Tunisia - Third, Fourth, and Fifth Societe Nationale d'Investissement Projects

Document of The World Bank FILE COPY FOR OFFICIAL USE ONLY Report No. 3611 PROJECT PERFORMANCE AUDIT REPORT TUNISIA--BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE (LOANS 648-TUN, 798-TUN AND 881-TUN) September 9, 1981 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT TUNISIA--BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE (LOANS 648-TUN, 798-TUN AND 881-TUN) TABLE OF CONTENTS Page No. Preface ......................................................... i Basic Data Sheet (Loan 648-TUN) ................................. ii Basic Data Sheet (Loan 798-TUN) ............................ iii Basic Data Sheet (Loan 881-TUN) ...................*....... ... iv Highlights .-................................................. .. v PROJECT PERFORMANCE AUDIT MEMORANDUM .............. 1 ATTACHMENT: PROJECT COMPLETION REPORT I. Introduction ......... ....................... . . 5 II. Background on Third, Fourth and Fifth Bank Loans .. . . . . . ...................... 5 III. Evolution of the Economic Environment in Tunisia .. .. . .. .... . .... . .. .. .7 IV. Institutional Performance ........................ 9 V. Summary and Conclusions ............................... 14 Annexes 1. Analysis of the Bank Loans and of SNI's Operational and Financial Performance during the Period 1970-1977 ........................ 15 2. Table 1: Cumulative Disbursement End of Calendar Years ........ .............. 21 Table 2: Sub-project Data ........ ........... 2 Table 3: Lending and Investment Operations ........... 23 Table 4: Balance Sheet Statements: Projections Compared with Actual .................. 24 Table 5: Income Statements: Projections Compared with Actual .............. 25 Table 6: Financial Ratios: Projections Compared with Actual ..................... 26 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.  - 1 - PROJECT PERFORMANCE AUDIT REPORT TUNISIA--BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE (LOANS 648-TUN, 798-TUN AND 881-TUN) PREFACE This report presents a performance audit of loans 648-TUN, 798-TUN and 881-TUN to the Societe Nationale d-Investissement (SNI) of Tunisia. Late in 1973, SNI was renamed Banque pour le Developpement Economique de la Tunisie (BDET). The latter name is retained in the audit memorandum; both names are used interchangeably in the project completion report (PCR). The three loans under review were the third, fourth and fifth Bank loans to BDET; they were approved in November 1969, January 1972 and January 1973, respectively, and closed in November 1974, December 1977 and March 1978 with total cancellations of US$2.20 million, i.e., 6.5 percent of their total amount of US$34 million. A performance audit report covering experience under the second loan was issued in May 1974 (Sec. M74-351). The attached PCR was prepared by the Bank's Europe, Middle East and North Africa Regional Office on the basis of information provided by BDET and of a country visit in October 1979. It provides an account of the use of Bank funds under the three loans and of achievements towards meeting the institu- tion building objectives set at appraisal. The memorandum comments on BDET's difficulties in meeting its development objectives while remaining suffici- ently profitable to attract private capital. No comments were received from the borrower.  - ii - PROJECT PERFORMANCE AUDIT REPORT TUNISIA--BANQUE DE DEVELOPPEMENT ECONQMIQUE DE TUNISIE THIRD DEVELOPMENT FINANCE COMPANY PROJECT (LOAN 648-TUN) BASIC DATA SHEET Amounts (in US$M) As of 02/28/81 Original Disbursed Cancelled Repaid Outstanding Loan 10.00 9.30 .70 8.39 .91 Cumulative Loan Disbursement 1970 1971 1972 1973 1974 (1) Planned n.a. n.a. n.a. 9.0 10.0 (ii) Actual 1.9 4.9 8.0 9.0 9.3 (iii) (ii) as % of (i) - - - 100 9.3 PROJECT DATA Original Actual or Loan Date Re-estimated ioard Approval 11/25/69 oan Agreement 12/24/69 - .ffectivenes 03/31/70 05/05/70 Loan Closing 12/30/73 11/30/74 MISSION DATA Month/ No. of No. of Man- Date of Item Year Weeks Persons weeks Report Appraisal 03/69 - - - 11/10/69 Supervision I L . 02/71 - - - - Supervision II 11 08/71 - - - - Supervision IIL/a 07/72 - - - - Supervision IV 11/73 2 2 4 01/11/74 Supervision V 05/74 0.2 1 0.2 07/01/74 Supervision VI 09/74 0.2 1 0.2 09/11/74 FOLLOW-ON PROJECTS BDET IV, Loan No. 798-TUN, approved January 1972, in the amount of US$10.0 million. BDET-V, Loan No. 881-TUN, approved January 1973, in the amount of US$14.0 million. BDET VI, Loan No. 1189-TUN, approved January 1976, in the amount of US$20.0 million. Industrial Finance - BDET Component, Loan No. 1504-TUN, approved December 1977, in the amount of US$30.0 million. /a Combined with appraisal mission of new loan to BDET.  - iii - PROJECT PERFORMANCE AUDIT REPORT TUNISIA--BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE FOURTH DEVELOPMENT FINANCE COMPANY PROJECT (LOAN 798-TUN) BASIC DATA SHEET Amounts (in US$M) As of 02/28/81 Original Disbursed Cancelled Repaid Outstanding Loan 10.00 9.14 .86 4.65 4.49 Cumulative Loan Disbursement 1972 1973 1974 1975 1976 1977 (i) Planned 0.6 4.8. 8.8 9.9 10.0 10.0 (ii) Actual 2.5 4.4 6.6 8.7 9.0 9.1 (iii) (ii) as % of (i) 417 92 75 88 90 91 PROJECT DATA Original Actual or Loan Date Re-estimated Board Approval 01/18/72 Loan Agreement 02/09/72 - Effectivenes 05/22/72 04/13/72 Loan Closing 03/31/76 12/31/77 MISSION DATA Month/ No. of No. of Man- Date of Item Year Weeks Persons weeks Report Preappraisal . 02/71 - - - - Appraisal 08/71 - - - 12/28/71 Supervision I /a 07/72 - - - - Supervision II 11/73 2 2 4 01/18/74 -Supervision III 05/74 0.2 1 0.2 07/01/74 Supervision IV 09/74 0.2 1 0.2 09/11/74 Supervision V La 06/75 1 1 1 08/31/75 Supervision VI 07/76 2 3 6 07/13/76 Supervision VIL 03/77 2 3 6 04/08/77 Supervision VIIT/a 06/77 2 3 6 07/06/77 Supervision IX 07/78 1 3 3 07/14/78 FOLLOW-ON PROJECTS BDET V, Loan No. 881-TUN, approved January 1973, in the amount of US$14.0 million. BDET VI, Loan No. 1189-TUN, approved January 1976, in the amount of US$20.0 million. Industrial Finance - BDET Component, Loan No. 1504-TUN, approved December 1977, in the amount of US$30.0 million. /a Combined with appraisal mission of new loan to BDET.  - iv - PROJECT PERFORMANCE AUDIT REPORT TUNISIA-BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE (LOAN 881-TUN) BASIC DATA SHEET Amounts (in US$M) As of 02/28/81 Original Disbursed Cancelled Repaid Outstanding Loan 14.00 13.36 .64 8.51 4.85 Cumulative Loan Disbursement 1972 1973 1974 1975 1976 (i) Planned 0.8 6.5 12.8 13.9 14.0 (ii) Actual 1.2 9.8 11.8 13.2 13.4 (iii) (ii) as % of (i) 150 151 92 95 96 PROJECT DATA Original Actual or Loan Date Re-estimated Board Approval 01/23/73 Loan Agreement 02/20/73 - Effectivenes 05/21/73 05/24/73 Loan Closing 03/31/78 06/30/79 MISSION DATA Month/ No. of No. of Man- Date of Item Year Weeks Persons weeks Report Appraisal 07/72 2 - 12/29/72 Supervision I 11/73 2 2 4 01/18/74 Supervision II 05/74 0.2 1 0.2 07/01/74 Supervision III 09/74 0.2 1 0.2 09/11/74 Supervision IV /a 06/75 1 1 1 08/31/75 Supervision V 07/76 2 3 6 07/13/76 Supervision VI 03/77 2 3 6 04/18/77 Supervision VIL/a 06/77 2 3 6 07/06/77 Supervision VIII 07/78 1 3 3 07/14/78 Supervision IX FOLLOW-ON PROJECTS BDET VI, Loan No. 1189-TUN, approved January 1976, in the amount of US$20.0 million. Industrial Finance - BDET Component, Loan No. 1504-TUN, approved December 1977, in the amount of US$30.0 million. /a Combined with appraisal mission of new loan to BDET.  - v - PROJECT PERFORMANCE AUDIT REPORT TUNISIA--BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE (LOANS 648-TUN, 798-TUN AND 881-TUN) HIGHLIGHTS The three loans under review were the third, fourth and fifth Bank loans to the Banque de Developpement Economique de Tunisie (BDET). Besides addressing BDET's resource gap in foreign exchange, the three loans sought to help improve the institution's appraisal and supervision capa- bilities and management structure. BDET's operations have increased rapidly over the years, Bank funds representing an average of 16 percent of total disbursements over the period 1970-1977. The borrower's financial performance also improved notably over the period, actual results being somewhat better than projected at appraisal. Following a reorganization in 1972, the institution's appraisal capabilities improved markedly. In comparison, its supervision and promotion activities have shown slow progress; the latter will have to be stepped up in the future if BDET is to maintain its growth momentum and make a significant contribution to the development of small-scale industries, export-oriented enterprises, new entrepreneurs and projects located outside Tunis. Overall, the experience under the loans points to the difficulties private DFCs may have in meeting their development objectives while remaining sufficiently profitable to attract private capital. Other points of interest are: - the differences of opinion between BDET and its auditors regarding the size of provisions for bad loans and doubtful equity investments (PPAM, para. 8; PCR, paras. 4.12 to 6.18); - BDET's successful resource mobilization efforts (PPAM, para. 6); and - the concentration of sub-projects in Tunis (PPAM, para. 3; PCR, Annex 1).  -2- the loan proceeds went to hotel projects. The 149 sub-projects financed by the Bank created about 8,000 jobs at an investment cost ranging from D6,500 (US$15,000) to D24,400 (US$58,000) per job created. The capital-labor ratio increased over the years with the growing size of the sub-project with which it was found to be correlated, being however highest for hotel projects. The Bank's financial contribution reached 55 percent of BDET's investments in 1972 and averaged 16 percent of total disbursements over the period 1970-1977. 