1977-83, the sector expanded at an average rate of 8.7 percent p.a. in real terms, with little structural change. The sector remained mostly composed of artisans and small enterprises producing commodities, such as banana wine, bricks, baskets, wood furniture and other basic goods for the domestic market. There are now a total of only 80-100 medium and large enterprises with more than 30 employees, including some 20 public and mixed-ownership enterprises. They generate close to 40 percent of industrial value-added and provide some 14,000 jobs, or about 35 percent of the sector's total, the rest being generated by SSEs (i.e., about 60 percent of industrial value-added and about 65 percent of industrial employment). 23. Modern manufacturing is concentrated mainly in Kigali, the capi- tal, wbere infrastructure is relatively more developed, although some towns in the regions are also gaining importance. With the exception of food processing, modern manufacturing relies mainly on imported inputs. Produc- tion is typical of the first generation of import substitation, with ex- ports confined largely to processed coffee, tea, pyrethrum and a few other products. Most firms enjoy a relatively high level of effective protection and some have the benefits granted hy the investment code (paras. 33-34). 24. Since 1977, a significant change has taken place in ownership pattern of modern industry. Out of 22 industrial projects completed during 1977-82, 13 became fully or majority state-owned. At present, about a third of the capital in modern manufacturing belongs to the public sector. The Government's increased intervention in manufacturing resulted from its desire to step up economic growth by increased investment in industry, to promote companies of nationeml interest, and to utilize bilateral aid. However, most of these interventions were poorly planned and implemented. Many ventures b'-came part of the administration when completed, as no structure was created to operate them. The Government recently decided to reduce its direct involvement in industry. It is trying to reorganize and improve the operation of public industrial enterprises, with a view to finding private investors to share in the ownership or to acquire some ventures in full. 25. The SSE sector is diverse and dynamic and has, compared to other African countries, developed without much institutional assistance. Manufacturing represents only about 25 percent of SSE activity, with agriculture, commerce and other services accounting for the rest. Virtually all SSEs are owned by Rwandese nationals and are either family businesses or cooperatives. Exceptionally, few SSEs are limited liability companies. Unlike modern manufacturing, SSEs are spread countrywide. They are productive, labor-intensive and less dependent on imported inputs than many modern enterprises.  2b. For better perspective in this subsector, including its con- straints and potential, IDA financed a comprehensive study on SSEs in Rwanda under its third BRD credit (para. 20). The work, carried out by BRD and the Centre Africain de Statistiques et d'Economie Appliquee, covered 220 enterprises, 146 of them located in rural areas and 74 in Kigali and the other three major cities of Rwanda. The study, which was submitted to the Government in December 1984, revealed considerable development poten- tial for SSEs in Rwanda. SSEs are less constrained by the small size of the local market and there is ample scope for development in such diverse areas as food processing, agricultural tools, livestock, construction materials, etc. However, the sector also faces severe constraints. More than modern enterprises, SSEs suffer from a lack of incentives, limited availability of training facilities, and difficult access to bank credit, although they appear to have substantial borrowing capacity as their debt/ equity ratios are generally low. The project proposed in this report will address to resolution or alleviation of some of these problems (para. 33- 3b). The USAID has also been active in Rwanda's SSE sector and current and prospective USAID operations in this field will contribute to the attain- ment of some of the proposed IDA project's objectives. Notably, a recently approved Rwanda private enterprise development project by the USAID will offer training facilities from which SSEs may benefit. Another USAID pro- ject, concerning -economic program initiative for Rwanda-, will include support for policy reforms related, inter-alia, to investment code, Special Guarantee Fuad, price controls and import tariffs. During preparation of the proposed IDA project, we have had close coordination with the USAID, and the IDA project will benefit considerably from the improved policy framework for private investments which the current USAID project is aiming for. Industrial Policies and Issues 27. Responsibility for the formulation and implementation of indus- trial policies in Rwanda is shared by four agencies. The Ministry of Mining, Industry and Artisanat (MINIMART) has the overall responsibility for industrial policy formulation, but mainly concentrates its efforts on industrial promotion, including SSEs, and on issuance of industrial licences. It also plays a role in the granting of benefits under the investment code and supervises the Bureau for Industrial Promotion (BPI), created in 1977 to prepare industrial projects and to assist enterprises in difficulty. The Central Bank (BNR) controls import licensing and all foreign exchange transactions, and sets interest rates charged by most financial institutions. The Ministry of Finance and Economy (MINIFINECO) influences industrial development through taxes and import tariffs and approves the benefits proposed by MINIMART under the investment code. The Planning Ministry (MINIPLAN) coordinates the preparation and follows the execution of the country's industrial investment program. 2s. To regulate and assist the industrial sector, the Government relies essentially on the same policy instruments used by most other African countries, but applies them with a certain liberalism, allowing some competition from imports and often amongst local producers. Recent measures taken to protect the country's foreign exchange reserves have, however, increased the level of protection accorded to local industry (para. 29).  -9- 29. Import licencing. All imports are subject to licencing, but licences are usually granted without difficulty. Restricting imports of competing goods to protect local producers has not been Government policy, though, in a few cases, imports have been temporarily banaed. However, in an effort to reduce the level of imports, the Government iatroduced in March 1983 a scheme requiring importers of certain goods to place a 100 percent local currency deposit with BNR at the time the import license is issued. The deposit earns interest, but since imports take four to six months to arrive, this measure imposes considerable strain ona the liquidity of importers. Goods covered by the regulation include those carryiag high import duties as well as most goods produced locally. Industrial inputs are not affected. Applications for import licences have fallen since the scheme was introduced. 30. Import tariffs. Imports are subject to a customs duty varying from 5 percent to 20 percent, a fiscal duty of up to 130 percent of the c.i.f. value, and a bonded storage tax of 3 percent of the c.i.f. value. A preferential rate of 1 percent is applied to items exempted from import duties. The tariff structure shows a familiar pattern: exemptions or low duties on raw materials, intermediate and capital goods, as well as con- sumer goods not produced locally; and high rates on consumer goods and luxuries. In general, duties increase with the stage of processing. 31. The large differences ia duty rates result in widely different rates of protection to various manufacturing activities. The fact that Industrial inputs attract low duties (or are exempted) provides little incentives to develop local substitutes and may discourage iadustries that could be viable in Rwaada. On the other hand, high duties on consumer goods encourage local production that is economically inefficient. The Government is aware of the need for changes in the tariff system and has initiated a comprehensive study of the tariff structure with a view to in- troducing better incentives for local producers. The study is expected for completion by early 1986. The CGoernment, by July 1, 1986, will review with IDA the study's recommendations and its proposed timetable for introducing necessary changes to Rwanda's import tariffs structure. 32. Price controls. In March 1983, the Government introduced a form of ex-ante price control under which all prices should be approved before goods are sold. Manufacturers can claim a net margin of 15 percent, while traders are allowed a gross margin of 15 percent (wholesale), or 25 percent (retail). However, the system is not applied strictly. While importers are required to submit price information to the Commerce Department, their proposed prices are approved automatically within 24 hours. With regard to local manufacturers, only two enterprises having a special agreement with the Government have had their prices reviewed by the administration. Twenty others (all large enterprises) submit the information requested, but are, in fact, free to sell their products at prices they deem reasonable. SSEs are not affected by this ruiling as the Commerce Department does not require tnem to provide price information. As applied now, the system operates merely as a price registration procedure and does not constrain importers and manufacturers.  - 1u - 33. Investment Code. The code offers eligible investors the follow- ing guarantees and fiscal advantages: (i) repatriation of interests, divi- dends, and the original capital invested at the official exchange rate; (ii) exemption from import duties on imported inputs and equipment; (iii) tax holidays for five years, and reduced taxes for the subsequent two years; and (iv) a wide range of benefits in the case of projects that are of special importance to the country and involve large investments, includ- ing, inter alia, relief from various duties and taxes for a period of up to 15 years, restrictions on competing imports, and guarantee of Government orders. 34. To qualify for the advantages, a project should have a minimum investment of RwF IU million (US$95,250) for local iavf.stors and RwF 20 million (US$190,500) for foreign investors. Twenty eight enterprises (all large enterprises) have benefitted from the code. The SSE study (para. 20) shows that only 15 percent of the surveyed SSEs fulfill the minimum in- vestment requirement and that 94 percent had never heard about the code. Thus, to be of value for large number of SSEs, certain amendments have to be made to the code. For SSEs, the benefit package should emphasize the tax exemption aspects, as many small enterpreneurs are less interested in the other benefits: they do not repatriate profits; they seldom import their equipment and other inputs directly, and they resort to limited bank borrowing. Some medium-sized enterprises are also constrained in their access to benefits under the code. Aware of these special needs, the Government has recently set up an interministerial committee to recommend on a special regime within the code for small and medium-sized enter- prises. The committee's work is fairly advanced; its report and recommen- dations will be submitted to the Council of Ministers for approval by June, 1986. The special regime would emphasize the tax exemption aspects of the benefits package and set clear eligibility criteria for small and mediua- sized enterprises. 35. Special Guarantee Fund. A Government action in favor of SSEs has been the creation of a Special Guarantee Fund to help enterprises unable to meet normal security requirements to obtain loans from financial institu- tions. Contributions to the Fund come from a 10 percent levy on the pre-tax profits of participating institutions, which comprise BRD and com- mercial banks. To obtain a guarantee, a request must be submitted, accom- panied by a substantial amount of documentation, to an interministerial committee composed of 7 members and chaired by a Director General of the MINIFINECO. Because of its cumbersome procedures, the Fund has not been functioning well. Over the last six years, only a handful of guarantees were issued and the more than RwF 1bU million (USS 1.5 million) contributed to the Fund have remained largely idle in a BNR account. 36. The Government intends to amend the existing Policy Statement and Procedures for the Fund to allow for: (i) clearer eligibility criteria, to reserve the Fund benefits for promoters of productive projects in agricul- ture, fisheries, industry and mining; (ii) simplified requirements for documentation by applicants; and (iii) automatic payments out of the Fund, up to 80 percent of any principal arrears, when a loan is more than one year in arrear. These changes are adequate and should improve the func- tioning of the Fund.  - 11 - B. The Financial Sector Present Set-up 37. Rwanda's financial system consists of the followvng institutions: BNR, three commercial banks (Banque Commerciale du Rwanda, Banque de Kigali, and Banque Continentale Africaine, Rwanda - BACAR), a network of cooperative banks (Banques Populaires, created in 1975 with Swiss aid), a Savings Association (Caisse d'Epargne du Rwanda), a mortgage bank (Caisse Hypothecaire) and a development bank (BRD). With the exception of commer- cial banks, all other institutions are totally, or to a significant extent, Government owned. 38. Financial institutions are supervised by BNR which sets strict criteria for their lending operations, including the following: institu- tions accepting deposits from the public will have, at all times, to main- tain their equity to at least 10 percent of their liabilities. Equity participations by banks will not, unless approved by BNR, exceed 10 percent of the bank's net worth and 20 percent of the company's share capital; the financial institutions' exposure limit to any single enterprise is set at 25 percent of their net worth, which may not he exceeded without BNR's express authorization. However, BRD's lending policy is not subject to BNR regulations, hut is set by its Board. 39. Commercial banks are active mainly in import and export financing and overdrafts to local companies and their medium-term lending (^^e to four years) is limited to a maximum determined by their net worth plus deposits of more than a year's duration, minus fixed assets, equity parti- cipations, doubtful debts and losses. Among commercial banks, Banque Com- merciale du Rwanda is the most active in medium-term lending, holding about 30 percent of total medium-term credit outstanding at the end of December 1984. Because of the short-term nature of their resources, commercial banks require BNR authorizion before engaging in long-term lending (more than four years) which is the domain of institutions such as BRD and the Caisse Hypothecaire. BRD is the most important long-term lender, account- ing for 72 percent of the RwF 2,712 million (US$25.8 million) long-term credit outstanding at the end of December 1984. The rest is mostly held by commercial banks which are allowed by BNR to provide long-term housing finance, given the still relative inactivity of Caisse Hypoth&caire in this field. Availability of Finance for Industry 40. In general, BNR ceilings assigned to commercial banks have not prevented them from meeting manufacturing firms' short-term credit needs. However, term credit is more limited and mainly provided by BRD through lines of credit from foreign donors, notably IDA. 41. SSEs, nevertheless, have more difficulty than other ventures in finding hank credit for their development. According to the SSE study (para. 20), less than 25 percent of the enterprises surveyed had obtained a bank loan, although about a half of the total bave had at one time or another applied fo-r credit. Credit to SSEs is not well-organized or deve-  - 12 - loped. Short-term credit, mainly to small enterprises in commerce, trans- port and construction, is provided to a limited extent by commercial banks, and by the Banques Populaires. The Banques Populaires are also active in term lending, but most of their loans have been to individuals for housing improvements and car purchases. Expansion of credit for directly produc- tive activities is hindered hy the Banques Populaires' lack of appraisal capability and the high administrative costs of processing and supervising a large number of small loans. Commercial banks are reluctant to engage in term lending to SSEs because of the high risks involved. Uowever, with pending improvements in the operation of the Special Guarantee Fund (para. 36), it is expected that commercial banks will be more active in SSE lending. Since 1982, BRD has been giving attention to SSE lending and, with support of this proposed project, would substantially improve the availability of term credit for them. Interest rates 42. Interest rates by lending institutions (except BRD) are regulated by BNR. Following a substantial upward revision in November 1979, interest rates remained unchanged, except for minor modifications. The maximum rates for term deposits of one month to one year vary from 7.5 percent to 8.75 percent and range from 8.75 percent to 11.87 percent for savings bonds of one to five years maturity. These rates include a tax equivalent to 20 percent of the rate to be paid by the banks to the Treasury. On the lend- i-ag side, short-term credits for general imports and overdrafts, which represent the bulk of commercial banks' lending, carry rates of 14 percent and 14.5 percent, respectively. Other lending rates are 4 percent for tea exports, 5 percent for mining exports and 9 percent for coffee collection and imports of essential goods. The lending rate for productive invest- ments approved by BNR is 11 percent. Other term loans carry rates between 12 percent and 14 percent. In view of the present and projected rates of inflation (about 6.0-6.5 percent), all the rates are positive in real terms, excepting those on short-term credits for tea and mining exports. Foreign Exchange System 43. Rwanda has a relatively open economy compared to those of some other African countries. The foreign exchange system is virtually free of restrictions on current international transactions. Repatriation of dis- tributed dividends is allowed and resident expatriates can transfer up to 60 percent of their net annual income. Manufacturers can obtain foreign exchange from BNR for imports of equipment, inputs and spare parts. With the approval of BNR, foreign investors can repatriate the capital invested after sale or liquidation of their business ventures. C. IDA Lending Strategy 44. The project completion report on the first ERD project, prepared in 1983, concluded that BRD has successfully emerged as a sound development finance company capable of appraising medium-sized development projects and has become the main support for the development of the private sector in Rwanda. The execution of the second and third BRD projects, which will further strengthen BRD, has been progressing satisfactorily.  - 13 - 45. The Third IDA credit for BRO helped BRD and Government in start- ing a concerted effort to assist SSEs, which included simplifying and improving BRD's SSE appraisal and loan processing procedures, funds to sup- port BRD's SSE pilot program and a study of issues facing SSE development in future. The IDA manufacturing sector report (para. 20), which was dis- cussed with the Government in June 1985, confirmed the importance of assisting SSEs and identified a number of issues for particular attention (paras. 