Turkey - Special economic report : policies for the financial sector

Report No. 4459-TU Turkey Special Economic Report Polides for the Financial Sector Septenber 21, 1983 Country Programs Department II Europe, Middle East and North Africa Region FOR OFFICIAL USE ONLY Document of the VoId Bank 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 wihout Ubrld Bank authorization  TURKEY CURRENCY EQUIVALENT (Annual and Quarterly Averages) Value of US$ 1959 TL 2.80 1970 TL 11.50 1971 TI 14.92 1972 TL 14.15 1973 TL 14.15 1974 TL 13.93 1975 TL 14.44 1976 TL 16.05 1977 TL 18.00 1978 TL 24.28 1979 TL 31.08 1980 First Quarter TL 61.60 Second Quarter TL 75.53 Third Quarter TL 80.10 Fourth Quarter TL 86.93 1981 First Quarter TI 94.59 Second Quarter TL102.84 Third Quarter TL118.89 Fourth Quarter TL128.56 1982 First Quarter TL142.25 Second Quarter TL153.27 Third Quarter TL171.86 Fourth Quarter TL182.85 1983 First Quarter TL194.15 September 15 TL242.65 FISCAL YEAR Republic of Turkey March 1 to February 28 - through 1981 March 1 to December 31, 1982 January 1 to December 31 - from 1983 SEEs and TCZB January 1 to December 31 ABBREVIATIONS EEC - European Economic Community IMF - International Monetary Fund LIBOR - London Inter-Bank Offer Rate M&LT - Medium and Long-Term OECD - Organization for Economic Cooperation & Development SEE - State Economic Enterprise SPO - State Planning Organization TCZB - Agricultural Bank of Turkey TSKB - Industrial Development Bank of Turkey EBK - Meat and Fish Authority TMO - Soil Products Office  -FOR OmCIAL USE ONLY This report was based on the findings of a mission that visited Turkey in November 1982, and incorporates information obtained up to February 1983. The report was written by Jose da Silva Lopes (mission. chief), Jacques Coudol (deputy mission chief), Cesare Calari, William Hayden, Seok Hyun Hong, Millard Long, Katrine Saito and Alain Coune and Leif Hansen of the International Monetary Fund. 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.   ABSTRACT The report critically assesses the capability of the financial sector in Turkey to fulfil the financial needs of Turkey's development strategy. It analyses rece.nt developments in the sector, in terms of the mobilization of financial savings, the allocation of credit, the high real cost of credit, t:he sources of corporate finance, and the structure of financial inst:itutions. It makes recommendations on ways of improving the efficiency of the financial sector in mobilising and allocating resources by means of an appropriate interest rate policy, banking reform, and a reduction in the costs of intermediation. The elaborate system of selective credits is reviewed, and recommendations made on how to reduce market fragmentation and improve the allocation of scarce credit funds. Recommendations are also made to streamline and expand the medium and long-term credit market. The reasons for the recent deterioration in the financial and debt position of Turkish industrial enterprises are examined, and means of financial rehabilita- tion explored. The need to develop the capital market is stressed, and suggestions made on specific measures to promote the bond and money markets, the equity market: and an efficient secondary market. Such developments would not only provide competition to the banking system and permit a reduction in intermediation margins, but also diversify the supply of financial instruments, thereby encouraging a better allocation of iesources.   TURKEY SPECIAL ECONOMIC REPORT POLICIES FOR THE FINANCIAL SECTOR Page No. Table of Contents COUNTRY DATA SUMMARY OF CONCLUSIONS AND RECOMMENDATIONS i-xuiii Chapter I: THE SIZE OF THE F'INANCIAL SYSTEM AND THE MOBILIZATION FINANCIAL SAVINGS 1 A. The Size of t:he Financial System 1 B. The Role of Interest Rates in the Mobilization of Financial Savings 2 C. Increasing the Size of the Financial System 6 Chapter II: THE GROWTH AND ALLOCATION OF CREDITS FROM THE BANKING SYSTEM 9 A. Credit Expansion and Monetary Policy 9 B. The Allocation of Credits 10 C. Financing of the Deficit of the Consolidated Budget of the Central Government 12 D. Financing of State Economic Enterprises 13 Chapter III: THE COST OF CREDIT 16 A. The High Level of the Real Costs of Credits 16 B. The Need to Reduce the Real Cost of Credits 18 C. Reduction of the Cost of Bank Deposits 18 D. Reducing the Burden Imposed by the Liquidity and Reserve Requirements 22 E. Contributions to the IRRF 25 F. Elimination of the Financial Transaction Tax 26 G. Cost of Bank intermediation 26 H. Lending Interest Rates and the Supply of Credit 27 Chapter IV: SELECTIVE CREDIT POLICIES 29 A. Controlling Credit Allocation 29 1. IntroducItion 29 2. Improvements in the System of Selective Credits 30 B. Agricultural Credit 31 1. The Agricultural Credit System 31 2. Interest Rate Policy in Agricultural Credits 34 3. Institutional Problems 35 4. Differentiation in the Types of Agricultural Credits 35 5. Crop Purchase Activities 36 6. Input Financing 36 7. Medium Long-term Credit 36 8. Small-farmer Credit 37  Page No. C. Export Credit 37 1. The Growth of Export Credits 37 2. The Export Credit Scheme 38 3. Improvement of the System of Export Credits 40 D. Housing Credit 41 1. The Availability of Housing Credits 41 2. The Housing Credit Scheme 42 3. Improvements of the System of Housing Credits 42 E. Medium and Long-Term Credits 43 1. Introduction 43 2. The State Investment Bank 44 3. TSKB and SYKB 45 4. Main Problems in Medium- and Long-term Financing 46 5. Mobilization of Resources in Domestic Currency for the Investment and Development Banks 46 6. Fixed or Floating Interest Rates in Medium- and Long-term Loans 48 7. Subsidization of the Interest Rates of Mediumr and Long-term Credits 50 8. The Coverage of Exchange Risks in Medium- and Long-term Loans 51 9. The Liquidity Problems Created by High Nominal Interest Rates in Medium and Long-term Loans 52 Chapter V: CORPORATE FINANCE 53 A. The Financial SituationL of Industrial Firms 53 B. The Liquidity Squeeze 56 C. Assistance to Firms in Difficulties 58 D. Enterprise Accountancy 60 Chapter VI: STRENGTHENING THE BANKING SECTOR 63 A. Introduction 63 B. High Cost of Intermediation 64 C. Undercapitalization 65 D. Maturity Transformation 66 E. Nonperforming Loans 67 F. The Interbank Market 69 G. Stabilization of the Banking Sector 69 H. Interlocking Ownership and Flow of Funds Between la-k. and Corporations 70 I. Shortcomings of Accounting and Auditing System 71 J. Supervision of the Banking System 71  -iii - Page No. Chapter VII: THE CAPITAL MARKET 75 A. The Case for Strengthening the Capital Market 75 1. The Case for Equity 75 B. Constraints on the Development of the Capital Market . 76 1. Need for a Long-term Program 76 C. Strengthening the Regulatory, Supervisory and Institutiomal Framework 78 1. The Capital Market Board 78 2. DiscloE,ure of Information and the Screening of New Issues 78 3. Strengthening the Supervision of Securities Market Activities 79 4. Non-baik Financial Institutions 80 D. The Primary and Secondary Markets for Securities 82 1. The Bond for Money Markets 82 2. Public Sector Securities 82 3. Corporate Bonds 83 4. Proposals for a Reform of the Bond Market 85 5. The Momey Market 87 6. The Equity Market 87' 7. The Supply of Equity 88 8. The Demand for Equity 89 9. Tax Issues and the Equity Market 90 E. The Secondary Market 91 1. The Present Framework for Trading 91 2. Establishment of a New Stock Exchange 92 ANNEX A ANNEX B STATISTICAL AlNEX  - iv - LIST OF TEXT TABLES Page No. Chapter I. THE SIZE OF FINANCIAL SYSTEM AND THE MOBILIZATION OF FINANCIAL SAVINGS Table 1.1 Financial Aggregates 1 1.2 Corparative Financial Statistics 1980 2 1.3 Financial Ratios 2 1.4 Interest Rates on Deposits 3 Chapter II: THE GROWTH AND ALLOCATION OF CREDITS FROM THE BANKING SYSTEM Table 2.1 Direct Credits from the Finiancial System 9 2.2 Distribution of Financial Sector Advances 11 Chapter III. THE COST OF CREDIT Table 3.1 Cost of Credits to the Borrower in the Last Quarter of 1982 17 3.2 The Interest Cost of Resources from Deposits Available for Lending by Commercial Banks in 1982 20 3.3 Interest Cost of Resources from Deposits Available for Lending by Commercial Banks After January 1, 1983 21 3.4 Effects of the Proposed Changes in the Interest Rates Paid on Required Reserves 24 Chapter IV: SELECTIVE CREDIT POLICIES CONTROLLING CREDIT ALLOCATION Table 4.1 The Supply of Agricultural Credit 32 4.2 Agricultural Credit by Recipient 33 4.3 Export Credits Extended 38 4.4 Final Interest Cost to Borrowers and Subsidy to Banks 39 Chapter V: CORPORATE FINANCE Table 5.1 Private Enterprise and Industrial Sector Debts 55 Chapter VII: THE CAPITAL MARKET Table 7.1 Bond Issues 82 7.2 New Stock Listings 1978-82 88  TURKEY - COUNTRY DATA Population: 45.5 million (1981) GNP Per Capita: US51540 (1981) Amount: Average Annual Increase (2) Share of CDP at MarkeiPices!T (milliolL USS (at constant 1980 prices) (at current prices) Indicator at current prices) 1982 1965-70 1970-75 1975-80 1965 1970 197'S 1980 NATIONAL ACCOUNTS Gross domestic product /a 52,890 6.6 7.5 2.8 100.0 100.0 100.0 100.0 Agriculture 10,324 3.1 4.4 2.7 30.7 26.4 26.2 21.4 Industry /b 13,527 9.5 9.5 2.8 16.6 17.2 18.0 28.6 Services 26,073 8.2 8.0 3.7 42.9 46.5 46.0 44.3 Consumption 44,345 5.8 7.0 2.7 84.6 82.8 85.2 81.8 Gross investment 10,705 11.7 12.9 0.6 16.7 20.1 23.3 26.4 Exports of goods and NFS 7,618 7.9 7.3 4.4 6.1 5.8 6.1 7.1 Imports of goods nod NFS 9,778 11.2 13.8 -3.1 7.4 8.7 14.5 15.2 Gross national savings 9,250 11.6 11.9 2.4 15.8 18.8 18.1 18.3 Average Annual Increase (X) Composition of Merchandise Trade (X) (at constant 1980 prices) (at current prices) 1972-75 1975-80 1972 1975 1980 MERCHANDISE TRADE /c Merchandise exports 5,746 -6.1 2.8 100.0 100.0 100.0 Primary 2,317 -6.3 4.0 72.6 64.1 64.0 Industrial products 3,42S9 -5.8 0.9 27.4 35.9 36.0 Merchandise imports 8,734 11.2 1.2 100.0 100.0 100.0 Agriculture and livestock 176 27.9 -23.8 2.2 4.3 0.7 Mining and quarrying 212 17.4 6.8 1.2 1.6 1.8 Petroleum 3,640 5.4 11.0 9.9 17.1 48.8 Machinery and equipment 2,352 14.0 -12.1 45.0 35.6 18.2 Other industrial products 2,354 9.9 4.5 41.7 41.4 30.5 1978 1979 1980 1981 1982 PRICES AND TERMS OF TRADE GDP deflator (1980 - 100) 29.0 49.4 100.0 142.1 181.5 Exchange rate 24.3 31.1 76.0 111.2 162.6 Export price index 63.0 78.2 100.0 91.9 92.4 Import price index 61.2 71.9 100.0 101.8 102.6 Terms of trade index 102.9 108.8 100.0 90.3 90.1 As Y of GDP (at current prices) 1965 1970 1975 1980 1982 PUBLIC FINANCE Current revenue 15.0 22.6 22.0 19.8 17.0 Current expenditure 10.0 11.8 12.6 11.5 8.1 Surplus (+) or deficit (-) -2.0 -2.3 -0.4 -4.8 -1.0 Investment expenditure 4.7 5.7 4.2 3.9 3.9 Transfers 5.0 7. 