4. During the years 1970 to 1976, BDET's overall approvals and dis- bursements increased fivefold from D6.7 million to D30.6 and from D3.9 million to over D17.9 million, respectively. BDET-s financial performance, as shown in the following table, improved also notably, actual performance being somewhat better than projected at appraisal (PCR, Annex 2, Table 6): Financial Ratios (Percent) 1970 1977 As percentage of average total assets: Gross Income 6.6 8.3 Administrative Expenses 1.5 1.5 Financial Expenses 3.7 4.8/a As percentage of equity: Profit before tax 6.5 13.7 Debt-Equity Ratio: 3.1:1 5.9:1 /a 1976 figure. 5. During negotiations for the fourth loan, the Bank expressed its concern over the deterioration of BDET-s management and procedures which had taken place in 1969 and 1970 as a result of staff turnover (paras. 4.01 to 4.03 of the PCR). Major efforts were then made to strengthen BDET; a new management team was appointed in 1971 and the institution reorganized. An appraisal department was created, which was instrumental in improving BDET-s appraisal capability. Under the sixth Bank loan, BDET started calculating economic rates of return on sub-projects. BDET's supervision and promotion activities, however, have not shown a similar improvement. The Bank's efforts to persuade its borrower to set up separate supervision and promotion depar- tments have not been successful. Because of a low profits margin and lack of appropriate expertise, BDET was not in a position to take up promotion seriously. As a result, over the years, its contribution to the promotion of small scale industries, export-oriented enterprises, new entrepreneurs and projects located outside Tunis (with the exception of hotels) was modest, and PROJECT PERFORMANCE AUDIT MEMORANDUM TUNISIA: BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE (LOANS 648-TUN, 798-TUN AND 881-TUN) 1. Loans 648-TUN, 798-TUN, and 881-TUN were the third, fourth and fifth loans made by the Bank to the Banque de Developpement Economique de Tunisie (BDET)1/. The three loans, for a total amount of US$34 million, were approved in November 1969, January 1972 and January 1973, respectively. Besides addressing BDET's resource gap in foreign exchange, the three loans under review sought to help improve the institution's appraisal and supervi- sion capabilities and management structure. New economic policies had been announced by the Government in 1969 and the Bank felt that, as these policies were being implemented, BDET would be in a better position than in the past to stimulate private investment in industry and tourism, and hence contribute to the improvement of the balance of payments, employment creation, and regional development in Tunisia. 2. The liberal policy measures announced in 1969 resulted during the 1970s in a more open export-oriented economy in which private initiative was able to play an increasing role. As part of this liberalization program, the Government simplified its procedures for investment licensing; import restrictions were gradually relaxed to introduce more competition in the domestic market; moreover, the role of the banking system in allocating investment funds was strengthened. In the years 1970-1977, real GDP grew at an average annual rate of 8.6 percent, almost twice as fast as during the 1960s. Undeniably, the country's economic performance benefitted from favorable weather conditions which resulted in good crops, as well as from increased world market prices for phosphates and petroleum products. Never- theless, fast growth in industrial investment, to which private investors increasingly contributed when they realized that the Government's past nation- alization policy was being set aside, was another important element in this improved performance. Tourism remained the favorite sector for private investors until 1973 and 1974 during which years tourist arrivals declined, only to regain their upward trend in 1975. Although reduced from 35 percent to 25 percent of the labour force between 1970 and 1977, unemployment remained the major problem facing the Tunisian economy. 3. Annex I of the PCR presents a review of the utilization made of Bank funds. Altogether, some 149 sub-projects were financed under the three loans, sub-loans amounting to US$215,000 on the average. 64 percent of the sub- projects financed under the third and fourth loans and 77 percent of those financed under the fifth were located in Tunis while, overall, 29 percent of 1/ The Banque de Developpement Economique de Tunisie was until 1973 known as Societe Nationale d'Investissement. -3- over many years, BDET maintained its growth momentum by financing a dispropor- tionate amount of tourist facilities until the Bank insisted on restrictions. However, since the mid-1970's, BDET has become somewhat more active in project promotion. Nonetheless, with increasing competition from commercial banks in providing foreign exchange term credit for equipment imports and the recent creation by the Government of the Agence de promotion des Investissements to promote high priority industrial projects, BDET will have to step up further its promotional capabilities if it wants to maintain a prominent position in the financing of Tunisia's industrial sector. 6. In recent years BDET has taken successful steps to increase its capital and obtain official foreign exchange resources outside the Bank to finance its fast growing operations. However, BDET-s efforts to raise local resources through disposal of equity investments, bond issues or attrac- tion of deposits from the public have been less successful; as a result it has had to rely on the Central Bank's rediscount facilities and on undue conver- sion of its foreign exchange resources, a practice which has been criticized by the Bank. BDET's liquidity position was under strain in 1974-75 because of arrears in sub-loan repayments and more recently because of an upsurge in loan disbursements. Greater attention should be paid by both BDET and the Bank in the future to BDET's cash flow projections. 7. Although BDET has made some effort to assist in the development of the local capital market, the private sector in Tunisia is apparently not yet ready for the spread of share-ownership. Most companies, especially the smaller ones, are indeed reluctant to open up their accounts and give out information even to public auditors. Corporate laws have not provided for sufficient measures to safeguard minority rights; nor is the recently created Stock Exchange (1970) sufficiently experienced to do this either. The per- centage of BDET shares owned by Tunisian individuals has been, over the years, reduced from around 33 percent to less than 20 percent. With frequent share increases and the slow development of profitability and dividend distri- butions, it was not expected that private shareholders would maintain their relative position; this is however an indication that the appeal of BDET shares to local investors has somewhat diminished as well as a reflection of there being little inducement for the public to participate in the share market. 8. The relationship between BDET and its auditors has, at times, been strained, causing concern in the Bank (paras. 4.12 to 4.18 of the PCR)1/ The main problems have been the size of annual provisions for bad loans and doubtful equity investments. BDET was reluctant to agree to a reduction in its profit position, since it felt that in the long run collection of most of the arrears outstanding was feasible and that the total value of its equity 1/ BDET audit is an area where the Bank's institution-building effort has been extensive and has produced successful results in strengthening the local auditing profession (PCR, para. 4.18). portfolio would remain above cost. Although the audit rules involved on the valuation of equity investments were not subject to question, the Bank staff took, after some discussions, a more lenient view, siding with BDET managementl . It would probably have served BDET's profitability as well as its accounting procedures better, had BDET been able to dispose of some of its equity investments in its portfolio, realizing book profits and avoiding controversy with its auditors. On a number of other issues, the Bank sided with BDET's auditors and, overall, should be credited for having been success- ful in convincing its borrower of the benefits to be derived from complying with external audit procedures; progress is only partial, however, as audit reports are distributed to shareholders on too short notice before their annual meeting to permit any meaningful discussions to take place (PCR, para. 4.17). 9. After a few years of rapid development, BDET has faced a number of somewhat conflicting objectives. To maintain its growth momentum it has had to employ, train and keep highly qualified and well-paid staff, make adequate provision for doubtful debts, entering untried business sectors, while lending at moderate rates to meet the development objectives of the Government; yet as a private institution, it has had to remain solvent and be able to dis- tribute sufficient dividends to keep its commitments to its shareholders and be able to increase its capital. Discussions between the Bank and the Government have centered on these conflicting interests. However, much of the difficulties in meeting both sets of objectives could have been avoided if, at the very beginning, BDET had been able to obtain a larger subordinated loan free of interest. The difference in terms of lower financial charges would have covered much of BDET's developmental expenses and avoid the subsequent disputes between the Government, BDET, its auditors and the.Bank group. 