27 to 3b). IDA's development strategy in the industrial sectol is now aimed for: (a) continued institution-building at BRD as it is going through a major reorientation of its operations towards SSE and agricultural projects financing; (b) transfer of increased resources through BRD to assist development of private sector in Rwanda; (c) improvement of the policy framework for SSEs to include: (i) improvement of the operation of the Special Guarantee Fund (para. 36); (i_) simplification of administrative procedures for esta- blishing new enterprises; and (iii) establishment of a special regime for SSEs and medium-sized ventures in the investment code (para. 34); Cd) strengthening of the Government's capacity to analyze sector deve- lopments and to formulate investment programs for industrial sector; and (e) assisting the Government in reforming the incentives system to include: Ci) establishing clear criteria for access to benefits under the investment code and improving its administration (paras. 33-34); (ii) abolishing the 100 percent import deposit requirement for imports competing with local manufactures (para. 29); iii) adopting a realistic exchange rate system; and (iv) establishing a more neutral incentives system for all manufacturing firms. 46. These IDA objectives may be attained through IDF operations, technical assistance, dialogue with the Government and, if warranted, policy-based lending. The proposed project would make a significant con- tribution towards IDA's attainment of the foregoing development objectives in the sector. PART IV. THE PROJECT 47. The proposed project would be the fourth IDA operation in support of BRD. An appraisal mission visited Rwanda in February-March 1985. Nego- tiations were held in Washington from October 1b to November 8, 1985; the Rwandese delegation was led by Mr. Theodore Mpatswenumugabo, Director of Study Bureau, MINIFINECO. A detailed description of the project components is presented in the Staff Appraisal Report No. 5791-EW, dated October 31,  - 14 - 1965, being circulated separately to the Executive Directors. Supple- mentary project data are included in Annex III of this report. A. THE PROJECT INSTITUTION Background 48. The Rwandese Development Bank (Banque Rwandaise de Developpement - BRD), which would be the beneficiary of the proposed IDA credit, was established by the Government in 19b7 as a limited liability company to encourage creation and development of enterprises in Rwanda, to promote diversification of the country's economic structure, and to help attract foreign investment into the country. To fulfill these objectives, BRD has been authorized to grant term loans or guarantees and make equity partici- pation in enterprises. It may receive deposits from enterprises it assists, administer funds, issue notes, and borrow in Rwanda or abroad. The institution is autonomous. Its investment decisions are made on the basis of staff appraisal results deriving from financial, economic and technical merits of projects proposed for its financing. Capital Structure 49. BRD's share capital consists of -A7 shares (which must account for at least 55 percent of total share capital and are reserved for the Government and public Rwandese agencies and institutions), and 'B' shares (designated for private and foreign shareholders). There are no differ- ences between the rights and privileges of either category of share- holders. Since 1967, five share capital increases have taken place, including a raise in June 1983, when the authorized capital was set at RVF 1.12 billion (US$10.7 million). Ownership of the share capital is now distributed as follows: public sector 55.1 percent, domestic private sector 12.5 percent, and foreign institutions 32.4 percent (including French, German, Dutch, Belgium and Japanese participations). Management, staffing and organization 50. BRD's Board of Directors consists of 13 members, seven of whom represent the "A' shareholders, and six the "B" shareholders. The Presi- dent of the Republic appoints the Board Chairman. The Board meets on a regular basis to determine BRD's policy and to approve equity participa- tions, guarantees and loans exceeding RwF 3 million (US=28,571). 51. BRD's Managing Director, a Rwandese national, is capable and well respected by the Government and the business community in Rwanda. In the day-to-day management, he is assisted by two competent Rwandese directors, one in charge of BRD's Department of Finance and Administration and the other responsible for its Investments Department. There are also two expa- triate technical assistants: (i) a management specialist recruited in 1984, who works in BRD's project supervision division and advises clients on management matcersj and (ii) a highly experienced development banker who has worked closely in various capacities at BRD since 1975 and has been instrumental in helping make BRD the strong financial institution it now is. The latter's contract, which is financed under German technical  - 15 - assistance, will expire at the end of 1985. The Government will seek an extension. 52. BRD's staff, 90 in total, including thirty-nine professionals, has almost doubled within the last three years in line with the expansion of BRD's operations. Salaries are adequate and there is little staff turn- over. BRD's staff training program relies mainly upon practical on-the-Job training by the expatriate development bhnker and other BRD senior staff. Such training has yielded good results. However, with the increasing com- plexity of the operations, BRD's staff will also need external training to keep them abreast of developments in their fields of expertise. Under an expanded training program, to be financed from BRD sources, BRD will have three staff members per year to participate in seminars and development finance-related courses of up to three months each outside Rwanda for the next two to three years. 53. BRD's Investment Department is responsible for project appraisal (including SSEs), supervision and technical assistance; the Administration and Finance Department is in charge of disbursements, loan collections, and relations with foreign donors. In addition, a new Studies and Development Department is now being organized to: (i) focus on agriculture and sector studies as a basis for new operations; (ii) undertake macroeconomic stu- dies; and (iii) examine the development impact of BRD-financed projects. BRD's increased activity and exposure in the regions have resulted in a wide geographical dispersion of its client base. BRD, therefore, plans a phased establishment of small regional offIces (initially two) to provide assistance to its clients in the interior of the country. Operating Policies and Procedures 54. BRD's policy statement, which was amended in May 1985, places emphasis on analysis of economic impact of projects and reflects the re- orientation of BRD's lending towards SSEs and agriculture. It provides for BRD's avoidance of the foreign exchange risk on loans from foreign donors and clarifies its procurement and dividend policies. The statement is satisfactory. 