5 5. 5 9.2 6.0 Foreign financing 1,8 1.6 0.3 0.2 -0.01 1965-70 1970-75 1975-80 OTHER INDICATORS GNP growth rate (X) 6.8 7.7 2.6 GNP per capita growth rate (Y) 4.1 5.0 0.3 ICOR 2.9 2.9 5.7 Marginal savings rate (X) 28.2 19.5 30.8 Import elasticity 1.7 1.8 -1.3 /a At market prices; components are expressed ac factor cost and will not add due to exclusion of net indirect taxes and subsidies. 7T Includes mining and quarrying, manufacturing, and electricity, gas, and water. /c In accordance with Turkish Government's specLfications, which are not compatible with SITC's.  TURKEY - BALANCE OF PAYMENTS * XENLCAPITAL, AIND DEBT /a (million US$ at current price-) Population: 45.5 million (1981) GNP Per Capita: USi1540 (1981) Actual Proijeted 1970 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 BALANCE OF PAYMENTS Net exports of goods 6 NFS 342 1984 2442 4658 3476 -2021 -1960 -1788 -1771 -1956 -2229 Exports of goods 6 NFS 754 3075 3247 4102 6416 7620 8894 10613 12523 14545 16909 Imports of goods 6 NFS 1096 5059 5689 8760 9892 9641 10853 12402 14293 16501 19138 Workers' remittances 273 983 1694 2071 2490 2187 2350' 2450 2622 2779 2946 Net transfers 91 - - - - - - - - - - Current account balance -58 -1741 -1771 -3207 -2089 -1035 -861 -703 -489 -607 -799 Direct private investment 92 147 200 148 129 125 127 131 144 158 174 Public MRLT (gross) /b 271 1017 4321 2354 2188 2076 1756 1716 1652 1890 2193 Amortization on M4LT /b -146 -336 -414 -434 -545 -1230 -1105 -1258 -1783 -2063 -2233 Public M5LT (net) /b 125 681 3907 1920 1643 846 651 458 -131 -173 -40 Other capital /c 27 1061 -2410 1642 983 264 131 -160 764 1119 1070 Change in reserves (- - increase) -186 -148 74 -5tl3 -667 -200 -48 274 -288 -339 -405 International reserves 612 874 800 1303 1970 2076 2124 1850 2138 2477 2881 Reserves as months of imports 7 2 2 2 2 2 2 2 2 2 2 Actual 1972 1977 1978 1979 1980 1981 1982 GROSS DISBURSEMENTS Gross dssb.rsements of M5LT loans 372 759 857 4198 /d 2279 2116 2020 Official grants - - - - - aRi -m Concessional 261 193 228 588 812 522 355 Bilateral 139 100 129 406 749 499 283 IDA 4 19 8 3 - - - Other multilateral 118 74 91 179 63 23 71 Non-concessional 111 566 629 3610 /d 1466 1294 1465 Official export credits 1 47 133 250 288 355 400 IBRD 25 146 165 277 313 454 665 Other multilateral 27 5 35 15 150 162 147 Private 58 368 296 3068 /d 715 323 253 EXTERNAL DEBT Debt outstanding and disbursed 2450 4293 6322 10942 13415 13804 14296 Official 4273 3657 T489 189 8281 8906 10103 IBRD 92 512 648 890 1158 1546 2144 IDA 99 181 188 190 189 188 187 Other 2082 2964 4653 6109 6934 7172 7772 Private 177 634 833 3753 5134 4898 4193 Debt outstanding including undisbursed 3560 712.8 9879 14620 16807 17093 19561 DEBT SERVICE Total debt service /e 224 363 428 627 1001 1231 2385 Payments 161 196 264 403 405 546 1229 Interest 63 167 164 224 596 685 1155 Total debt service as 2 exports of goods + NFS + workers' remittances 11.8 10.3 10.6 12.7 16.2 13.8 24.2 Totil debt service as 2 GNP 1.3 0.8 0.9 0.9 1.7 2.0 4.4 Average interest rate on new loans (;) 4.4 7.7 6.9 11.3 6.5 7.6 7.3 Official 4.5 7.7 5.6 3.5 5.5 5.3 6.2 Private 6.8 7.8 8.2 13.6 10.6 15.4 12.0 &verage maturity of new loans (years) 22.1 11.6 13.3 11.1 17.4 15.0 18.4 Official 26.0 12.7 15.2 25.1 16.6 16.3 21.1 Private 11.0 9.2 7.6 7.1 6.4 4.5 8.0 BANK GROUP EXPOSURE (Z) IBRD DOD/total DOD 3.7 11.9 10.2 8.1 8.6 11.2 15.0 IBRD disbursements/total gross disbursements 6.7 19.1 18.4 6.5 13.7 25,0 32.9 IBD debt service/total debt service /e 5.1 17.2 19.2 16.8 13.3 14.0 9.0 IDA DOD/total DOD 3.9 4.2 3.0 1.7 1.4 1.4 1.3 IDA disbursements/total gross di3bursments 1.1 2.5 0.9 0.1 - - - IDA debt service/total debt service /e 0.4 0.6 0.5 0.4 0.3 0.2 - As I of Debt Outstanding at End of Most Recent Year (1982) TERMS STRUCTURE Maturity structure of debt outstanding (2) Maturities due within 5 years 39.9 Maturities due within 10 years 79.5 Interest structure of debt outstanding (2) Interest due within first year 6.5 /a All entries on external debt are defined as in the Bank's Debtor Roporting Systm. (only public and private guarant*ed debt). 7i Includes private guaranteed and non-gueranteed debt, debt relief, and grants. /c Includes errors and emissions, and for projected years it includes net IMP, short-term, and unidentified capital inflows. /d Includes t2.638 million of consolidated short-term debt. /e Takes account of debt relief due to debt rescheduling, and excludes interest on short-term debt and private non-guaranteed.  (i). SUMMARY OF CiDNCLUSIONS AND RECOMMENDATIONS I. Introduction 1. By comparison with other countries of similar and even lower per capita income, the financial system of Turkey is very small. Its size declined substantially in real terms during the second half of the 1970's, as measured by deposit bank assets/GNP, as well as by M2/GNP. After interest rates were liberalized and inflation started to decelerate, the average real interest rate paid on sight savings and time deposits rose from -46 percent in 1980 to +18 percent in 1982. On six-month time deposits, the real interest rate reached 25 percent in 1982 before taxes and 11 percent after the withholding tax. In consequence, the demand for M2 rose from 15 percent of GNP in 1980 to 22 percent in 1982, but in that year it was still below the level of 24 percent reached in. the mid-1970s, and below that of other middle income countries. 2. While this increase i.n deposits did permit banks to extend credits beyond what would otherwise have been possible, the increase of deposit interest rates to very high levels, together with high intermediation costs of the banking system and heavy t:axation of financial operations, gave rise to difficult problems. Lending interest rates increased sharply, particularly on non-preferential credits. It is estimated, although very roughly, that alt the end of 1982 the average annual interest cost paid by non-agricultural enterprises, excluding State E5nterprises, artisans and small traders, were of the order of 50 percent in nominal terms or 20 percent in real terms. For many credits these rates exceeded 70 percent in nominal terms or 35 percent in real terms. 3. Few enterprises in Turkey can earn enough on their operations to pay such high real rates on any substantial fraction of their debt. For those firms which are profitable, the effective burden is eased since such interest payments are tax deductible, Iriowever many firms are being forced to borrow further from the banks in ordier to finance part of their interest payments. For such firms, there is the dauger that the debt-equity ratio will rise progressively and that over an extended period, this could lead to insolvency. Under such pressure, many firms are thus attempting to increase their equity rather than their debts. 4. Since a large part of the interest payments had to be refinanced, the demand for credit remained very strong, even with declining investment in. inventories and fixed capital. Credits to the private sector expanded by about 48 percent in nominal terms during 1982, but a large proportion of that increase was absorbed by the refinancing of high interest payments. The stock of M2 grew by 57 percent during the same period, but more than half of the increase was accounted for by the reinvestment of interest accrued in timie deposits. Taking into account that the wholesale prices increased 25 percent in the course of 1982, it is thus understandable that both banks and industrial corporations have faced serious liquidity problems, in spite of the rapid expansion of credit and. money supply. This explains why banks have. intensified their competition for deposits.  (ii) 5. The authorities are faced with three objectives. First, they must continue to encourage the growth of the financial system. For that purpose, they must maintain attractive real rates of interest for financial assets. Second, they must pursue their efforts to reduce inflation. That requires the maintenance of strict controls on the expansion of credits. Third, they must seek the rapid reduction of the excessively high real lending interest rates for most private credits. This objective is to a significant extent in conflict with the first two; among other mneasures, it would require the reduction of interest rates on deposits, and it might also make it necessary to relax the credit policy. 6. The recommendations in the first ithree chapters of the Report take all these three objectives into consideration. But in face of the conflicts between them, compromise solutions are indispensable. The balance of the compromise will, hqwever, tend to change in face of future developments of the Turkish economy. 1/ II. Mobilization of Financial Savings and the Cost of Bank Credits 7. In order to promote the mobilization of financial savings and reduce the cost of bank credits, the following recommendations are made in the Report: (i) the expansion of the financial system should continue to be stimulated. The most frequentiLy suggested instrument for increasing the mobilization of financial savings is the payment of higher interest rates on deposits. However, since the lending interest rates are too high, there is limited scope for raising time deposit rates further. Okn the contrary, it would be desirable that they should fall gradually in real terms. However, in order to avoid risks of a substantial slowdown of the growth of real M2, or even of negative growth of that aggregate, the decline in the real interest rates for deposits should be combined with reductions of the withholding tax as proposed in recommendation (iv) below. In any case, in the next few years, the real interest rates for sight savings deposits, which at present are close to zero, should not be allowed to become significantly negative and those for time deposits should not be lower than 5 percent after the withholding tax; (ii) it will probably be necessary to impose progressively lower ceilings on deposit interest rates in order to achieve their 1/ Since the report was written, some of the report's policy recommendations have been introduced. For example, the Banking Reform Act was enacted and there has been some reduction in the interest rates on deposits. These changes are footnoted in the report.  (iii) gradual reduction.. 1/ At the urging of the authorities, the lending banks have already decided to lower the ceilings on interest rates on. time deposits from January 1, 1983. This decision may help improve the profitability of banks, but at the end of the first semester of 1983, it was still unclear whether it would be sufficient to bring down real lending interest rates to more acceptable levels. Further reductions in the interest rate ceilings on time deposits were introduced in July 1983, togelther with reductions in the lending rates. While this may help the financial position of some firms, if the liquidity situation remains very tight, there is the danger that the ceilings will be circumvented by the practice of blocking a certain percentage of the loans in comLercial deposits; (iii) if there was a persistent lag in the inflationary expectations of the population in, face of significant declines in the actual inflation rates, it might be worthwhile to consider the possibility of allowing banks to offer indexed savings depos:its. However, it appears that the very high real interest rates wlhich have recently been paid on time deposits have their origin mainly in the pressures of bank competition resulting from the tight liquidity situation and not in the inflationary expections of the savers. Moreover, if a system of indexation was introduced it would have to be extended to credits and it would create rigidity in the conduct oi- monetary policy, which would tend to make the fight against inflation more difficult; (iv) consideration should be given to further reductions of the withholding tax. In January 1983, the rate of the tax was cut from 25 to 20 percent for deposits and bonds whose owner is identified. The withholding tax is levied not only on real interest income, but also on the component of the interest rate which compensateii inflation. In principle, the withholding tax should be levied only on the real portion of the interest revenues. However, the administrative difficulties of introducing a system of that type might be considerable. In addition, if the inflation rates continue to decline rapidly, the problem will become progressively less important. But if the real interest rates and the infWlation rate remains at high levels, the authorities should consider reducing further the rate of the 1/ At the time this report a-s written, the Government's official inflation target for 1983 of 20 percent seemed reasonable, in view of the Government's success over the previous three years of reducing inflation from over 100 percent to about 25 percent. However, estimates of inflation for the first nix months of 1983 indicate that this target is likely to be exceeded by 5 to 10 percentage points. The Government 'has indicated that it remains committed to policies designed to reduce inflation, and on this basis the recommendations of this report would still be valid. If, however, the accelerating inflation which occurred prior to the 1980 economic reforms should resume, then this recommendation would no longer be appropriate.  (iv) withholding tax by a significant margin, in spite of the budgetary difficulties which would have to be faced. This reduction should preferably be combined with reductions in the ceilings for deposit interest rates. It should take place, however, only if there are good prospects that it will contribute to reduce lending interest rates or that it will be positively reflected in the demand for real M2; (v) the costs imposed on the operations of the banks by the liquidity and reserve requirements and by the transaction tax should be reduced. The measures introduced by the authorities in January 1983 contributed to that objective, but they should be reinforced by: (a) a reduction of the ratio of the reserve requirements or an increase in the average interest rates paid on those requirements. Either of these solutions would result in reductions, which might be substantial, of the cost of the funds used by banks in their credits, but at the same time the subsidization of selective credits would have to be reduced. Thus, in order to maintain the control of the money supply, the reduction of the ratio of the reserve requirements would have to be accompanied by the reduction of credits extended by the Central Bank to the public sector or as rediscounts to other banks, and would therefore be negatively reflected in the amount of selective credits refinanced by the Central Bank. Similarly, in order to avoid a decline in the profitability of the Central Bank, the payment of higher interest rates on reserve requirements would have to be compensated by higher rates of rediscounts charged by that bank and would therefore have a negative impact on the cost of selective credits. In any case, the interest rates paid on the reserves required from time deposits should be substantially higher than those paid on reserves from sight deposits and in particular from commercial deposits. Such a change would reduce the existing wide disparity between the costs to the banks of funds from time deposits and from sight deposits; (b) the elimination of the transactions tax of 15 percent at present applied to the interest earnings of credits and other revenues received by the banks. The transactions tax has a strong negative impact on the costs of financial intermediation and consequently on savings and investment; although it is a significant source of fiscal revenues (4.5 percent of the total) it should be replaced by other taxes with less adverse effects on economic growth. If the elimination of the tax in one step is not feasible because of budgetary constraints, it should at least be made in two or three steps, which should take place within a predetermined period of, for instance, two years. The tax in interbank transactions should be abolished immediately, since it is not producing any fiscal revenues and it is an important obstacle to the development of an interbank market;  (v) (vi) the contributions to the Interest Rate Rebate Fund should be maintained. Although such contributions add a significant charge to the cost of non selective credit, their reduction would omly be possible if the expenditures of that Fund on the subsidization of selective credits were sharply curtailed; (vii) the reduction of lending interest rates should also be assisted by cutting the excessively high intermediation costs of the banking sector. The measures which are required for that purpose are mentioned below in section 4.1. However, it is not realistic to expect that the reduction of the operating costs of the banks can be achieved very rapidly, given the structural nature of the problems to be faced; and (viii) since lending interest rates are determined mainly by the supply and demand of credits, it is not certain that the cost reductions proposed in recommendation (ii) and (v) would be passed on to the borrowers, if the liquidity of the banks and enterprises would remain very tight in the near future. The evaluation, after a period of several months, of the actual effectiveness of the measures introduced on January 1, 1983, in reducing lending interest rates may provide important guidelines for future action. If the reduction in such rates is significantly lower than what would be desirable, this may suggest that the totaL supply of credit is a constraining factor. If so, the Turkish authorities will have to make a difficult choice between two conflicting objectives: either to hold to their targets for a rapid reduction of inflation while maintaining very high real interest rates; or to try to reduce substantially lending ral:es and run the risk of not being able to achieve the desired improvements in the rate of inflation. It will be essential to avoid the rekindling of inflation and at the same time it must be recognized that high real interest rates have positive effects not only on the demand for financial assets but also as regards the improvement of thie debt/equity ratios of enterprises, since enterprises will tend to show more interest in equity financing as opposed to credits. But in spite of that it would be desirable to reduce as far as possible the conflict between the objectives of the stabilizationL policy and the need for bringing real interest rates to a lower level. Several types of measures recommended in the Report may help to resolve the conflict, but their contribution will be only partial. In addition to what is recommended above with regard to the transactions tax, the withholding tax, ceilings to interest rates on deposits and interest rates paid on required reserves, such measures include: (a) further efforts to reduce the borrowing requirements of the administrative public sector in order to create room for a faster expansion of credits to the productive sector within the constraints of the overall credit supply. It must be emphasized that substantial progress has been achieved since 1980 in reducing the borrowing requirements of the Treasury from the banking system. In spite of that, given the scarcity  (vi) of financial resources in the Turkish economy, it would be desirable to generate a financial surplus in the budget, in order to help financing the productive sector. A further reduction of the borrowing requirements of the administrative public sector could also be achieved if a significant volume of issues of Treasury bonds and bills was placed in the market at competitive interest rates. The marketing of such instruments could increase the demand for financial assets thereby contributing to a net increase of the resources available for credits to the productive sector; (b) adjustment of the total supply of credits in line with the increase in the demand of M2; (c) increase in the availability of general credits, by reducing the share of total funds at present allocated to selective credits, in accordance with recomendation (iii) of Section 2.1 below; and (d) rapid development of capital markets and, in particular, the revitalization of the bond market so as to contribute to the increase of the demand for financial assets and, consequently, to the expansion of the real amount of finance available for firms. III. Selective Criedit Policies Controlling Credit Allocation 8. The complex system of selective credits in Turkey results in fragmentation of the financial market and in extremely wide differences between the costs of different types of credits: while, for