10. In spite of the foregoing, BDET's performance has been at a high level and the Bank's efforts in providing it with finance and in improving its organization and procedures over time have enabled it to contribute signifi- cantly to the development of Tunisia in the fields of industry and tourism. 1/ BDET's argument was that reserves based on a case-by-case assessment of its equity portfolio would be unnecessarily large since unrealized book profits more than compensated for doubtful investments. The Bank staff considers that the position it took at the time was reasonable, even if it conflicted with the auditors' view. ATTACHMENT PROJECT COMPLETION REPORT ON SOCIETE NATIONALE D'INVESTISSEMENT BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE I' LOANS 648-TUN, 798-TUN, 881-TUN I. INTRODUCTION 1.01 The present report (PCR) covers the three following Bank loans that were signed respectively on December 24, 1969 for loan 648-TUN (SNI-III), on February 9, 1972 for loan 798-TUN (SNI-IV) and on February 20, 1973 for loan 881-TUN (SNI-V). SNI-III and IV were for $10 million each and SNI-V was for $14 million. The second loan was subjected to a performance audit and the report was submitted to the Board in 1974 (No. 445 dated May 15, 1974). 1.02 The Soci6tg Nationale d'Investissement (SNI) was established as a semi-public institution in 1959. Its objective was to promote investments and to compensate for the scarcity of private capital and the reluctance of private investors to undertake industrial ventures. Between 1959 and 1965, SNI invested most of its D2 million capital in the equity of new or expanding firms with slow growth prospects. In 1965 SNI requested the Bank Group's assistance to increase its equity base and to transform itself into a DFC making term loans as well as equity investments. The Bank granted its first loan to SNI for $5 million in May 1966; at the same time, IFC took a 20% participation in SNI's capital of D.1.5 million. A second Bank loan of $10 million was made in September 1967. II. BACKGROUND ON THIRD, FOURTH AND FIFTH BANK LOANS A. Third Bank loan (SNI-III) 2.01 By early 1969, SNI had committed $6 million out of the second Bank loan. A Bank appraisal mission visited Tunisia in March 1969. The loan was signed on December 1969, and became effective on March 31, 1970. The interest rate was 7%, the maturity was 17 years, and the loan amount $10 million. 2.02 SNI III had two main objectives: (a) to fill a resource gap, and (b) help SNI improve its appraisal and supervision capability and its management structure. At the time of appraisal (March 1969) SNI had a resource gap in foreign exchange of $16.2 million on a commitment basis for the period 1969-1971. SNI-III was designed to cover this resource gap for approximately eighteen months beginning in 1970. Since SNI's resource position in dinars was also extremely tight, the Bank allocated $1 million out of SNI-III to finance local expenditures. Bank assistance was not only financial but also institutional because appraisal capability, supervision activity and organization all needed strengthening. High turnover among the staff and changes in SNI's top management in 1969 also required close attention (para 4.05). SNI's appraisals, though weak on the market analysis side, were considered thorough and of good qcua1ity for a relatively young institution. For this reason, the free-limit was.raised from $100,000 to $200,000. 1/ The name of SNI was changed to BDET in 1973, and both are used interchangeably in this report. 2.03 SNI was considered creditworthy overall. In 1969, only 2.5% of SNI's loan portfolio was affected by serious arrears in large part however because SNI's loans were recent: -SNI started making loans only in 1966 and about 10% of them were covered by Government guarantees. At the Bank's urging, SNI started to make provisions for doubtful loans in 1969. Debt servicing was not a problem. A share capital increase was planned for late 1970. Because of SNI's good overall financial situation, the ceiling on the debt- equity ratio was raised from 3.3:1 to 4:1. 2.04 The main covenants in the Loan Agreement were: (a) a free-limit of $200,000; (b) financial assistance to public enterprises limited to 35% of the loan amount. The foreign exchange risk was assumed by the Government. B. Fourth Bank Loan (SNI-IV) 2.05 The fourth Bank loan to SNI was appraised in August 1971, signed in February 1972 and became effective on April 13, 1972. The interest rate was 7.25%, the maturity was 14 years, and the loan amount $10 million. 2.06 The main objectives for SNI-IV were the same as for SNI-III; to fill a resource gap and to strengthen the institution, with major emphasis on the latter. At the time of appraisal, 94% of the previous Bank loan was committed and SNI projected a resource gap in foreign exchange of $29.1 million for 1972-1973. Therefore, SNI needed fresh resources quickly. However, the appraisal mission recommended that the Bank loan cover SNI's commitments only for 1972 because progress made in improving SNI's organization and appraisal procedures had to be reviewed before considering further lending. 2.07 The debt-equity ratio was only 2:1 at the end of October 1971 because SNI had doubled its share capital to D3 million in January 1971 (para 2.03)and the debt-service coverage was comfortable. The debt-equity ratio limit was, therefore, left unchanged at 4:1. But the staff turnover, especially at the middle management level, caused some deterioration in the quality of project appraisal in 1970-1971. The proportion of the portfolio affected by arrears rose to 16.3% at the end of 1970 compared to 3% in 1968, and, for the first time, SNI's auditors seriously qualified their opinion on the 1970 accounts (para 4.12).Changes in SNI's top management compounded the difficulties linked to the staff turnover. A new president and general manager, Mr. Bourguiba Jr., was named in March 1971, who in turn appointed a new deputy general manager, Mr. Fakhfakh. (para 4.05). 2.08 The main covenants in Loan Agreement 798-TUN were: (a) a free- limit left unchanged at $200,000, which was equivalent to a reduction in constant terms, and the introduction of an aggregate free-limit equal to 25% of the loan amount; (b) a reduction of the portion allowed for public enterprises' financing to 25% of the loan amount, down from 35% in SNI-III (para 2.04). -7- C. Fifth Bank loan (SNI-V) 2.09 A bank mission visited Tunisia in July 1972 to assess progress made in correcting organization and appraisal weaknesses observed during appraisal of the previous loan. A new $14 million loan to SNI was presented to the Board in January 1973 and became effective on May 24, 1973. The interest rate was 7.25% and the maturity was 20 years (para 2.11). 2.10 The main justifications for SNI-V were similar to those for the two previous loans; to cover a resource gap in foreign exchange after other funds had been obtained, and to continue the institutional building assistance to SNI. A resource gap of $13.5 million for 1973-1974 was projected, assuming that the Government would allocate to SNI $2.6 million from a Swedish loan and $2.9 million from the first KfW loan to Tunisia. Continuous attention to institution building was necessary though the ongoing effort was bearing fruits as SNI was reorganized in January 1972. The Bank staff noted also significant improvement in appraisal quality in 1972 and recommended that the free-limit be doubled to $400,000. 2.11 The main covenants in Loan Agreement 881-TUN were: (a) an aggregate free-limit raised from $2.5 million to $4 million, representing a small change in proportion to the whole loan; (b) the subloans to hotels could be extended up to 20 years against 15 years for industrial projects to alleviate the debt servicing for hotel sponsors because current repayment schedules were not realistic and caused an excessive number of arrears. III. EVOLUTION OF THE ECONOMIC ENVIRONMENT IN TUNISIA (1969-1977) 1/ 3.01 During the 1960's, the economic policies in Tunisia centered on an inward-looking investment strategy based on state predominance in all important economic activities,and economic management relied on a complex system of government regulations. Tunisian industrial policies during this period favored the development of large state-owned enterprises, primarily for capital intensive projects in import substitution products such as steel, refined oil, vehicles, tools and appliances. In the 1960's, there was little foreign investment, the Government's involvement was dominant and the private sector accounted for less than one sixth of total investment in manufacturing and mining during that period. As the Appraisal Report for SNI III pointed out, the results of these policies were disappointing. 3.02 The appraisal of SNI-III coincided with a basic change in the economic policy in Tunisia. In 1969, new liberal policy measures were announced, which during the 1970's resulted in a more open export-oriented economy in which private initiative could play an increasing role. A new Investment Code, consolidating in one law various advantages already granted private investors, was passed in May 1969. There was a growing appreciation of the relative efficiency of the private sector, and efforts were made to reassure private investors that earlier nationalisation policies had been abandoned, and particularly to attract foreign private investors. As part of their liberalization 1/ "Tunisia: Economic Position and Prospects", Report No. 2201-TUN, November 6, 1978. - 8 - program, the Government also simplified the procedure for investment approval; import restrictions were gradually relaxed to introduce more competition in the domestic market; and the role of banks in allocating investment funds was strengthened. 