55. BRD has made significant progress in improving the quality and standard of its appraisal work. Projects are now more thoroughly analyzed financially and technically and, in some cases, redesigned to ad;, t them better to conditions in Rwanda. Under the second and third IDA credits, BRD has undertaken to calculate economic rates of return for all non- services sector projects above the free limit of US$200,000. As of June 30, 1985, the 63 projects financed under these credits, which extended over several economic sectors (39 percent in industry, 38 percent in agriculture and the remaining 23 percent in services, including tourism), were all *udged to he viable with economic and financial rates of ret-ura in most cases ahove 15 percent. Individual project size varied from USS10,000 to US$250,000 equivalent for SSE operations, with an average cost per job of about USS7,000, and up to USS4 million equivalent for larger enterprises, with an average cost per job of USS26,000 equivalent. Although 3RD has progressed in strengthening economic standards of its appraisal work, how- ever, there is still room for improvement. The project completion report,  - 16 - which was prepared in 1983 on the first IDA project (para. 44), underlined the need for improvement in economic analysis of projects. BRD's expanded training program (para. 52) will give priority to staff training in this area. BRD's supervision procedures are satisfactory, with staff visiting each project at least twice a year. BRD's policies regarding foreign exchange risk, interest rates, procurement, disbursement and audits are discussed in paras. 74, 80, 81 and 83 below. BRD Lending to SSEs 56. Until 1982, BRD required SSEs to provide the same information asked from larger enterprises as a basis for its lending. Many SSEs could not comply and their applications were not processed. According to a BRD review of its experience with SSE financing (conducted in 1981), during the six year period ending in 1981, BRD received an average of 63 SSE loan applications per year amounting to about RwF 140 million/year, but pro- cessed only 15 percent of them. Staff constraint and a higher priority given to processing of larger loans have also played a role. 57. The lower priority given by BRD to SSE financing was due to the higher administrative costs involved and bad debt losses associated with such lending in the early years of the SSE progam. The BRD's review of SSE financing showed that, at March 31, 1982, most of the arrears over three months was on account of BRD's SSE clients. However, this analysis also showed that several factors could alleviate the negative financial impact of SSE lending. First, more experience in the sector should lead to improved repayment performance through better criteria for selecting projects and promotors. Second, streamlining of loan processing procedures for SSEs should hold down administrative costs. Third, an increase in BRD lending rates should help improve its average spread and profitability. Fourth, access by BRD to an effective loan guarantee scheme could help minimize the financial impact of bad SSE loans. 58. To expand its SSE lending operations, BRD is implementing a two- phase program. In the first phase, which has been supported under the third IDA project, the main emphasis remains on: (i) processing of large backlog of SSE loan applications which were not investigated in the past; and (ii) adapting BRD's 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 in the Investment Department. Supervision of SSE projects was reinforced. BRD also in- creased 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, using counterpart resources from KfW lines of credit. In the second phase, which is expected to start in late 1985, BRD intends to take a more active, promotional role with emphasis on assistance to enterprises. Preliminary assessment by BRD of its SSE pilot program shows encouraging results. BRD has succeeded in decreasing its SSE-related administrative costs from 7.9 percent of SSE portfolio in 1983 to 6.1 percent in 1984 and plans to reduce them further to 4.5-5.0 percent in 1985-86. An increase in BRD's lending rate to 12.5  - 17 - and 13.5 percent for all loans, as agreed with BRD (para. 74 (ii)), would also help improve the profitability of SSE lending. Operations 59. As of December 31, 1984, BRD had approved loans and equity in- vestments totalling RwF 3.8 billion (US$36.2 million). Virtually all of BRD's financing has been to private (or majority private) enterprises. While most of the equity investments were made before the mid 1970s, loan approvals became important starting in 1974 and have since risen rapidly to reach RwF 616 million (US$5.9 million) in 1984. Long-term lending now represents between 95 percent and 98 percent of the total approvals. Over the past six years, the BRD operations have shown certain trends which underline: {i) an increase in loan maturities from medium towards relati- vely longer terms; (ii) a decline in the average size of projects due to increased SSE financing; (iii) a more evenly balanced sectoral mix among the projects financed; and (iv) a broadening of the geographical distribu- tion of the BRD-financed projects. bO. During 1982-84, BRD approved 129 projects with a total cost of RwF 2.8 billion (USS26.7 million), of which RwF 1.4 billion (US$13.3 mil- lion) was financed by BRD. Manufacturing accounts for about 32 percent of the BRD project approvals, and agriculture/livestock for about 22 percent. In general, BRD (total) lending to a single enterprise may not exceed 20 percent of its own net worth. BRD's loan approvals have steadily increased from 1981 to 1984. The rate of increase in loan disbursements was even faster between 1981 and 1983. Disbursements in 1984 declined somewhat, mainly due to completion of certain investments in previous years. How- ever, an increase in rate of disbursements is expected in 1985 and subse- quent years. Portfolio and Provisions 61. As of December 31, 1984, BRD had an equity portfolio of RwF 335 million (US$3.2 million) in 14 companies and a loan portfolio of RwF 1,975 million (US$18.6 million) in 130 projects. The majority (77 percent) of BRD's projects were totally owned by private Rwandese entrepreneurs. BRD's total arrears of more than three montbs at the end of 1984 amounted to RwF 138.4 million (US$1.3 million equivalent), or 7 percent of its total loan portfolio. They affected 39 projects (more than half in manufacturing and mining) representing 19 percent of the value of the toral loan portfolio. Forteen and a half percent out of the latter ratio comprised the overdue principal and interest payments on the SOMIRWA and RWAKINA project ac- counts1/. The remainder was in agroindustry and other sectors, excepting 1/ Two large investment projects, RWAKINA and SOMIRWA, account for 76 per- cent of the BRD portfolio affected by arrears, and for 69 percent of its actual arrears. The companies involved were set up with participa- tion by French and Belgian interests to produce pharmaceuticals and to engage in mining, respectively. Due to various managerial, financial and technical problems, both companies have been incurring losses; SOMIRWA is now being liquidated.  - 18 - transport sector projects which have had no arrears. In 1984, BRD allo- cated out of its profits provisions of RwF 107.4 million (US$1.0 millioa) against its claims from SOIRIWA and RWAKINA accounts. It intends to make further provisions in 1965. Also, it has general reserves of RwF 64.8 mil- lion (US620,VOOU), representing 3.3 percent of the outstanding portfolio. b2. Although the overall quality of BRD's portfolio is good and ade- quate provisions have been set aside for probable losses from doubtful investments, BRD's collection performance has not improved adequately over the last few years when the collection ratio remained under 100 percent. In 1984, of the RwF 305.9 million (US$Z.9 million) in principal billed, only 89 percent (US$2.6 million) was collected. If the collection ratio remains at this level, arrears are bound to increase at a time when BRD is reorienting its lending activity towards SSEs, a sector with potentially difficult collection problems. Against this background, a major objective of the proposed IDA project would be to assist BRD in improving its arrears position. At the credit negotiations, BRD has agreed to raise progressi- vely its collection ratio to 105 percent by 1988. IDA will closely monitor BRD's collection performance, including its progress in implementing appropriate action programs related to the SOMIRWA and RWAKINA accounts. Financial Results and BED's Financial Position b3. Over the last five years, BRD's income from loans showed fast growth, at an annual compounded rate of 22 percent, mainly as a result of a doubling of the loan portfolio between 1980 and 1984, and an increase in BRD's loan interest rates (2 percent on average) over the same period. In 1984, BRD's total income was 15 percent above the 1983 figure, but net pro- fits were substantially lower (RwF 22.1 million - or US$210,000 equivalent in 1984, compared to RwF 72.5 million in 1983) primarily due to record pro- visions related to the RWAKINA and SOMIRWA accounts (para. 61). The 1984 provisions amounted to 5.7 percent of the average loan portfolio, up from 2.5 percent in 1983. Financial charges as a percentage of the average debt increased from 4.1 percent in 1983 to 4.7 percent in 1984 and reflected the increasing cost of BRD's borrowings. Administrative expenses were,however, tightly controlled, and amounted to 2.3 percent of the average total assets in 1984, down marginally from 2.5 percent in 1983. b4. BRD is a financially sound institution. Over the past 5 years, its term debt/equity ratio has never exceeded 1:1. At December 31, 1984, it stcod at 0.9:1.0, well within the 3:1 limit previously agreed with IDA. BRD's debt service coverage ratio is also good at 2.U. The Government bears the foreign exchange risk on BRD's term debt which comes from foreign sources. This is appropriate for a small, relatively unsophisticated DFC which lends, inter alia, to SSEs. BRD has substantial equity resources and 47 percent of all its assets were financed from net worth. Its loan port- folio is adequately protected, with the necessary provisions for probable losses. Lending Resources 65. As of December 31, 1984, BRD had the equivalent of US$2.1 million available for new loan approvals, of which US$110,500 equivalent was in  - 19 - local currency and US$2.0 million in foreign exchange. The low level of local currency resources should improve with the expected collection of USS3.0 million equivalent in loan repayments in 1985, of which only about 75 percent has to be immediately reimbursed to lenders. In addition to its equity, BRD's local resources include loans from BNR and Government, two investment funds and retained earnings. 66. BRD's foreign exchange resources consisted of eight lines of credit from four donors (KfW, CCCE, IDA and EIB), totalling the equivalent of US$12.6 million, and a large credit line provided by the BNR from an IMF Trust Fund Loan. Interest rates on the foreign loans range from 0.75 per- cent to 6.9 percent and the maturities from 9 to 50 years, with grace periods of up to 10 years. The CCCE loan is tied to French procurement and was intended to finance a specific project at 9 percent interest rate. The BNR line of credit has relending rates of 8-12 percent depending on the maturity of subloans and the sector. All other loans are relent at BRD- established rates and carry no restrictive conditions on procuremeat. Prospects 67. Projected Operations. BRD's project pipeline for the next two years exceeds by far its resource availability for that period. These projects include 23 medium and larger lending operations with a total investment cost of about US$70 million equivalent and numerous SSE ventures. Although the level of possible BRD financing is yet to be decided for a number of large projects, BRD has already been approached for lending amounting to about US$14 million equivalent. With regard to SSE lending, BRD expects to at least maintain the level of approvals reached in 1983-84 of US$1.5 million equivalent per year. 68. Despite the strength of its project pipeline, BRD plans to main- tain a cautious approach to its approvals, particularly because it recog- nizes the potential risks in proceeding too fast with SSE operations. Because of this and the fact that the forecasts are based on the operations of a particularly active year (1984), the BRD projections assume a steady but cautious 5 percent nominal growth ia loan approvals over the next 3 years. On that basis, total approvals would grow from US$6.7 million equivalent in 1985 to USS7.6 million equivalent by 1988. Commitments over the same period (also the commitment period of the proposed IDA credit) would total US$28.5 million equivalent of which about 66 percent would be in foreign exchange (para. 79). Disbursements over the period would be about USS26.5 million equivalent. These RRD forecasts are realistic. 