instance, agricultural credits pay 20 percent, the cost of non-preferential credits often exceeds 70 percent. Moreover, the supply of preferential credits has been severely restricted and they have to be rationed. This results in arbitrary allocations and delays. In addition, selective credit policies have a substantial impact on the cost of non-selective credits, through the reserve requirements imposed on deposits and the contributions to the Interest Rate Rebate Fund. 9. In these conditions, it is reconmended that; (i) at a minimum the interest rates in all special credits should be equal to the rate of inflation. But in the present situation, in which the real interest rates on general credits are abnormally high, this will not be sufficient. As long as the real interest rates on non-preferential credits exceed 20 percent, the real interest rates in preferential credits should be at least 5 percent; (ii) the large variety of interest rates on preferential credits should be reduced; and  (vii) (iii) the priorities in ,selective credit policies should be defined more narrowly in order ito reduce market fragmentation and to improve the allocation of scarce credit funds. A larger percentage of funds should be al:located to general credits. Agricultural Credits 10. Although the real level of agricultural credits is now much lower than in the mid-1970s, the proportion of those credits financed by the Central Bank increased from 30 percent in 1971 to more than 80 percent in 1981. The interest rates in agricultural credits are of the order of 20 percent to 22 percent1 as compared with, for instance, around 30 percent to 40 percent in export credits and 60 percent to 70 percent or even more in general short-term credits. A large proportion of agricultural credits consists of credits to agricultural sales cooperativeis and it may be argued that they are trade credits rather than credits to the farmers. Very little credit is going directly to small farmers, who often have to get other types of financing at extremely high interest rates. 11. The following recommendations on agricultural credit policy should be considered: (i) interest rates on agricultural credits, which, with declining inflation, are at present close to zero in real terms, should not be allowed to become negative in real terms. In accordance with recommendation (i) of Section 2.1, the authorities should consider the possibility of raising those real rates to around 5 percent if the real costs of other credits continue to remain very high; (ii) significant improvements in the operational efficiency of the Agricultural Bank should be sought, among other means, by encouraging more competition in the agricultural credit market. It is unlikely thal: commercial banks or private investment banks would enter the agricultural credit market in a big way, but they could become interested in extending credit to agro-industrial enterprises and large farmers if they were given equal access to rediscounting faci]Lities for agricultural loans at the same terms given to the Agricultural Bank. This could release some of the funds of the Agricultural Bank to small and medium-scale farmers; (iii) the role of TCZB and of the Central Bank in financing crop purchase activitiesi should be more clearly defined. Given the cbnstraints on Central Bank credits, it may be preferable for the conduct of m6netary policy that part of the loans needed by the sales cooperatives be provided by TCZB rather than by the Central Bank alone; (iv) it may be necessary to increase input financing over the medium-term, especiially because of the increases of fertilizer prices which will r esult from the gradual phasing-out of the fertilizer subsidy;  (viii) (v) medium and long-term credits, which at present account for less than 20 percent of total agricultural credit, should be increased within the constraints imposed by the availability of resources; the interest rates on such credits should be adjustable. (vi) the authorities should study the possibilities of improving the allocation of credits to small farmers; the share of such credits in total credits of the Agricultural Bank should be increased, but it must be recognized that the achievement of such an objective will be difficult. Those credits could be greatly assisted by a credit risk guarantee scheme. Export Credits 12. The share of export credits in total loans from the banking sector has steadily increased in recent years with the exception of 1982. The subsidization of the interest rates has been justified by the need to correct the bias against exports, stemming from the protection of the domestic market. The preferential treatment of export credits has been a crucial factor in stimulating exports in a situation in which credit availability was severely constrained. In 1982, the rediscounts of export credits decreased and the interest rates paid on those credits increased to about 8 percent to 10 percent in real terms. 13. In order to deal with the difficulties currently being faced in the area of export credits, the following recommendations should be considered: (i) the share of export credits in total credits may have to be enlarged; even with an increase in the proportion of export credits financed by commercial banks on the basis of their own resources, those credits should continue to receive a high priority in the rediscounting policy of the Central Bank; (ii) since in any case the refinancing of export credits will continue to be subject to severe constraints, it will be necessary to allocate them more equitably, either by lowering the portion of the credit which is rediscountable at the Central Bank or by shortening the maturity of the rediscounts for a faster turnover; (iii) the amount of subsidized export credits should be related to domestic value added in the goods exported. However, a system of that type might be difficult to implement from an administrative point of view. At present exports with less than 20 percent of domestic value added are excluded from the benefits of preferential export credits. There may be justification for raising their percentage to 25 or 30 percent; (iv) a system of medium and long-term export credits should be established in order to support the expansion of exports of capital goods and engineering services;  (ix) (v) an export credit insurance scheme to cover the risks usually covered in most countries which export manufacturing goods, slhould be introduced; and (vi) the authorities should study the possibility of introducing a system under which indirect exporters could also have access to preferential export credits. Housing Credits 14. The share of housing credits in total credit from the banking system is extremely small, but it camnot be easily increased given the overall constraints on credit supply and the long-term nature of such credits. The interest rates charged to the borrowers have been highly negative in real terms, but there is a tight rationing of credits. 15. In order to improve the system of housing credits in Turkey, the following recommendations should be considered: (i) the real interest rates on those credits should be increased, in accordance with recommendation (i) of Section 2.1 above; (ii) a scheme of floating mortgage housing loans, with partial refinancing of interest payments, should be envisaged, in accordance with recommendations of Section 2.6; (iii) in an environment of high and variable inflation, bonds with adjustable interest rates, mentioned in recommendation (iii) of Section 5.3 may be a viable financial instrument in mobilizing resources for housing credits; and (iv) as the availability of resources improves, the upper limit of housing credits should be increased gradually, since at present the average rise of loans is far from adequate to cover the cost of housing. Credits to State Economic Enterprises 16. The share of SEEs in total credits from the banking system, including development and investment banks, which had reached 36 percent in 1979 declined to about 28 percent by the end of 1982. Credits from deposit money banks to SEEs are subsidized by the Interest Rate Rebate Fund, which grants a subsidy of 4 percentage points to the borrower and 1 percentage point to the lender. In addition, the State Investment Bank grants medium- and long-term loans to SEEs with a fixed interest rate of 21.5 percent, which is heavily subsidized. Those subsidies impose a burden on the Treasury and on other credits and distort the social profitability calculations of the SEEs.  (x) 17. It is recommended that: (i) the special interest rate subsidies to SEEs should be eliminated. Those enterprises should pay the same interest rates as private enterprises. This implies that they should have the same conditions of access to preferential credits; and (ii) SEEs with a more solid situation should try to cover part of their financial needs by issuing bonds in competitive market terms. Medium and Long-Term Credits 18. About 60 percent of medium and long-term credits are distributed by deposit money banks, but a large proportion of them is directed to companies of the banks' own groups. Investment and development banks, which concentrate their activity practically entirely in medium and long-term credits, are better equipped than commercial banks to appraise the quality of the projects and to follow their implementation. However, their share in total credits has been declining rapidly, due essentially to lack of resources in domestic currency. The volume of medium and long-term credits has been negatively affected by the risks resulting from the system of fixed interest rates in an inflationary context, by the exchange risks of loans in foreign currency, and by the liquidity problems created for the borrowers by high nominal interest rates. 