3.03 Overall, the new policy orientation benefited the country; Tunisia's economy performed remarkably well during the 1970's. In the years 1970-1977, real GDP grew at an average annual rate of 8.6%, nearly twice as fast as during the 1960's. Per capita GNP reached $840 in 1977, 74% above its 1969 level in real terms. According to the 1979 World Bank Atlas, Tunisia was one among the dozen countries in the world whose GNP per capita during 1970-1977 increased by 6% or more each year. Agriculture generated some 20% of GDP and accounted for some 40% of total employment during 1970-1977 partly because favorable weather conditions resulted in good agricultural crops. The change in world market prices during 1973/74 for phosphates and petroleum brought windfall profits. In the industrial sector, economic performance was characterized by a relatively fast growth of both production and investment. However, it took about two to three years before private investors recognized that the time of nationalization was over; as a result, the industrial investment boom started only in 1972-73. Tourism, which had been sheltered against nationa- lization in 1960's, remained the favorite sector of private investors until 1973, when the tourism trend reversed and arrivals started to decline. 3.04 The Government traditionally kept inflation under control in Tunisia mainly by maintaining a balanced budget and also by applying strict price controls and granting Government subsidies on basic commodities. The cost of living index (COL) showed an average increase of 5.4% per annum between 1970 and 1977. Investment costs rose faster than the COL; but that did not deter investments which rose by 12.9% per annum in real terms over the same period. 3.05 The most important socio-political problem facing the Tunisian economy remained widespread unemployment and underemployment. Some 220,000 people, about 15% of the non-agricultural labor force, were officially unemployed in 1977. Taking into account the considerable underemployment in agriculture,real unemployment probably reached about 25% (yet this compared favorably with 1969 when this percentage was 35%). During 1970-1977, Tunisia made important, though insufficient, efforts to create jobs by giving generous tax incentives to investors in industry and tourism, providing investment financing on favorable terms, and creating jobs in public enterprises. -9- IV. INSTITUTIONAL PERFORMANCE A. Background 4.01 It was against this economic background that SNI emerged as the development bank of Tunisia (see Annex 1 for a detailled analysis of SNI's operations). It was not, however, without some institutional difficulties. When the Bank appraised the third loan in March 1969, it saw few problems with SNI except perhaps the need to improve middle management staffing and to strengthen further its appraisal of subprojects. SNI was put on the list of projects with long intervals between supervision missions. As a result, no basic supervision of SNI took place for the next twenty months. This judgment was later on seen retrospectively as a serious mistake, and lessons were taken into consideration in setting up supervision standards for DFCs (frequency of missions, topics, reporting requirements, etc) the Bank applied in the 1970's. 4.02 When the Bank started to prepare for SNI-IV in February 1971, SNI's middle and top management were in disarray. The President had just been changed; experienced middle management had left; staff morale was very low and the quality of operations had declined. A few months thereafter the audit report was strongly qualified; it described a deterioration of SNI's portfolio and pointed to inadequate provisions and supervision of investments. 4.03 Given Tunisia's resource gap the Bank accepted, nevertheless, to make the loan. However this was done only after SNI reached a number of agreements with the Bank to improve its internal structure and operations. SNI also agreed to strengthen its financial situation and to increase its provisions for bad loans. Basic measures were called for,given the important role SNI should have and indeed had in the Tunisian economy as the leading source of term financing for the private manufacturing sector and for tourism. By the time of appraisal of the fifth loan (July 1972), the key issues were under control though additional measures had to be taken to further strengthen middle management and BDET's financial situation. 4.04 Given this background, this PCR will focus on three aspects which are of interest in terms of the institutional performance of SNI: SNI's management, organization, staffing; quality of appraisal; audit matters. An analysis of the three Bank loans' utilization is found in Annex 1. Quantitative data are grouped in Annex 2. B. SNI's Management, Staff, and Organization 4.05 In the early 1970's, SNI's performance suffered from staff turnover and management changes. Actually SNI had experienced a fairly high rate of turnover even before 1970 because its credit department was widely used as a training school for future middle managers before taking up jobs in industry. But the management, organization and staff strength of SNI began to weaken in mid-1969 with the departure of several senior officers, among which the general manager and the head of the credit department. These losses among experienced staff affected adversely the quality of project appraisal and follow-up work at a time when SNI's operations and portfolio were growing rapidly. Changes in the top management compounded somewhat the difficulties. - 10 - The Government became increasingly aware of the need to strengthen SNI at a time when Tunisia's new economic policies needed strong institutional support. Thus in March 1971, Mr. Habib Bourguiba Jr. was appointed as SNI's President. He was assisted by SNI's previous President, Mr. Belkhoja, who had agreed to serve as managing director during a limited period of time and Mr. Mokhtar Fakhfakh, deputy general manager. 4.06 In addition to contributing to fill SNIs resource gap, the Bank chose to focus its efforts with SNI on institution-building objectives. The team Bourguiba Jr.-Belkhoja-Fakhfakh proved quite effective 1; Mr. Bourguiba Jr. showed quick understanding of SNI's problems and set in motion steps to_remedy the organization's shortcomings. An International Executive Service Corps consultant was hired to advise SNI on a reorganization designed to improve its project appraisal work as well as to develop its promotional capability. The French Caisse Centrale de Coopgration Econonique, a shareholder, also provided help to strengthen SNI's accounting and portfolio supervision. The Bank kept closely involved with the reorganization effort which became effective in January 1972. The former credit department was expanded into an appraisal department divided into two divisions: one for financial and economic appraisal and the other for technical evaluation. A separate department was established to process projects promoted by SNI. At the Bank's urging during negotiations for SNI-IV, a division for project supervision was also created. 4.07 Overall the reorganization was a success. Two factors played a major role. First, the decision to appoint Mr. Crifo, a French technical advisor who had been with SNI since 1964, as head of the new appraisal department,proved to be wise. Under his guidance, appraisal quality improved rapidly and a systematic search for promising young Tunisian professionals was undertaken with very positive results as some of them are now heading SNI's departments. Second, the timing was right because the strengthening of SNI's appraisal capabilities coincided with a long lasting surge of investments in Tunisia which translated into a steep increase of operations for SNI (see Annex 1). 4.08 Some aspects of the reorganization took, however, disappointingly longer than expected to be implemented. Two cases in point are the promotion department and the follow-up division. The promotion department was not really active for a long period because of lack of qualified personnel, leadership, and commitment on the part of SNI's management. Only in 1978 did it start to contribute significantly to SNI's loan approvals. 1/ When Mr. Belkhodja decided to leave SNI in 1974, the Bank voiced its preoccupation that this would leave a void and could unduly weaken SNI's management. This caused some strain in the relations between the Bank and SNI until mid-1976, when the Bank recognized Mr. Fakhfakh had established himself as a very competent manager. - 11 - 4.09 To a large extent at the insistence of the Bank, the follow-up activity was undertaken reluctantly by SNI and was hampered until recently by a lack of recognition by management and a narrow perception of the scope of project supervision. Some of these shortcomings still exist. The follow-up division produced eighty supervision reports in 1978, but it was mainly to comply with requests from lenders such as KfW and the Arab Fund for Economic and Social Development. Moreover, the reports' recommendations were rarely followed by any action or fed back to the entrepreneurs. Lastly, the follow-up division's staff, dealing only with supervision of old projects and not with appraisal of new ones, suffered from their isolation from the main stream of SNI's activities. Some corrective measures were recently taken, however, with weekly reviews of all follow-up reports by the Management Committee to decide on action to be taken. Serious consideration is also now given to a closer integration of supervision and appraisal work with the supervision staff being given responsibility to appraise repeater projects, in exchange for the participation of the technical staff in the appraisal department in some supervision activities. C. Quality of Appraisal 4.10 The quality of SNIts appraisal deteriorated in 1970-71, because of the loss of experienced staff noted above. Technical and marketing evaluations in particular were weak, the economic merits of projects were often not adequately discussed and the financial rate of return calculation was sometimes missing. As a result, the Bank had frequently to raise questions when reviewing projects. In both SNI-III and SNI-IV the Bank left the free limit unchanged (para 2.08) in order to be able to review as many subprojects as possible. 