69. Projected Financial Condition and Financial Results. BRD's financial forecasts are made in current terms and assume an increase in the average cost of BRD's new borrowings from about 7 percent at present to about 8 percent p.a., taking into cnasideration the anticipated costs of external borrowing. In the face of increasing resource cost and given the necessity to continue making substantial provisions against certain invest- ments (para. 63), BRD will increase its interest rate on loans from about 12 percent on average at present to 12.5 percent for directly productive projects and 13.5 percent for service-type projects. This wotild permit BRD to have a spread of about 5 percent on its new borrowings (para. 74).  - 20 - 70. On the basis of the above adjustments, BRD's net income should resume a steady rise from its low point in 1984 (para. 63) to reach about US$610,000 equivalent in 1988. Although the latter is still low in real terms compared to past years, it conceals the underlying financial strength of BRD, due to its conservative provisions policy (with reduced future risks of loss and ample cashflow gains in the process). The projections also indicate that BRD's financial structure will remain sound. Over the period 1985-88, BRD's debt to equity ratio should not exceed 1.2 to 1, well within the 3 to 1 limit previously agreed with IDA, and a debt service coverage ratio of above 2 at all times. BRD's asset/liability structure is expected to retain its current favorable characteristics (i.e. debt liabi- lities will continue to have much longer maturities than loan assets). This should contribute to BRD's financial strength, to allow the institu- tion to remain liquid over the forecast period, and enable it to overcome any portfolio difficulties in the foreseeable future. B. THE PROJECT Project Objectives and Justification 71. The proposed IDA credit would represent a continuation of the Bank group'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. Specifically, it would: (a) support BRD as the institution moves more forcefully into the areas of SSE and agricultural lending; (b) improve BRD's capabilities for economic evaluatioa of projects; (c) provide the needed foreign exchange resources for priority investments by private entrepreneurs; and (d) help improve the policy framework for SSEs and develop indigeneous entrepreneurial capabilities in Rwanda. The support for SSEs would be particularly valuable to the economy of Rwanda as this sector offers good potential to alleviate Rwanda's pressing unemployment problem. Project Content 72. Under the proposed project, IDA would lend SDR 8.5 million (US$9.0 million equivalent) to the Government of Rwanda, which would onlend the proceeds to BRD under a subsidiary loan agreement. Execution of a satisfactory subsidiary loan agreement would be a condition of the IDA credit effectiveness. BRD would use the IDA credit proceeds to finance the foreign exchange cost of its eligible. subloans, allocating SOR 4.7 million equivalent to finance medium-sized and larger projects and SDR 3.8 million equivalent for SSEs. The maximum loan size for SSE projects would be US$230,000 equivalent. Relending of Credit Proceeds 73. Selection of Subprojects. IDA credit proceeds would be used to fisance the foreign exchange cost of capital investments, including initial working capital, in subprojects approved by BRD. All projects that are within BRD's scope of operations would be eligible, provided they are of  - 21 - high economic priority, selected by BRD in accordance with its investment priorities and justified on the basis of a full economic analysis, includ- ing calculation of an economic rate of return for all non-service sector projects costing above US$200,000. SSE projects would be selected and appraised according to BRD's simplified procedures (para. 58). 74. BRD Borrowing and On-Lending Terms. Since IDA would be providing nearly a half of BRD's foreign exchange resource needs during 1985-88, BRD's borrowing and on-lending rates under the proposed project would weigh heavily on its future profitability. The following are the rates which BRD would use for its borrowing and on-lending operations: (i) Government onlending terms and foreign exchange risk. Consistent with the Bank Group policy for IDF operations, IDA funds for re- lending to medium-sized and larger enterprises (RDR 4.7 million) would be passed on by Government to BRD at IBRD's interest rate prevailing at the time of the credit documents' distribution to the Executive Directors. However, recognizing the higher admi- nistrative costs aoA risks associated with SSE lending, funds for relending to SSEs (SDR 3.8 million) would be onlent by Goverament at a rate of 7 percent, so as to afford BRD a greater spread. The foreign exchange risk would be borne by the Government. The average cost of IDA funds for BRD would be about 8 percent p.a., compared to about 7 percent p.a. under the third IDA BRD pro- ject. This average cost would be positive in real terms as inflation in Rwanda is estimated at about 6-6.5 percent p.a. during the next few years; (ii) BRD's lending rates. BRD would further simplify its current interest rates structure which consists of six rates reflecting term (medium and long-term loans) and sector (agriculture/live- stock, industry and services) differentials of its projects. There would be two rates, one for projects in directly productive sectors, to he set at 12.5 percent p.a., and the other for the service-type projects which would have a rate of 13.5 percent p.a. The new rates would represent an increase of 0.5-2.5 per- centage points over the existing rates and would be highly posi- tive in real terms. They would yield an average spread of 5 percent for BRD to cover administrative costs, to make adequate provisions, and to allow profit. In future, the level of BRD's interest rates would be reviewed with IDA at least once a year to ensure its adequacy and competitiveness compared to alternative sources of finance; and (iii) Commitment charge. As under the previous IDA credits, BRD would pay to Government a commitment charge of 0.75 percent p.a., from the respective dates of subproject approval by IDA. 75. Free limit. The individual subproject free limit would be ia- creased from USS200,000 equivalent under the third IDA credit to US$300,000 equivalent, in recognition of the good quality of BRD's appraisal work. The aggregate free limit would be USS3.5 million equivalent, or about 39  - 22 - percent of the total amount of the line of credit. As most SSE projects would be below that free limit, IDA would review in detail the first 15 SSE subprojects to be financed by BRD. 76. Amortization schedule. In conformity with the Bank Group policy for IDF operaLions, funds for relending to medium-sized and large enter- prises (SDR 4.7 million) would have a flexible amortization schedule that would substantially conform with the aggregate of the amortization sche- dules of the subloans approved by BRD and financed under the proposed credit, subject to a maximum period of 15 years, including a grace period not to exceed five years. Given the large number of small loans to be made under the SSE credit allocation (SDR 3.8 million), this portion of the IDA credit would have a fixed amortization schedule of 10 years, including five years of grace, starting from the date of credit effectiveness. 