19. In order to deal with the main problems in the area of medium and long-term credits, the following recommendiations are presented in the Report: (i) investment and development banks should try to mobilize more resources in domestic currency by issuing medium-term bonds to be placed in the market. The interest rates of such bonds should be adjustable in accordance with the system mentioned in recommendation (iii) of Section 5.3. The Government might assist the investment and development banks by providing guarantees for their bonds if that could contribute to lower significantly the bond. interest rates; (ii) the State Investment Bank should have the same conditions of access to domestic and foreign currency resources as the other investment and development banks. Loans granted by the State Investment Bank to SEEs should receive the same subsidies from the IRRF and benefit from the same rediscounting facilities in the Central Bank as loans with similar characteristics granted to private enterprises. But the special privileges of the State Investment Bank in the utilization of low interest rate loans from the Treasury and from the Social Security System should be eliminated;  (xi) (iii) given the needs of a strict control of money supply, it would not be realistic to envisage a rapid growth of Central Bank rediscounts of medium and long-term credits. If reserve requirements are reduced, the amount of such rediscounts may even have to be cut. However, their share in Central Bank credits could be increased if it was possible to reduce the borrowing requirements of the Treasury and the share of the Central Bank in the financing of crop purchases (TMO, Tobacco monopoly and agricultural sales cooperatives). Besides, the rediscounts of medium and long-terum credits should be more equitably distributed among all development and investment banks instead of being granted mainly to the State Investment Bank; (iv) the Central Bank rediscounts of medium and long-term credits should depend more strictly upon the evaluation of the projects being financed; at present, such evaluation is conducted by the State Planning Office; there might be some benefits in contributing also for that purpose the project evaluation capabilities of investment and development banks; (v) in view of the unceirtainties associated with high and unpredictable inflation, a clearly defined system of floating interest rates for mnedium and long-term credits should be introduced. These reference interest rates could be chosen equal to the average cost of resources of commercial banks, to be computed and published periodically by the Central Bank, plus a. spread fixed by the authorities or, preferably, determined by market conditions; (vi) the system of floating interest rates proposed in recommendation (v) would not create any special difficulties for commercial banks, but it could lead to lending rates by the investment and development banks significantly lower than their average cost of resources and, in particular, lower than the cost of the floating interest rate bonds proposed above in recommendation (i). In order to correct this difficulty, the Interest Rate Rebate Fund should grant to the investment and development banks a subsidy on their medium and long-term loans equal to the difference between the average cost of the resources of those banks and of commercial banks, minus the estimated difference between the average operations costs of both types of banks. At present, the IRRF is already providing subsidies to banks engaged in preferential credits, but on different groundsB; (vii) given the high cost: of the resources of commercial banks, under present conditions the interest costs of medium and long-term credits would be too high without subsidies. Some subsidies from the IRRF may thus continue to be necessary. However, in view of the scarcity of resources for subsidization and of the need to  (xii) alleviate the burden imposed by subsidies on non-selective credits, the level of subsidization should be reduced and the disparities in the subsidies provided to the different types of medium and long-term should be gradually narrowed; (viii) In spite of the losses with exchange risks which borrowers of loans in foreign currencies suffered in recent years, there are no strong reasons to fear similar losses in the future if the Turkish authorities keep their present exchange rate and interest rate policies. If, however, borrowiers continue to show reluctance in accepting foreign currency medium and long-term credits, it may be useful to create a special funid to cover the exchange risks of such credits. Borrowers wishing the coverage of exchange risks should pay to the fund a contribution equal to the difference between the interest rates of medium and long-term credits in TL and in foreign currencies. Foreign credits, benefiting from the exchange risk coverage, should receive the same subsidies from the IRRF as if they had been obtained on TL in the domestic market. On the basis of present policies, and with such contributions, it is likely that the fund would accumulate profits rather than suffer losses; and (ix) the high level of nominal interest rates creates serious liquidity difficulties for the borrowers of medium and long-term loans. In order to reduce those difficulties, the lending banks should refinance part of the interest payments, corresponding, for instance, to the inflation of the previous year less a margin of 5 percent or, alternatively, they should increase the initial amount of the loan by, say, 20 percent and retain the increase for the payment of part of the interest of the first two years. The refinancing system could however be eliminated once the nominal interest rates in medium and long-term credits decline to levels below 20 percent. IV. Corporate Finance Rehabilitation of Industrial Firms 20. The profitability of Turkish industrial enterprises deteriorated sharply in 1982 in consequence of the recession of the domestic market and, especially, of the jump of real interest rates to very high levels. However, these difficulties have been partially eased by the decline in real wages and more recently by tax reductions. The average debt/equity ratios of industrial enterprises, estimated on the basis of revalued assets, are still satisfactory, but there is the risk that they will become progressively worse if the present high real interest rates persist. The high interest rates have also created a severe liquidity crisis for most enterprises. Firms were forced to increase the real value of their debts, without a similar increase  (xiii) in real assets, in order to finance their losses and to refinance their hilgh interest payments. In such corLditions, the pressures on the Government to assist the rehabilitation of industrial enterprises have been growing. This complex problem involves many aLreas of economic policy, apart from credit and subsidies (for instance, taxation, labor regulations, trade policy, etc.). Concerning financial measures, the following recommendations should be considered: (i) one of the most important contributions to the rehabilitation of industrial enterprises should come from the reduction of real interest rates. The measures dealt with above in Section 1 are oriented towards that objective; (ii) the rehabilitation of enterprises in a particularly difficult financing situation often requires the increase of their equity. The high real interest rates are already contributing to that increase. It will be desirable to encourage further the increase in equity by tax incentives to the reinvestment of profits and to the partial conversion of existing liabilities into equity; (iii) a special Fund for the Rehabilitation of Industry may also be necessary. To the extent possible, the resources of the Fund should come mainly from external credits, but they might be complemented by budgetary allocations and by Central Bank rediscounts. The Fund could provide medium-term credits to meet working capital needs, with priority given to the partial refinancing of the interest payments of medium and long-term credits and to easing the difficulties with the service of credits denominated in foreign currency. A scheme of exchange risk coverage of the type mentioned in recommendation (viii) of Section 2.6 should be applied to the loans of the Fund. (iv) a scheme of exchange risk coverage of the type mentioned in recommendation (viii) of Section 2.6 should be applied to the loans of the Fund. In order to ensure the efficient use of the resources of the Fund for Rehabilitation of Industry, the following principles should be followed; (a) there should be careful evaluation of the firms to be assisted; banks and authorities should accept that firms of dubious economic merit be liquidated; (b) the assistance should be made conditional upon commitments from the owners and managers of enterprises to improve their performance; (c) the administration of the assistance program to each firm should be entrusted to reliable institutions; the use of the expertise and of the operational capacity of investment banks could be very helpful; and  (Xiv), (d) if a system of assistance to banks with liquidity difficulties was simultaneously in operation, it should be combined with the program for the rehabilitation of industrial enterprises. Enterprise Accountancy 21. The shortcomings of the systems of accounting and auditing in Turkey constitute a major obstacle to the efficient management of both SEEs and private enterprises and to the development of financial and capital markets. In addition, the accounting results of recent years are grossly distorted by the effects of inflation on the value of assets, on the costs of depreciation, and on the real cost of the liabilities. In order to correct those shortcomings, it is recommended that: (i) the Government should create a commission to establish, if necessary with the technical assistance of experts from reputable accounting institutions, an action plan for the phased development and implementation of modern accounting procedures; (ii) the plan mentioned in (i) should cover: (a) the assessment of the manpower, educational, and training requirements for accountants and auditors; (b) passing as soon as possible legislation for the organization and control of the accotnting profession; (c) the establishment of accepted accounting and auditing standards; and (d) changes in the statutory accounting and auditing requirements. V. Strengthening The Banking Sector High Costs of Intermediation of the Banks 22. The intermediation margins of Turkish banks, including operating costs and profits, corresponded on average in 1980 and 1981 to more than 11 percent of their total assets in 1980 and 1981 and to 8.1 percent in 1982. The comparable ratios for industrialized countries are of the order of 3 to 4.5 percent. The excessively high costs of bank intermediation are one of the major obstacles to the reduction of real lending interest rates to a more sustainable level. This is a structural problem, and it would be unrealistic to expect that it may be solved