4.11 As mentioned earlier, the 1972 reorganization aimed mainly at improving appraisal work. Additional engineers and financial analysts were recruited not only to cope with the rapid growth in lending, but also to enable SNI to assess more thoroughly the projects' technical merits and to subject them systematically to financial rate of return tests. The effects of this reorganization were very rapid. At the time of SNI-V, in 1973, the Bank noted that SNI's project .appraisals had improved considerably, yet continued guidance was necessary because weak assessments occurred sometimes, especially as regards market studies and economic justification of subprojects. The review of 49 subprojects above the free limit between 1970-1976, follow-up missions and negotiations for further loans offered opportunities to the Bank to comment on progress in appraisal quality. In agreement with the Bank, SNI calculated the internal financial rate of return on"R subprojects only during the period under review, though rarely on an incremental basis for expansion projects. In late 1973 the Bank introduced SNI's staff to the economic rate of return calculation methodology, which SNI began applying to subprojects under the sixth Bank loan. SNI has since become quite proficient in the economic analysis of projects, although there is still some room for improvement (sensitivity tests, incremental analyses, shadow pricing of non- tradeables). - 12 - D. Audit Reporting Requirement 4.12 In 1971, Peat, Marwick, Mitchell and Co., SNI's auditors issued a disclaimer on SNI's 1970 accounts. They refused to express an opinion on SNI's accounts because they were unable to get information on fifteen SNI's important clients in time to perform the audit before the May 31-deadline stipulated in the successive loan agreements with the Bank. The previous year, they had qualified SNI's 1969 accounts, questioning the adequacy of provisions for doubtful loans, and during 1970, arrears increased considerably (para 2.07)For these reasons, the auditors asked to undertake a detailed review of selected clients in arrears. Only after they had done it, in September 1971, the auditors expressed a positive opinion on the 1970 accounts, provided D16,000 of old accounts receivable were written off which SNI did out of the 1971 income. 4.13 In 1973, Peat, Marwick, Mitchell and Co. associated with the Tunisian Cabinet Finor issued again a disclaimer on the 1972 accounts. The reasons were SNI's failure to make D300,000 of provisions on two doubtful loans and the over-valuation of its equity portfolio. The same auditors issued another disclaimer on the 1973 accounts for the same reasons. Another auditor, Arthur Anderson and Co., produced also a disclaimer on the 1974 accounts for similar reasons, estimating necessary provisions to more than double those set aside by SNI. 4.14 There were three main reasons for these audit problems. First, especially in the case of the first disclaimer, it was difficult to obtain the annual statements of SNI's clients in time for SNI's audit because, by law, Tunisian companies were not required to issue such statements before June 30. This conflicted with the May 31-deadline imposed on SNI's audit report to allow Board members to obtain an outside opinion on SNI's accounts before deciding on how to allocate profits. During negotiations of SNI-IV, the Bank reached an understanding with SNI to correct progressively the situation by requiring its borrowers to provide SNI with provisional financial statements in the loan contracts within three months after the end of their fiscal year, generally December 31. 4.15 Second, according to Tunisian law, annual allocations to provisions for doubtful loans are not deductible from taxable income, our only actual losses are. SNI had no financial incentive to take large provisions out of its income after tax. Nevertheless, at the insistence of both the Bank and the auditors, SNI gradually set aside sufficient provisions to cover the net risk on its portfolio. In more recent years, however, the auditors have identified three large borrowers (one hotel and two industrial firms) which were in difficulty and for which SNI had not made any specific provisions leading the auditors to single out SNI's loans to those borrowers as potential losses, in their reports on the 1976-78 accounts. Since then, SNI has taken measures to minimize its risk, which the auditors found satisfactory. - 13 - 4.16 Another cause of controversy between SNI and the auditors was the treatment of the blocked reserves of deferred taxes (R6serves pour Investissements Exon6r6es). Income from equity investments in sectors specified under Law 62-75 are tax-exempt if the securities are held for five years; the corresponding income must by law be accumulated in non- distributable reserves for the same length of time. SNI considered these non-distributable reserves as equivalent to a provision of a general nature and thus claimed that, added to the formal provisions, they covered adequately the risk on portfolio. Peat, Marwick, Mitchell and Co. eventually accepted SNI's argument, though reluctantly,because according to generally accepted accounting principles these reserves could be assimilated to net worth, but not so easily to provisions. However, Arthur Anderson and Co. considered these reserves neither as net worth because they could not be distributed as profits before five years, nor as provisions because they were not available immediately to cover actual losses. Time showed, however, that SNI maintained its portfolio of tax-exempt securities well beyond the five-year limit, and the auditors eventually accepted them as covering part of the risk on SNI's portfolio, particularly that of equity investments, on the grounds that, should a loss be incurred on such investments, the tax liability would also be accordingly reduced, and the corresponding reserve would become usable to write off the loss against it. 4.17 Third, auditing was somewhat negatively perceived by SNI's management for some time. SNI has always been reluctant to distribute widely the audit report. For many years, the Board members were not given the audit report on the ground that it contained confidential information on SNI's clients. Since the mid-1970's, the directors have received the audit report, but on too short notice to enable them to raise questions on SNI's financial situation before the annual shareholders meeting. In 1979, for example, the 1978 audit report was distributed only the day before the general assembly meeting. The Bank has been trying to correct this with little success. The Bank itself received the 1978 audit report only in September 1979, more than three months after it was ready. Since 1976, however, the auditors attend the general assembly meeting in order to answer questions from the shareholders regarding the audit report. This positive initiative would become even more useful if the shareholders receive the audit report with sufficient advance notice to prepare their questions. 4.18 One positive aspect of the audit experience was the cooperation which developed between the foreign and the local auditors over the years. After the 1971 disclaimer, SNI and the Tunisian authorities pressed for a Tunisian accounting firm to be agreed as SNI's auditors. In 1973, at SNI's request, the Bank sent a mission to evaluate the suitability of the proposed Tunisian audit firms. Eventually the Cabinet Finor was chosen, but was to operate in partnership with a foreign audit firm in order to gain the experience it lacked in auditing matters. Peat, Marwick, Mitchell and Co. and Cabinet Finor joined forces to audit the 1972 and 1973 accounts, and all audits after 1975 (the 1974 audit was undertaken by Arthur Anderson also in cooperation with Cabinet Finor). - 14 - SNI and Tunisia in general have greatly benefitted from this association which could serve as a model of technical assistance and cooperation in the auditing field for countries in need of strengthening the local auditing profession. V. SUMMARY AND CONCLUSION 5.01 The three Bank loans under review in this report achieved their dual objective of helping SNI fill resource gaps and of strengthening SNI's organization and ways of operating especially in project appraisal and follow-up. The Bank's financial contribution to SNI was significant, reaching 55% of SNI's disbursements in 1972, and 16.4% of total disbursements over the period 1970-1977. The timing of the three Bank loans was appropriate. First, these loans enabled SNI to fill its resource gap. Second, adequate time was available between loans for SNI to undertake necessary reorganizations and administrative improvements and for the Bank to assess them before processing the next loan. Considerable progress in terms of institution building was achieved: a major reorganization was planned in 1971 and carried out in early 1972; the credit department was expanded; a project promotion department and a follow-up division were created. Under Bank guidance, project appraisal was significantly improved. 