77. Commitment period. The funds woule be available for commitment by BRD until December 31, 1988.' BRD's Total Project Costs and Financing 78. Total commitments by BRD during the IDA project period (1985-88) are estimated at US$28.5 million, of which US$18.9 million would be in foreign exchange. The proposed IDA credit would meet US$9.0 million equi- valent (about 48 percent) of the foreign exchange needs and the remainder would be secured from other sources. Assuming that BRD would finance, on average, about 50 percent of the total investment costs of medium-sized and larger projects and about 80 percent of the costs of SSE projects, the US$28.5 million forecast commitments would result in about US$53 million in total iavestments. 79. The financing plan for BRD's operations during the project period would be as follows: Financing Plan, January 1985-December 1988 (US$ million equivalent) Local Foreign Total I. BRD's comaitments 9.6 18.9 28.5 II. Resources Availab'le for commit- 2.9 3.4 6.3 ments (as of December 31, 1984) Resource Gap 6.7 15.5 22.2 III. To be financed by: - BRD - rolled over donor funds 6.7 - 6.7 - EDF - 0.4 0.4 - ADF - 5.0 5.0 - IDA - IV - 9.0 9.0 - Other donors/lenders 1.1 1.1 Total financing 6.7 15.5 22.2  - 23 - Project Implementation B0. Procurement. Although the responsibility for selecting suppliers lies with project sponsors, BRD requires its clients to submit at least three quotations to ensure that the items to be purchased are suitable for the planned operation. In cases where substantial construction is in- volved, BRD usually requires competitive bidding. BRD also requires that equity funds are paid out first. Disbursements are made against presenta- tion of invoices or relevant documentation, with direct payments being made to equipment suppliers whenever feasible. Payouts are checked by follow-up site visits at various project stages. These procedures are acceptable. 81. Disbursement. The proceeds of the proposed credit would be dis- bursed on the basis of: (i) 100 percent of the c.i.f. cost of goods and services for eligible subprojects; (ii) 80 percent of the cost of previous- ly imported equipment purchased locally for the subprojects; (iii) 70 per- cent of the cost of equipment produced in Rwanda from imported components and raw materials; and (iv) 65 percent of the local cost of construction works included in subprojects and carried out by locally-based contrac- tors. Disbursements are expected to be completed over seven years, by December 31, 1993. 82. Special Account. To expedite disbursement of funds, a special account (revolving fund) would be set up at BNR, or another financial institution acceptable to IDA, into which IDA would make an initial deposit of US$300,000 equivalent from the proposed credit immediately after the credit effectiveness. This amount represents an estimated average disbur- sement of funds over a three-month period. Withdrav'al of funds for indivi- dual sub-loans not exceeding US$25,000 equivalent would be made from this account on the basis of statements of expenditures. The documentation would be reviewed by supervision missions and by independent auditors. Disbursement of funds under the technical assistance component and for in- dividual subloans in excess of US$25,000 equivalent would be fully docu- mented. The special account would be replenished when funds are below US$150,000 equivalent. 83. Audits and Reporting. As under the previous IDA projects, BRD will continue to have its accounts and financial statements audited by independent auditors acceptable to IDA and will furnish to IDA certified copies of its audited financial statements and its annual reports within six months of the end of the fiscal year. BRD will also submit to IDA quarterly reports, which would include, inter alia, financial statements, resource position, statement of arrears, collection data and notes on major projects in difficulty, including the conclusions and recommendations of BRD's supervision staff. Benefits and Risks 84. It is estimated that the proposed IDA credit would support in- vestments totalling about US$15.0 million equivalent and would help create more than 1,500 new jobs in the Rwanda economy at an average cost per job of USS10,C00 equivalent. The project would enable BRD to expand its sup-  - 24 - port of SSEs and agriculture. Continued IDA support would also enhance BRD's capability to mobilize additional foreign exchange resources from other aid agencies and enable it to play a more significant role in the development of Rwanda economy. Finally, the IDA project would play an importamt role in initiating a dialogue with the Government on a reform of Rwanda's import tariffs structure and in improving the policy framework for SSEs. 85. The main risk in this project relates to possible deterioration of BRD's arrears position. This risk, however, should not be overrated: (i) because BRD has developed a sound program for assisting the SSE sector, the first phase of which has yielded encouraging results; and (ii) because of the actions BRD has already initiated to deal with its two major problem projects (para. 61) and the measure proposed in the context of this project to improve its collection performance (para. 62). PART V - RECOMMENDATION 86. I am satisfied that the proposed Credit would comply with the Articles of Agreement of the Association and recommend that the Executive Directors approve the proposed Credit. A. W. Clausen President Attachments Washington, D.C. November 26, 1985  -25 - Annex I iiSLU . Page 1 of 6 3MI,D - OCtA leicayg Oa^t IU!! EIDI IUinUCZ G3OIPS (MiLsU?gO *aVZ3ctS1 mm (a mts uc U ?sbin 1m In&. - ,'oLk ,usTum,i d Sm of SWARA I, 1. ot SrAi mu (lemma me. m ToraL a36*3 303 26.1 ZiOiLIMUIL 1A.0 15.2 13.3 mw cl uhrr (u) *- .- 270.0 238.3 306 . I amosmina r3 CAPIT (tELOC,5r OF OIL . ;v _ALEN-) 10.0 11.0 13.0 62.1 Sol.S POUULiTIow AM viaL zran Cs poPUL.AID.3C0-rt" taTw0sh"si 2732.0 3603.0 s6n.0 IoA 00301A210 11 Of TOTAL) 2.4 3.2 *4 20.1 2.0 POPLAII0U Pm..JgctIoIs POPsATtW to tElt 2000 HTILL) 10.2 rT naTa