too soon. However, in order to increase the pressure on banks to improve their productivity, it is recommended that:  (xv) (i) the growth of the operating expenditures that can be deductedl from total revenues for the purpose of determining the tax profits should be subject to a limit set significantly below the rate! of expansion of the value of total assets of each bank; (ii) the number of branches and the revalued real estate assets of each bank should be subject to limits related to its total equity; the authorities are already considering the introduction of such limits; (iii) more competition should be created to the banking sector by revitalizing the bond market, in accordance with the recommendations of Section 5 below; since the bond market canI operate with low intermediation costs, it can become a highly effective instrument of competition with the banking system; and (iv) more competition should also be stimulated by reducing the specialization among banks, particularly as regards Government-owned specialized banks and investment and development banks. Undercapitalization of Banks 23. In order to solve the problem of undercapitalization, the authorities are already requiring a minimum of capital. The authorities are also giving consideration to the introduction of a capital adequacy requirement based on a solvency ratio as implemented in an increasing number of countries; that ratio would relate the amount of unimpaired own funds to total liabilities. That ratio could be set as a higher level in the case of development and investment banks, given the lower liquidity of their assets. Maturity Transformation 24. Excessive maturity transformation by Turkish banks stems mainly from a worsening of the quality of- their loan portfolio and from the unavailalbility of resources with maturities in excess of one year. Although an improvement of this situation is dependent primarily upon economic and monetary policy measures, it is recommended t:hat: (i) the degree of trcmnsformation by the banks should be more closely monitored by the supervisory authorities; for that purpose, it would be useful t:o use the breakdown by maturities of the data on assets and liabi]Lities provided by the Central Bank on a monthly basis; and (ii) the authorities tihould put into effect their intention of improving a limil: on the banks' possibilities of investing depositors' savinigs into fixed assets and capital participations, equal to the amoamt of their net worth.  (xvi) Non-Performing Loans 25. The proportion of bad debts in the portfolio of banks is increasing, but the procedures for monitoring those debts are not satisfactory and the constitution of the provisions which they require has been far from sufficient. It is recommended that: (i) the criteria for classifying bad debts should be revised, and a tighter system for reporting thiose debts should be established; (ii) banks should not be allowed to accrue uncollected interest payments on loans when principasl or interest is in default for 90 days or more; and (iii) the conditions under which provisions for loan losses may be set up in the deduction of taxable income should be relaxed. The Interbank Market 26. An interbank money market could contribute usefully to ease the liquidity problems of individual banks, particularly when there are large and sudden shifts of deposits from some banks into others, as occurred recently. The creation of an interbank money market would require: (i) the immediate abolition of the financial transactions tax on interbank transactions; and (ii) maintenance or even increase of the penalties imposed on the banks which do not comply with the minimum liquidity and reserve requirements: if the cost of not meeting those requirements is sufficiently high, banks will show interest for an interbank market. Stabilization of the Banking Sector 27. The liquidity crisis of the corporate sector has been reflected in difficulties for the banks, which have been forced to refinance a large proportion of the amortization and of the interests in their loans outstanding. Those difficulties were compounded by deposit withdrawals from the weaker banks after the brokers' crisis in the middle of 1982. In order to ensure the stabilization of the banking sector, the authorities should consider the following recommendations: (i) the system of deposit insurance should be improved. At present, only the face value of the deposits, but not accrued interest, is guaranteed by the Government; in addition, the principal of the deposits is refunded to the owners only at the end of 3 years. In an environment of high interest rates, a guarantee of the interest due as well as the face value of the deposit may be needed to  (xvii) prevent deposit withdrawals. The deposit insurance scheme should cover only deposits up to a certain size, should not cover interest rates in excess of the announced rates, and should not cover commercial deposits; and (ii) mergers of smaller and weaker banks into larger and more solid ones may be desirable to strengthen the banking system, although to some extent they might contribute to increased concentration. Interlocking Ownership Between Banks and Corporations 28. The widespread interlocking ownership between banks and corporations results in distortions in the allocation of credits and in excessive concentration of lending risks in particular corporations. In the absence of an adequate equity market, the Government cannot require the divestiture of shares. However, in order to reduce the problems created by interlocking ownership of banks and corporations: (i) the authorities should put into effect their intention of imposing lower limits on the proportion of credits granted to large customers and to companies related to a bank; and (ii) the introduction of the solvency ratio suggested above in section 4.2 could also be used to attack the problem of interlocking interests. Shortcomings of Accounting and. Auditing Systems 29. There are no accounting standards for the banking sector in Turkey, and the reliability of the accounting and auditing system is far from satisfactory. Improvements of the accounting and auditing standards shouLd be coordinated with the measures proposed for the corporate sector as a whole in the recommendations of Section 3.2. The authorities have been considering the following changes which shouldl be implemented: (i) introduction of a system of standardized accounts and accounting procedures especially adapted to the requirements of the banking sector; (ii) refinement in the short-run of the contents of some sensitive items that banks show in their compulsory financial statements, since the implementation of the system mentioned in recommendation (i) will take time. Some basic accounting guidelines should lbe laid down, in particular for the recording of doubtful and bad debts, for the writing-off of loans, for accrued and uncollected interest payments,, and for provisions for loan losses;  (xviii) (iii) close monitoring by the Board of Sworn Bank Auditors (SBA) of the effectiveness of the auditing system of each bank and of the remedial actions taken to correct the shortcomings detected. However, SBA should not be perceived by banks as a substitute for deficient internal auditing systems. Supervision of the Banking System 30. The system of bank supervision has on the whole been considered satisfactory, but the tasks which it must handle have become increasingly complex, mainly because of the present economic and financial situation. In order to improve the effectiveness of the system, the following recommendations are made in the Report: (i) some rearrangements of the respective responsibilities of the various units in charge of the supervisory process are needed to eliminate current duplication of effort and to put the present operational capabilities to their best use. In particular the role of the Central Bank should be enlarged. The Central Bank should play a greater part in the provision and analysis of financial information on each individual bank and on the banking system as a whole. Individual information on the creditworthiness of banks' borrowers, based on the risk centralization scheme, on rediscount files and other sources, should be provided by the Central Bank to SBA. Central Bank staff members could also assist occasionally the SBA in the conduct of their on-site examinations. Bank examiners should be relieved from the inspection of brokerage firms, since this task will be of the responsibility of the Capital Markets Board; and (ii) bank examiners should be assigned mainly the role of checking into the quality of assets, the solvency, the liquidity, the profitability and the equity base of banks, and assessing their managerial skills by means of on-site examinations. Bank examiners should have the possibility of getting assistance from outsiders in the course of examinations of banks, and the Banking law should be amended accordingly. VI. The Capital Market Need for a Long Term Program for the Development of the Capital Market 31. The development of the bond market would create competition to the banking system and would permit a reduction of the intermediation margins, diversify the supply of financial instruments, and result in a better allocation of resources and financial savings. The need for an equity market is now higher than in the past in view of the deteriorating debt equity ratios of most companies and of their liquidity difficulties.  (xix) 32. The Capital Market Bcard, in cooperation with other relevant public agencies and with the private sector, should formulate a long-term progrua for the development of the capital. market in an orderly and regulated environment. The main focus of the program should be, as a long term goaL, developing an equity market. A shorter-term goal is the development of an efficient market for corporate bonds and other debt instruments. 