5.02 The timing of the loans under review was especially propitious, in that it coincided with the shift of Tunisia's economic policies away from centrally planned socialism toward an open market economy. SNI being the sole institution in Tunisia offering long-term foreign exchange financing for private sector investments, played a crucial role in the process of converting the national economy to the new private sector orientation. The Bank's loans, audits, institution building efforts with SNI, contributed effectively to the financing of the private investment boom which began in the early 1970's in Tunisia, and to the re-definition of SNI's role, from that of a supplier of equity funds for public sector investments, to that of financier and advisor of private investors in industry and tourism. 5.03 In the first half of the 1970's, there was disagreement between SNI and the auditors concerning the adequacy of provisions to cover the risk of loss on the portfolio. The thorough and constructive audits since 1971 led SNI to gradually bring its provisions for doubtful loans and equity investments to satisfactory levels. In recent years, only limited qualifications and acceptable SNI's accounts have been issued. The audit report, however, is still not distributed to SNI's board members and to the Bank in a timely manner. -15 - ANNEX 1 Pg. 1 ANALYSIS OF THE BANK LOANS AND OF SNI'S OPERATIONAL AND FINANCIAL PERFORMANCE DURING THE PERIOD 1970-1977 I. UTILIZATION OF THE LOANS A. Third Bank Loan 1. The Appraisal Report for SNI-III assumed that the loan would be committed within two years and fully disbursed within four and a half years. Actually, it was committed in slightly over two and a half years and disbursed in about five years. However, as shown in Table 1, Annex 2, 50% of the loan was disbursed in the first two years, and 89% in the first three years, while 7% was eventually cancelled. During these three years (1970-72), the loan proceeds accounted for 45% of total SNI foreign exchange disbursements. 2. The Bank initially approved 14 projects above the free-limit ("A" projects) of $200,000 and authorized 51 projects below the free-limit ("B" projects), but cancellations reduced the latter figure to 45. Based on these initial approvals, the size of "A" subloans averaged $527,430 and ranged from $202,000 to $1,525,000; the size of "B" subloans averaged $77,420 and ranged from $7,000 to $191,000. Based on a sample of 52 projects, the greatest share of the loan proceeds was for hotels (29%) followed by services (25%), textiles (10%), electrical and mechanical industries (9%) and leather (7%). The majority of the projects was located in Tunis (64% of the loan amount). B. Fourth Bank Loan 3. The Staff Appraisal Report for SNI-IV assumed that the loan would be committed within one year and fully disbursed within four years. Actually, it was committed in two and a half years and disbursed in over five and a half years. 66% of the loan proceeds was disbursed in the first three years, but it took three more years-to disburse another 25%, and 9% was eventually cancelled. During the period 1972-1974, the loan -proceeds accounted for 20% of SNI's total disbursements in foreign exchange, down sharply compared to the previous loan (45%). 4. The Bank originally approved 12 "A' projects and 37 "B" projects, but cancellations reduced these figures to respectively 10 and 35. Based on original approvals, the size of "A" subloans averaged $719,430 and ranged from $250,000 to $2,200,000; the size of "B" subloans averaged $67,520 and ranged from $6,300 to $279,000. Compared to the previous Bank loan, the average size of "A" subprojects under SNI-IV increased (+ 36%), while "B" subprojects became fewer and smaller size on average (- 13%). This was because of the large number of big hotels financed (66% of loan amount) with projects in services (13%) and textiles (8%) far behind. The majority of projects (64% innumber) was still located in Tunis. - 16 - ANNEX 1 page2 C. Fifth Bank Loan 5. The Staff Appraisal Report for SNI-V forecasted that the loan would be committed within two years and fully disbursed within five years. The loan was declared effective three months after signature and it was committed in two and a half years and disbursed in slightly over six years. Disbursements were faster than projected in the first two years (70% against 46% projected), but slowed down afterwards. 6% of the loan amount was eventually cancelled. In the first three years (1973-1975), the loan proceeds accounted for 23% of SNI's total disbursements in foreign currencies. 6. The Bank initially approved 16 "A" projects and 34 "B" projects, but cancellations reduced those figures to respectively 15 and 30. Based on the amounts initially approved the size of "A" subloans averaged $741,465 and ranged from $259,000 to $1,620,000 1/; the size of "B" subloans averaged $138,185 and ranged from $12,500 to $375,000. The doubling of the size of "B" projects (+ 105%) was a direct consequence of the doubling of the free-limit (para 2.10) The sector distribution of projects changed compared to SNI-IV's. Although the loan maturity was extended to twenty years to accomodate longer terms for hotels (para 2.11)only one hotel project was approved under SNI-V and it represented only 4% of the loan proceeds. The mechanical-electrical industries got the largest share (28%), followed by textile (18%), food industries (16%), and wood processing (13%). The Tunis location was emphasized more (77%) than in the previous two loans because of the shift away from tourism and toward manufacturing industries. II. CHARACTERISTICS OF PROJECTS FINANCED UNDER THE THREE BANK LOANS 7. 149 subloans for 146 companies were financed out of the three Bank loans for a total disbursement of $31.8 million. These projects are broken down as shown in Table 1. Table 1. PROJECTS FINANCED UNDER SNI-III, IV, V Projects financed under: Number of Amount Disbursed Projects in $ 000 in % of total SNI-III "A" -above free limit: 14 5,259 57 ($200,000) "B" -below free-limit: 45 4,038 43 59 9,297 100 SNI-IV "A" -above free-limit: 10 6,859 73 ($200,000) "B" -below free-limit: 35 2,519 27 45 9,378 100 SNI-V "A" -above free-limit: 15 9,693 73 "B" -below free-limit: 30 3,667 27 45 13,360 100 1/ Excluding two small loans to borrowers under former Bank loans, whose outstanding borrowings were above the free-limit. - 17 - ANNEX 1 page 3 The first column of Table 1 showsthat the successive Bank loans financed a decreasing share of Tunisian investments, both in terms of amount and in terms of number of projects. This was partly because Bank lending did not keep up with the rapidly increasing investments in Tunisia (which had been under- estimated at appraisal), and partly because of the growth of the average size of individual subprojects. A. Contribution to Employment Creation 8. The three Bank loans contributed significantly to the creation of new jobs in Tunisia in the 1970's as shown in the table below. Table 2. CONTRIBUTION OF BANK LOANS TO JOB CREATION IN TUNISIA Jobs created 1/ Cost/job created VJ(Dinars) "A" projects "B" projects TOTAL "A" projects "B" projects SNI-III 1,500 1,634 3,134 9,052 (10)2/ 10,372 (25) SNI-IV 1,514 884 2,398 24,414 (10) 6,490 (21) SNI-V 1,824 650 2,474 14,332 (11) 12,967 (21) 1/ Based on data in project appraisal reports 2/ Number of projects for which average was computed. 9. The 149 projects financed by the Bank created about 8,000 jobs during the 1970-1976 period when these projects were approved. SNI-IV generated less jobs than SNI-III because,although they were of same amount, they were two years apart; therefore, SNI-IV was actually smaller than SNI-III in real terms. SNI-V was 40% greater than the two previous Bank loans, but this increase merely met rising investments costs and, thus, the number of jobs created was about the same as under SNI-IV. 10. The average capital cost per job created under the three Bank loans was systematically higher than the average for all the projects approved by SNI during the period 1970-76; the latter was around D8,000 between 1972-74, declined to D6,000 in 1975, and rose to over D12,500 in 1976. The high proportion of hotels financed under SNI-IV explains why the average cost per job for "A" projects rose to D24,414 (about $54,900). In general, costs per job for "B" projects were lower than for "A" projects. This confirms that (a) hotel investments, which are generally large and above the free limit, are more capital intensive than industrial investments, and (b) that, among projects in the manufacturing sector, the.larger projects tend to be more capital intensive. B. Actual Performance of Bank-financed Projects 11. Table 2, Annex 2, gives selected data concerning the 39 "A" projects financed under the three Bank loans. -18 - ANNEX 1 page 4 12. For 10 projects for which expost data are available, the actual total project cost on completion was D13.1 million compared to appraisal estimates of D10.9 million which gives a 21% cost overrun. This percentage was higher (+ 22%) for projects with investment costs greater than Dl million, but lower (+14%) for projects smaller than Dl million. 