33. The recommended actions are presented below: A. Improvement of the Regulatory, Supervisory and Institutional Framework (i) the capability of the Capital Market Board to regulate, supervise and develop the market should be strengthened by hiring an advisor and a number of qualified financial analysts and accountants, and by training; (ii) the improvement in accounting and auditing practices mentioned in the recommendationL of Section 3.2 will take time. Meanwhile, special requirements concerning accounting, auditing and disclosure of information should be imposed on firms operating or issuing securities in the capital market. Such requirements should include external auditing by licensed firms of at least all financial institutions and companies offering securities to the public or having a large number of shareholders. At a later stage, large closely-held companies should also be audited; (iii) once adequate accciunting, auditing and disclosure standards have been established, the role of the Capital Market Board in screening new issues of securities should be limited to monitoring compliance with disclosure requirements, leaving the appraisal of their quality to t:he market; (iv) the Capital Market Board should be fully responsible for the supervision of securities market activities. It should encourage the establishment of an association of licensed securities firms to set professional standards and help in the detection of unlicensed operators; (v) small over-the-counter dealers not engaged in underwriting should be allowed to remaLin in business until a viable securities exchange is established; (vi) the Securities Regulation Fund should be converted from a rescue fund to an ordinary financing facility for securities firms;  (xx) (vii) the investment policies of Social Security Institutions and of insurance companies should be reviewed to channel their funds increasingly towards the capital market, including private sector instruments. Consideration should be given to the encouragement of funded private pension funds and other contractual saving schemes; and (viii) establishment of mutual funds, especially open-end ones, should not be allowed until the secondary market is larger and more efficent. A. The Bond and Money Markets (i) The interest rates of public sector bonds should be made competitive with those of other financial instruments in order to make them more attractive to private investors; they should be marketed increasingly through officially recognized dealers and commercial banks, instead of being offered only through the Central Bank and the Agricultural Bank; (ii) in view of the present difficulties of the corporate sector, it is not desirable to favor a large increase in the size of the market for corporate bonds until auditing and disclosure standards have been improved and real interest rates have decreased. The recent asset revaluation law allows in principle large increases in bond issues, by making the net-worth ratio less stringent. However, it improves the ability of the firms to service their debts only to the extent that the revaluation of assets is accompanied by an increase of subscribed capital, implying an inflow of new equity funds. The recent imposition of a 30 percent withholding tax for bearer bonds and 20 percent on registered ones will likely have the effect of preventing a flow of new issues, since they will be more expensive. In spite of that, reserves and free capital generated by assets revaluation should be temporarily excluded from the computation of the ratio. For the time being, only the apportionment of new equity or of retained earnings should allow the issue of bonds by firms which have presently reached the ceiling; (iii) the issue of bonds with adjustable interest rates should be stimulated. According to the regulations published in January 1983, it is already possible to issue bonds with interest rates adjusted annually on the basis of interest rates for time deposits. Although so far such bonds have not met with great success, they can offer considerable advantages by allowing longer maturities while at the same time reducing the risks both to savers and to borrowers;  (xxi) (iv) the recent abolition of repurchase guarantees by underwriters should eventually lead to the establishment of a genuine secondary market, based on over the counter dealers, who have already started making a market by purchasing bonds at listed market prices reflecting demand and supply conditions. The dealers should be encouraged to deal increasingly in bonds underwritlten by others. In the longer run, once a secondary market has been established, and an efficient Securities Exchange organized, trading should be channelled mostly through the Exchange; (v) the repurchase guarantees which have already been eliminated as far as new issues are concerned, should be replaced by the introduction of sinking funds, as a mechanism to ensure liquidity and price stability; (vi) the restraints over bond interests and maturity requirements should be gradually relaxed; (vii) the introduction of convertible bonds, should be tried; and (viii) the introduction of commercial paper should also be envisagedl. The issues of such paper should be subject to approval by thea Board on the basi.s of disclosure standards similar to those of bonds and equity. They should also be subject to the same withholding tax as bonds. A market of that type will, however, be feasible only after the reform of the transactions tax, since otherwise there would be risks of unfair competition with the banks. The Equity Market 34. Many companies are now badly in need of additional equity which cannot be supplied by their controlling groups. At the same time, they are reluctant to sell to the public, largely because of disclosure requirements and tax considerations. This; reluctance may well be reduced when disclosure requirements can be imposed also on large closely held companies, as a result of improvements in auditing standards. In the meantime, tax incentives for "open" companies should be considered. Since it is more difficult for them to avoid taxes than for closely held ones, this will not necessarily result in a loss of fiscal revenues. 35. The demand for equity is negatively affected by: (a) inadequate disclosure of information; (b) low yields compared to other investments, due largely to extremely high current interest rates, low dividend pay-outs and the inadequacy of the secondary market to translat,e increases in retained earnings into a corresponding appreciation in share prices; and  (xxii) (c) to a lesser degree, tax factors. 36. In order to overcome these problems, the authorities should give consideration to the following recommendatiLons: (i) a study should be carried out to determine which measures would increase the supply of equity through providing incentives for closely held companies to make public offerings. Potentially profitable State Economic Enterprises should also be included in the review; in this connection, consideration should be given to taxing more heavily those companies above a certain size which do not have widespread shareholditng. (ii) the effects of the requirement of minimum dividend pay-out ratios adopted by the Capital Market Board should be closely monitored. The negative effects of such requirement on the rate of savings of the private sector and on the improvement of the financial situation of companies in a dilfficult situation can be avoided if the dividends distributed are immediately invested into capital increases as allowed by the present legislation. Even so there are risks that the minimum dividend requirement may result in a disincentive for firms to offer shares to the public; and (iii) the taxation of equity income should be reanalysed in the light of budgetary needs and national priorities for resource allocation. A possible tax credit or tax deferment for investments in equities, particularly by pension funds and insurance companies should be studied. The Secondary Market 37. The present lack of an efficient secondary market hampers the development of the primary market as financial assets are not provided with the necessary liquidity and there is no effective and centralized pricing mechanism. 38. The over-the-counter market, for asll its shortcomings, provides however at present the only workable framework for trading, and should be allowed to function until a new stock exchange is in place. Work on the development of a new stock exchange should be started at an early stage, but such exchange would not be useful and viable until the proposed measures to improve accounting, auditing and disclosure standards and to encourage the growth in the demand and supply of securities have been implemented. The steps for the establishment of a new exchange should be as follows: (i) at an early stage, the Capital Market Board should be given responsibility for developing the exchange and start a program of study trips to acquaint its staff with the organization of foreign securities exchanges;  (xxiii) (ii) amendment of the Capital Market Law to include the exchange among the "auxiliary inst:itutions", thus subjecting it to the authority of the Capital Market Board; (iii) repeal of the 1929 Stock Exchange Law; (iv) implementation of new exchange, including logistics, training of personnel and of new members, and internal organization; and (v) channelling of all trading on the exchange. VII. Conclusion 39. This extensive list: of financial reforms is clearly a tall order for any government to undertake. Their adoption and implementation would constitute a fundamental restructuring of the financial system in Turkey, and would take considerable time and effort. The current situation, however, calls for urgent action to relieve both banks and enterprises of growing insolvency. It is, therefore, crucial that certain reforms, particularly those relating to the cost of credit, be undertaken without delay.   CHAPTER I - THE SIZE OF THE FINANCIAL SYSTEM AND THE MOBILIZATION OF FINANCIAL SAVINGS 1.1 The financial system of Turkey is dominated by the banking sector. The proportion of financial assets held by private savers in bonds, easily marketable shares, insurance policies, and other investments issued by nonbank institutions is very small. W[n such conditions, the money supply M2 provides an adequate measure of the size and growth of the Turkish financial sector. A. The Size of the Financial System 1.2 In nominal terms, the financial system in Turkey has been growing rapidly. The money stock grew fifteen fold between 1975 and 1982, from TL 129 billion to TL 2,052 billion; total advances from the financial system grew only slightly less rapid:Ly, from TL 185 billion in 1975 to TL 2,672 billion in 1982. But the rapid nominal growth reflects the high rates of inflation, which reached 100 percent in 1980. In real terms the system grew much less and considerably below the growth rate of GDP. In 1975 prices, using the wholesale price index (WPI) as a deflator, M2 was equivalent to only TL 160 billion in 1982 (average growth of 3.2 percent per annum) and total credit equivalent to onLy TL 207 billion (1.6 percent per annum). 