13. For 66 projects for which actual employment data are available, the number of jobs actually created was 4,437, that is 270 more than appraisal estimates; but most of these jobs were concentrated on 7 projects, while 2 other projects created 300 fewer jobs than estimated. 14. Actual sales in the third year of operation were on average 153% of the appraisal estimates on a sample of "A" projects financed under SNI-III; it was only 105% for "A" projects financed under SNI-IV, and as high as 269% for "A" projects financed under SNI-V. Similar figures for actual profits before tax compared to projection were 105%, 25% and 190% 1/ for samples of "A" projects financed under respectively SNI-III, IV, and V. These figures were largely influenced by the relative proportion of hotel projects (3 out of 14 under SNI-III; 6 out of 10 under SNI-IV; and 1 out of 15 under SNI-V) whose profitability had been overestimated at appraisal, thus explaining the low 25% actual/estimated ratio for SNI-IV. 15. "B" projects showed different patterns. First, actual sales three years after start-up were consistently higher than appraisal projections, and more so than for "A" projects. Actual profits were also greater than appraisal projections, except for "B" projects financed under SNI-IV where there was also a relatively high proportion of hotels financed. These figures are summarized in Table 3. Table 3. COMPARISON OF ACTUAL PERFORMANCE WITH APPRAISAL PROJECTIONS (in %) Actual Sales in 3rd Year/ Actual Profits before tax Projections in 3rd Year/Projections "A" projects "B" projects "A" projects "B" projects SNI-III 153 (12)1/ 180 (27) 105 (8)2/ 163 (19) SNI-IV 105 ( 8) 177 (18) 25 (6) 90 (11) SNI-V 269 (11) 174 (15) 190 (8) 156 (13) 1/ Number of projects for which average was computed. 2/ The difference with 1/ gives the number of projects which showed losses instead of profits. 16. As of September 30, 1979, 12 of the 146 companies financed under the three Bank loans were in arrears of more than three months. Their arrears amounted to only D518,787 in principal, but they represented 75% of all arrears of over three months in principal. They were caused mainly by bad management, frequent changes of owners, or business folding after a short period. According to BDET three of the twelve companies in arrears had good recovery prospects. In addition, loans to, and equity investments in fourteen other companies financed under the three Bank loans were provisioned against losses for a total of D724,500 at the end of 1978, of which D386,000 were for possible losses on loans. Finally, legal action had been engaged against 4 Bank-financed projects. 1/ Excluding the cases where losses were registered instead of profits. ANNEX 1 19 - page 5 III. OPERATIONAL AND FINANCIAL PERFORMANCES OF SNI A.Growth of Operations 17. SNI's actual loans and equity investments approved and disbursed from 1970 to 1977 are summarized in the table below. Table 4. ACTUAL OPERATIONS OF SNI (1970-1977) (D'OO0) 1970 1971 1972 1973 1974 1975 1976 1977 Dinar loans: Approvals 1,311 5,247 4,348 3,671 9,343 3,614 9,544 8,793 Disbursements 809 831 1,008 3,146 4,000 4,000 5,752 na Foreign exchange loans: Approvals 5,144 3,865 5,970 13,237 24,673 17,026 17,963 19,841 Disbursements 2,924 2,403 3,647 2,403 8,960 10,767 10,317 na Equity investments: Approvals 255 na 761 1,184 2,675 1,723 3,070 na Disbursements 124 209 221 1,069 1,185 945 1,866 1,858 Total Approvals 6,710 na 11,079 18,092 36,691 22,363 30,577 na Total Disbursements 3,857 3,443 4,876 6,618 14,145 15,712 17,935 na 18. Approvals of both loans and equity investments grew markedly in the 1970's. 1974, a boom year for investments, was also a boom year for SNI's overall approvals, but the upward trend started before. Approvals rose to record levels as early as 1971 for Dinar loans, 1972 for equity investments, and 1973 for foreign currency loans. However, as shown in Table 3, Annex 2, none of the projections prepared successively for SNI-III, IV, and V foresaw the great surge of activity that made SNI the development bank of Tunisia. Similarly, all the projections of commitments and disbursements proved to be overly conservative, especially since about 1973, by a factor of three or four. B. Financial Results 19. As a result of underestimated business projections, the major items of SNI's balance sheet and income statement, with few exceptions,grew larger than projected since about the same year. As shown in Table 4, Annex 2 dealing with balance sheets, this was true in particular of the loan and equity portfolios, of the long term borrowings made to finance them, and of the share capital increased to allow for more borrowings. As shown in Table 5, Annex 2, dealing with income statements, this was also true of interest received on loans and of interest paid on borrowings. However, dividends from equity investments were constantly lower than projected over the whole period 1970-1977 (due to a decrease from 9% to 8%, in SNI's lending rate in 1973), while administrative expenses were invariably higher. - 20 - ANNEX 1 page 6 20. The main financial ratios reflect some of the patterns described above. Interest income as percentage of average loan portfolio was always lower than projected; 8.3% in average over 1970-1977 instead of 9.4%. However, gross income in percentage of average total assets was better forecasted with an average actual yield of 7.7% against an average projection of 8.4%. Financial expenses in percentage of the same assets were about 10% lower than projected: 4.1% versus 4.6% in average; but because of the relatively fast growth in personnel, administrative expenses were about 20% higher than expected: 1.7% of total assets versus 1.4% projected. As a result, even though SNI's gross return on equity doubled from 6.5% in 1970 to 13.7% in 1977, it was still substantially lower than projected, except in 1976-1977. C. SNI'S Spread 21. This generally lower than expected profitability was not due in a significant way to lower a than expected spread on the Bank loans. Even before SNI-III was appraised, SNI increased its lending rate from 8% to 8.5% in 1968 to maintain a 2%-spread on Bank resources, and at time of appraisal (March 1969) a further increase to 9% was considered by SNI to keep up with the new Bank lending rate of 7%. On-lending rate of 9% on SNI foreign exchange loans became effective toward the end of 1970. SNI's spread on Bank funds was reduced, however, under SNI-IV because the Bank rate was raised to 7.25% (para2.05) while SNI's rate remained unchanged. While SNI-V was appraised (July 1972), the Government indicated that it was reviewing the interest structure in Tunisia. The Bank staff expected the changes to be made to have an important effect on SNI. Indeed, SNI's rate was lowered from 9% to 8%, effective in April 1973. The following month, SNI-V became effective with an interest rate of 7.25% plus the usual.75% commitment fee, thus reducing the spread on Bank funds to less than three quarter of a point. The Government justified the lower rate to borrowers to stimulate economic development and to harmonize term lending rates between SNI and the commercial banks. The Bank agreed to the change after the Government gave assurance that SNI's profitability would not be harmed and that the interest rate would be raised as soon as economic conditions would permit. Later, SNI concluded a stand-by arrangement with the Government guaranteeing a minimum - operating margin of 2.5% on borrowed resources. At the Bank's urging, an understanding was reached during negotiations of the 6th Bank loan (December 1975) to return to the 9%-rate for all new borrowers as of January 1, 1976. However, because of the steadily increasing cost of SNI's borrowed resources, which rose from 5.4% in 1974 to 6.1% in 1976, the arrangement with the Government to assure a minimum spread of 2.5% on SNI's operations had to be maintained. Its terms, however, were changed in 1976, following Bank prodding, to the effect that the minimum spread would be determined every year on an ex-ante basis, in view of BDET's projected borrowings and loans, rather than at year-end, on the basis of actual results. This change in the procedure was intended to safeguard BDET's incentive to maximize its profitability. The mimimum spread was later raised to three percentage points, to take into account increased administrative costs incurred by BDET to promote SSI and other priority sectors. - 21 - -1- TABLE 1 CUMULATIVE DISBURSEMENT END OF CALENDAR YEARS (million) 1970 1971 1972 1973 1974 1975 1976 1977 1978 SNI-III 648-TUN actual 1.9 4.9 8.0 9.0 9.3 estimated na na na na na SNI-IV 798-TUN actual - - 2.5 4.4 6.6 8.7 9.0 9.1 - estimated - .6 4.8 8.8 9.9 10.9 SNI-V 881-TUN actual - - - 1.2 9.8 11.8 13.2 13.4 - estimated - - - .8 6.5 12.8 13.9 14.0 -22- TABLE 2 Nature Amount Loan Creation(C, Implementa- Invest- Turnover in Profit befo4a Job of disbursed Maturity or tion Date: ment Cost: Tear 3: Year 3: created Project/Name Buainese ( (years) Extension Forecasted Forecasted Forecasted Forecasted Foreca (grace) (E) (Actual) (D'000) (Actual (Actual (Actual) (D'000)%) (D'OQO)Z) Loan 648-TULN-SNI-II 1. TANNERIE HODERNE Leather 477,947 10(1) E 1971(na) 370 1,564(113 52(253) 110(na) 2. STIA Autom- 144,064 12(2) E 1971(72) 250 3,399(306 324(123) na(na) bile 3. MATEX Textile 467,802 10(2) E 1971(72) 529 649(113 55( 46) 108(na) 4. SNI* Transportl,208,691 10(4) E 1971(72) 5,060 6,660(120 588( 37) 450(317) 5. TRANSTOURS* Tranaport 762,000 5(0) E 1971(71) 900 1,380(1171 296(loss) 36( 36) 6. SIOC Rubber 340,300 10(1) C 1971(74) 312 190(360) 41(109) 23( 94) 7. MEDIA aotel 501.962 12(2) C 1972(na) 750 329( 68) 29(loas) 130(na) 8. STIA II kutomo- 196,589 10(1) E 1971(72) 124 6,164(160) 607( 66) na(na) bile 9. ZARZIS HOTEL fntel 375,374 12(2) C 1971(72) 1,000 500( 58) 33(loss) 220(na) 10. HOBUGA Iatel 860.714 11(1) C 1973(73) 2,200 1,442(na) 292(na) 115(na) 11. STAT r-rental 209,500 3(0) E 1971(72) 490 512(147) 78(loss) na(na) 12. STUFIT textile 161.000 11(1) E 1972(73) 530 1,283(109) 175( 49) 124(354) 13. Ste PLEASURE DON Tourihm 207.130 10(2) C 1973(na) 750 438(na) 59(na) na(na) 14. COTUSAL alt 207,130 10(2) E 1971(71) 444 1,340(150) 92(156) na(na) Loan 798-TUN-SNI-IV 1. Hotel PHENICIA Hotel 764,898 15(3) C 1974(na) 2,500 1,065(133) 112( 45) 360(370) 2. SyTAPEX-COUVERTEI Textile 522,000 7(1) E 1973(77) 440 300(206) 51(Loss) 59(61) 3. STAT Car-rents 423,000 5(1) E 1973(na) 680 890( 97) 237( 27) 17(19) 4. TUNISIAN CONTINENTAL Hotel 861,534 12(2) C 1973(na) 1,575 703(112) I1l( 23) 144(na) 5. Ste HOTEL&TOURIS (M'Henll 1,172,539 14(2) E 1974(na) 2,520 1.642(na) 200(a) 584(na) IHotel 6. STe LE PARADIS Hoatl 258,035 11(2) E 1973(na) 700 384(114) 58( 32) 160(160 7. Ste de PATES ALIKENTAIRES 279,397 10(1) E 1974(na) 214 509( 30) 31( 4) 8(8) food S. Sce TOURISTIQUE ES ILES Hotel 361,692 13(2) C 1974(na) 630 333( 49) 38(lose) 52(ne) 9. TUNISIAN-TRAVEL- ERVICE rransport 294,300 5(0) E '1972 187 900(na) 37(na) 9(na) 10. STE RTH dotel Z200,000 14(2) C 1974(na) 4.500 2,538( 98) 341(20) 700(na Loan 881-TUN-SNI-V 1. NCR TUNISIE omputer easing 473,740 5(0) E 1973(na) 300 625(147) 74( 21) 4(na) 2. Ste TUNIS BOISSO S GAZEUSES severages 359,442 7(1) E 1973(na) 680 1,610(214) 356( 48) 20(20) 3. STe TUNIS DE PROq .ALIM. Iod 1CT98,532 10(1) C 1975(na) 620 2,076(317) 28(749) 71(62) 5. RESOURCES-TUNISI; - Chemicals 1395,920 10(1) C 1975(na) 1,250 3.440(na) 85(na) 30(30) 6. SAVONNERIE AFRIC INE Foo 221,847 10(1) E 1973(na) 194 1,100(135) 89(loss) 0(na) 7. HERITIERS BOUZ 2 Mechanic& 18,633 8(1) E 1974(ne) 80 239(na) 55(na) 30(30) 8. Ste GEN4ERALE IINDT, Mechanica 70&,630 10(1) E 1975(na) 700 1,345(234) 260(204) 106(na) 9. STE, LA BROSSel Brush 5,628 7(1) E 1974(na) 37 130(394) 22(80) 3(3) 10. SACEM Electrica 267,941 8(1) E 1974(na) 222 1,238(101) 95(180) 75(75) S11. CRANUPHOS Phosphate 975,058 12(2) E 1975(na) 1,140 2,790(50) 125(Loss) 530(530) 12. STE SINUTEX-' Textile 30g.373 10(1) C 1974(na) 306 350(na) 62(na) 300(135) 13. SOCER Ceramic 412,500 8(1) C 1974(a) 540 342(188) Loss(Profit 80(80) 14. BOGURA Iotel 495,000 12(1) C 1974(na) 2,350 1,442(na) 292(0a) 0(na) 15. FILTISS Textiles 1,365,910 10(2) K 1974(na) 550 1,200(273) 278(125) na(na) 16. STIB II Wood 86,910 10(1) C 1974(na) 910 888(105) 207(112) 110(110) C Public enterprisea 1/A-4 was cancelled 2/Treated as "A"project because aggregate free-limit was exceeded. - 23 - SOCIETE NATIONALE D'INVESTISSEMENT/BANQUE-DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE Lending and Investment Operations (D'OOo) TABLE 3 1970 1971. 1972 1973 1974 1975 1976 1977 Approvals Dinar loans Loan 648:Proj 1,750 1,775 1,900 2,000 Loan 798:Proj - - 1,215 1,600 1,800 1,900 2,030 - Loan 881:Proj - - - 2,000 2,200 2,300 2,400 2,500 Actual 1,311 5,247 4,348 3,671 9,343 3,614 9,544 8,793 Foreign exchange Loan 648:Proj 3,950 - 4,500 4,850 - - - - Loans Loan 798:Proj - 5,303 5,300 5,500 5,700 5,950 Loan 881:Proj - - - 4,900 5,100 5,300 5,575 5,900 Actual 5,144 3,865 5,970 13,237 24,673 17,026 17,963 19,841 EquI v investments Loan 648:Proj. 450 450 500 550 - - - - Loan 798:Proj. - - 500 600 600 700 750 - Loan 881:proj. - - - 600 600 700 825 900 Actual 255 n.a 761 1,184 2,675 1,723 3,070 n.a Commitments Dinar loans Loan 648:Proj. n.a n.a n.a n.a - - - - Loan 798:Proj. - - 1,254 1,567 1,670 1,835 1,945 Loan 881:Proj. - - 2,540 2,080 2,240 2,340 2,440 Actual 1,103 3,615 5,000 3,919 6,760 n.a Foreign exchange Loan 648:Proj. n.a n.a n.a n.a - - - - Loans Loan 798:Proj. - - 6,222 5,202 5,370 5,620 5,787 - Loan 881:Proj. - - - 4,756 4,980 5,180 5,410 5,705 Actual - - 4,224 8,457 10,733 13,414 12,478 n.a. Equity investments Loan 648:Proj. n.a n.a n.a n.a - - - - Loan 798:Proj. - - 451 580 600 680 740 Loan 881:proj. - - - 580 600 680 802 885 Actual - - 241 943 1,494 913 2,361 1,916 Disbursements Dinar Loans Loan 648:Proj. n.a n.a n.a n.a - - - - Loan 798:Proj. - - 1,011 1,356 1,623 1,779 1,929 Loan 881:Proj. - - - 1,945 2,344 2,199 2,241 2,350 Actual 809 831 1,008 3,146 4,000 4,000 5,752 n.a Foreign exchange Loan 648:Proj. n.a n.a n.a n.a - - - - Loans Loan 798:Proj. - - 4,372 5,254 5,542 5,685 5,906 - Loan 881:Proj. - - - 4,837 4,895 5,007 5,208 5,449 Actual 2,924 2,403 3,647 2,403 8,960 10,767 10,317 n.a Equity investments Loan 648:Proj. n.a n.a n.a n.a - - - - Loan 798:Pro. - 345 52( 592 648 753 - Loan 881:Proj - - - 592 648 753 852 Actual 124 209 221 1,069 1,185 945 1,866 1,858 n.a: Not Available - 24 - SOCIETE NATIONALE D'INVESTISSEMENT/BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE Balance Sheet Statements: Projections compared to Actual (D'O0O) TABLE 4 1970 1971 1972 1973 1974 1975 1976 1977 ASSETS Cash & Deposits Loan 648: Proj. 265 313 392 453 - - - - Loan 798: Proj. - - 3,032 3,263 3,715 4,273 5,212 - Loan 881: Proj. - - - 3,612 4,092 4,936 3,754 4,385 Actual 2,445 3,408 3,215 1,994 5,571 5,553 5,955 4,796 Loans outstanding Loan 648: Proj. 12,317 16,896 21,367 25,858 - - - - Loan 798: Proj. - 11,839 15,302 19,579 23,864 27,754 31,227 - Loan 881: Proj. - - - 18,751 23,110 26,742 29,968 32,826 Actual 9,068 11,588 14,398 19,794 31,464 43,627 54,803 65,624 Equity investments Loan 648: Proj. 2,562 2,712 2,912 3,162 - - - - Loan 798: Proj. - - 2,150 2,428 2,720 3,068 3,471 - Loan 881: Proj. - - - 2,623 3,015 3,413 3,891 4,443 Actual 1,907 2,005 2,013 2,908 3,925 4,603 6,576 8,293 LIABILITIES Deposits Loan 648: Proj. na na na na - - - - Loan 798: Proj. - - 2,579 2,579 2,579 2,579 2,579 - Loan 881: Proj. - - - 2,792 2,931 3,078 3,232 3,393 Actual 1,983 2,531 2,700 4,150 5,969 4,389 3,313 4,103 Current liabilities Loan 648: Proj. 583 563 513 463 - - - - Loan 798: Proj. - - 2,792 2,792 2,792 2,792 2,792 - Loan 881: Proj. - - - 3,616 3,813 4,022 4,242 4,474 Actual 3,546 3,251 3,592 5,520 7,868 8.350 9,000 - Long term borrowings Loan 648: Proj. 11,490 5,966 Z0,391 24,827 - - - - Loan 798: Proj. - - L4,098 L8,681 23,371 27,760 32,085 - Loan 881: Proj. - - 17,712 20,537 25,055 27,206 30,758 Actual 8,546 .0,214 L2,522 L5,332 27,839 39,360 54,548 59,023 Total equity Loan 648: Proj. 3,745 3,957 4,223 4,528 - - - - Loan 798: Proj. - - 4,301 4,454 4,721 5,056 5,492 - Loan 881: Proj. - - 4,672 6,875 7,021 7,214 7,477 Actual 2,788 4,543 4,425 4,430 6,710 8,411 8,688 10,445 na: Not available. - 25 - SOCIETE NATIONALE D'INVESTISSEMENT/BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE Income Statements: Projections compared with Actual (D 000) TABLE 5 1970 1971 1972 1973 1974 1975 1976 1977 Interest on Loans & advance; Loan 648: Proj. 896 1,316 1,705 2,090 - - - - Loan 798: Proj. - - 1,250 1,649 2,039 2,411 2,746 - Loan 881: Proj. - - - 1,549 2,016 2,389 2,726 3,026 Actual 613 834 1,092 1,373 2,039 3,145 4,228 5,390 Dividend income Loan 648: Proj. 80 92 98 105 - - - - Loan 798: Proj. - - 70 75 97 109 123 - Loan 881: Proj. - - - 126 158 170 194 219 Actual 53 59 117 113 120 104 128 149 Total income Loan 648: Proj. 1,090 1,563 1,965 2,365 - - - - Loan 798: Proj. - - 1,491 1,916 2,349 2,754 3,137 - Loan 881: Proj. - - - 1,896 2,433 2,857 3,260 3,632 Actual 842 1,132 1,579 1,901 2,604 3,878 5,150 6,463 Interest on.borrowings Loan 648: Proj. 607 903 1,193 1,489 - - - - Loan 798: Proj. - - 771 1,015 1,228 1,530 1,728 - Loan 881: Proj. - - - 982 1,245 1,505 1,747 1,969 Actual 391 562 714 894 1,167 1,927 2,804 3,659 Administrative expenses (Ind. depreciation) Loan 648: Proj. 140 147 150 160 - - - - Loan 798: Proj. - - 271 297 323 362 390 - Loan 881: Proj. - - - 422 489 547 598 646 Actual 235 188 305 437 626 769 912 1,120 Income before tax (D'000) Loan-648: Proj. 321 484 589 674 - - - - Loan 798: Proj. - - 306 433 556 606 716 - Loan 881: Proj. - - - 351 554 657 761 859 Actual 180 343 459 402 571 810 1,033 1,329 Net income after tax (D'000) Loan 648: Proj. 240 362 441 505 - - - - Loan 798: Proj. - - 266 383 506 556 666 - Loan 881: Proj. - - - 272 395 458 530 600 Actual 134 258 401 402 571 725 946 1,059 - 26 - SOCIETE NATIONALE D'INVESTISSEMENT/BANQUE DE DEVELOPPEMENT ECONOMIQUE DE TUNISIE Financial Ratios: Projections compared with Actual TABLE 6 1970 1971 1972 1973 1974 1975 1976 1977 Income from loans/average Loan portfolio (%) Loan 648:Proj. 9.2 9.0 8.9 8.9 - - - Loan 798:Proj. - - 9.2 9.5 9.4 9.3 9.3 - Loan 881:Proj. - - - 9.4 9.6 9.6 9.6 9.6 Actual 8.1 8.1 8.4 8.0 8.0 8.4 8.6 9.0 Gross income/average Tot l assets Loan (1 Loan 648:Proj. 8.3 8.6 8.6 8.6 - - - - Loan 798:Proj. - - 7.7 8.1 8.3 8.3 8.3 - Loan 881:Proj. - - - 8.0 8.5 8.5 8.7 8.9 Actual 6.6 6.9 8.1 8.1 7.6 7.7 7.9 8.3 Administrative expenses/ Average total assets (%) Loan 648:Proj. 1.2 .9 .7 .7 - - - Loan 798:Proj. - - 1.4 1.3 1.1 1.1 1.0 - Loan 881:Proj. - - - 1.8 1.7 1.6 1.5 1.5 Actual 1.5 1.5 1.9 2.0 1.8 1.6 1.5 1.5 Financial expenses/average Total assets (%) Loan 648:Proj. 4.7 5.0 5.2 5.4 - - - - Loan 798:Proj. - - 4.0 4.3 4.3 4.6 4.5 - Loan 881:Proj. - - - 4.1 4.4 4.5 4.7 4.8 Actual 3.7 3.3 3.9 4.4 3.3 4.5 4.8 - Profit before tax/ Average equity (%) Loan 648:Proj. 10.1 12.6 14.4 15.4 - - - - Loan 798:Proj. - - 7.1 9.9 12.1 12.4 13.6 - Loan 881:Proj. - - - 7.6 9.6 9.5 10.7 11.7 Actual 6.5 9.4 10.5 9.1 8.5 9.9 11.1 13.7 Long-term debt/Year end Net worth Loan 648:Proj. 2.7:1 3.6 4.4 5.1 - - - - Loan 798:Proj. - - 3.2 4.0 4.7 5.2 5.6 - Loan 881:Proj. - - - 3.8 3.0 3.6 3.8 4.2 Actual 3.1 2.2 2.6 3.1 3.2 4.3 5.8 5.9