1.3 The overall figures conceal sharp fluctuations in the holdings of financial assets. As Table 1.1 shows, M2 expanded in real terms between 1975 and 1977, declined by 37 percent in the next three years, and then increased from 1980 to 1982. But in spite of the growth in 1981 and 1982, the real value of N2 in the latter year was approximately the same as in 1977. Table 1.1: FINANCIAL AGGREGATES (TL Billion) Money stock a/ 197.5 1976 1977 1978 1979 1980 1981 1982 Ml (nominal) 102 133 177 243 354 559 827 1,147 M1 (1975 prices) 102 115 124 111 99 75 81 90 M2 (nominal) 129 163 210 283 416 682 1,201 2,052 M2 (1975 prices) 129 140 147 129 116 92 118 161 a/ Geometric mean of end-year figures. The wholesale price index (WPI) was used as a deflator. 1.4 In Table 1.2 figures are presented on the size of Turkey's financial system in 1980 in comparison with a range of other countries. By two alternative measures of size, the Turkish financial system was smaller than in the other countries, including those with considerably lower per capita incomes. Although the Turkis'h system has grown rapidly in the last two yreairs, it still remains one of the smaller financial systems in relation to GNP.  - 2 - Table 1.2: COMPARATIVE FINANCIAL STATISTICS, 1980 Per capita Deposit bank M2/GNP Country income assets/GNP (1980 US$) (---- percent -----) Turkey 1,470 28 15 Argentina 2,390 52 31 Brazil 2,050 39 23 Chile 2,150 65 30 Jordan 1,400 156 140 Korea 1,520 92 35 Morocco 2,090 28 25 Philippines 650 51 30 Portugal 2,350 120 110 Thailand 690 51 41 B. The Role of Interest Rates in the Mobilization of Financial Savings 1.5 To understand why the Turkish financial system is small and why its size has fluctuated widely during the last seven years, one must analyze what has happened to interest rates. The second half of the 1970s in Turkey was a period of growing public sector deficits that were financed in large part by borrowing from the Central Bank. As a result, the rate of nominal monetary growth, as shown in Table 1.3, increased every year, reaching 67 percent in 1980. The rapid rise in the monetary aggregates--when coupled with the fiscal deficits, imbalance in the balance of payments, and other factors--caused a rise in the rate of inflation from 10 percent in 1975 to 100 percent in 1980. Table 1.3: FINANCIAL RATIOS (percent) Ratio 1975 1976 1977 1978 1979 1980 1981 1982 a/ Ml/GNP 19 20 20 19 16 13 13 13 M2/GNP 24 24 24 22 19 15 18 24 Ml/M2 79 83 83 86 84 82 70 56 M1/M2 a/ 74 76 79 81 79 74 56 40 M2 growth b/ 29 24 34 35 61 67 86 57 a/ Commercial sight deposits have been sulbtracted from both M1 and M2. b/ From the beginning to the end of the year. 1.6 During this period nominal interest rates rose but, as shown in Table 1.4, did so less rapidly than the rate of inflation. For example, the interest rate paid on 6-month time deposits had increased only nine percentage points from 1975 to 1980, although the inflation rate had risen by 90 percentage points. Because inflation was rising so much more rapidly than  -3- Table 1.4: INTEREST RATES ON DEPOSITS (In Percent) 10/1/74 4/1/78 5/1/78, 3/1/80 7/1/80 1/1/80 2/1/81 7/1/81 1/l/82 1/1/83 Public 1.0 0 0 0 0 0 0 0 0 0 Commercial 0 2.0 0 0 0 0 0 0 0 0 Savings 1/ Sight 3.0 3.0 3.0 3.0 5.0 5.0 5.0 5.0 5.0 20.0 Time 1 Month n.a. n.a. n.a., n.a. n.a. n.a. n.a. n.a. 25.0 n.a. 2 Months n.a. n.a. n.a,. n.a. n.a. n.a. n.a. n.a. 35.0 n.a. 3 Months 6.0 6.0 8.0 8.0 n.a. n.a. m.a. 45.0 45.0 n.a. 6 Months 6.0 6.0 8.0 12.0 15.0 32.0 42.0 50.0 50.0 40.0 1 Year 6.0 9.0 12.0 12.0 15.0 32.0 42.0 50.0 50.0 45.0 2 Years 9.0 12.0 20.0 20.0 33.0 40.0 50.0 50.0 50.0 n.a. 3 Years 2/ 16.0 22.0 34.0 40.0 40.0 50.0 50.0 50.0 n.a. 4 Years 2/ 20.0 24.0 35.0 40.0 50.0 50.0 50.0 50.0 n.a. more than 4 Years 2/ 2/ 2/ 36.0 40.0 50.0 50.0 50.0 50.0 n.a. Inflation 3/ 27.8 48.1 47.6 133.2 109.8 94.7 26.4 37.9 36.7 23.0 Real interest rates on 6-month deposits -17.0 -28.4 -26.7 -51.8 -30.9 -30.9 15.8 13.3 14.3 17.1 Source: Central Bank 1/ As of 7/1/80 deposits in foreign exchange by emigrant workers have been receiving an additional 5 percentage points. Previously, they were receiving an additional 10-15 percentage points. 2/ Interest rate determined between the bank and the depositor. 3/ Based on the change in the wholesale price index in the previous 12 months.  -4- deposit rates, the real rate paid on those deposits declined sharply, and it reached about -50 percent in 1980. In the mid-1970s, to hold financial assets in Turkey was not very attractive and this helps explain the small size of the financial system. But from 1977 to 1980, financial asset holding became progressively even less attractive, and this caused a flight out of financial assets (see Table 1.3). Between 1975 and 1980, the ratio of M2 to GNP fell by more than a third, from 24 percent to 15 percent; in other words, the velocity of money (M2) rose from 4.2 in 1975 to 6.7 in 1980. To sum up, large public sector deficits led to excess money creation, which in turn led to an acceleration in inflation. The increase in the inflation rate was not matched by higher bank deposit rates; there was a consequent flight out of financial assets into inflation hedges; in turn, the real size of the financial system declined. The corresponding increase in the velocity of money caused prices to rise even faster than the increase in the monetary aggregates. 1.7 From July 1, 1980 onward, the Government took steps to reverse the financial decline. A major decision was to liberalize interest rates on time deposits. After that liberalization, the banks decided to determine the rates through a "gentlemen's agreement," according to which the higher rates at first were paid only on time deposits of longer maturity. By July 1981, the rates had been raised to 45 percent on deposits of 3 to 6 months and to 50 percent on deposit of longer maturity. Increased competition from within the banking system, from brokers, and from the higher rates paid on short-term Treasury bonds created strong pressures on the agreement, and in 1982 many banks were unofficially offering rates higher than those agreed (sometimes of the order of 60 percent for 6-month time deposits). 1.8 From 1980 to 1982, the increase in nominal deposit rates, combined with declining inflation, produced high real interest rates. During that period, the real rates paid on 6-month deposits, compounded annually, rose from around -50 percent to about 25 percent. Excluding commercial and public deposits, the average real rate paid to private savers on all deposits rose from -46 percent in 1980 to a positive 18 percent in 1982. It again became attractive to hold financial assets; in real terms the holdings of M2 rose by 63 percent from 1980 to 1982. As a percentage of GNP, M2 rose from a low of 15 percent of GNP in 1980, to 18 percent in 1981 and to 24 percent in 1982. By the end of 1982 the financial system had returned to roughly the size that it had been in the mid-1970s. 1.9 Tables 1.1 and 1.3 reveal more detail about the process. For the assets included in M1, the rates of interest paid were zero on currency and commercial sight deposits and 5 percent on personal sight deposits until the end of 1982. Hence, it was unattractive for savers to hold either currency or sight deposits. In real terms there was no increase in M1 holdings in the 1980-82 period. If commercial sight deposits are excluded from M1, there was an actual decline. These commercial deposits are often not held as a matter of choice since borrowers are usually required by banks to hold compensating balances against loans outstanding. By 1982 the quantity of sight deposits had fallen in real terms by 40 percent from the peak in 1977; exclusive of commercial sight deposits, household holdings of sight deposits have fallen more than 60 percent. To the public it was clearly unattractive  to hold sight deposits paying only 5 percent interest during a period in which inflation was close to 30 percent and savings deposits paid 50 percent. All of the growth in M2 was in the holdings of time and savings deposits, which in real terms increased fourfoLd in two years. 1.10 The liberalization and subsequent increase in interest rates played a major role in the stabilization of the economy. Higher real interest rates increased the demand for deposits, which in turn made it possible for banks to extend credits beyond what otherwise would have been possible, given the targets for inflation and the balance of payments. However, the increase in the interest rates to very highi real levels also gave rise to difficult problems. The collapse in mid'-1982 of a major nonbank financial intermediary had an adverse effect on confidence. Lower confidence increased the upward pressure on interest rates and led to an outflow of deposits from the weaker banks to the stronger banks, whiich proved unwilling to respond by reducing their deposit rates, and thus intensified the competition for deposits. Despite the increase in deposit rates and the consequent increase in credit rates effectively charged to b,orrowers, demand for credit continued to be strong, mainly because firms had to refinance a large part of their interest payments (see Chapter 5). The need to roll over credits and to refinance interest created liquidity difficulties for the banks and intensified their competition for deposits. This sequence of events explains why interest rates on time deposits remained high in real terms and often exceeded the limits of the gentlemen's agreement among the banks. 1.11 Faced with the problems of excessively high real interest rates, the Government decided in late 1982 to reform interest rate policy. As of January 1, 1983, the leading banks decided to increase the interest rate paid on sight savings deposits from 5 percent to 20 percent, while reducing the rate on 1-year time deposits from 50 percent to 45 percent and on 6-month deposits from 50 percent to 40 percent. These rates became binding on the nine largest banks, but smaller banks were allowed to pay a premium above these rates of from 2.25 percentage points to 2.5 percentage points. At the same time the withholding tax on deposits whose owner is identified was lowered from 25 percent to 20 percent (while it was increased from 25 to 30 percent in bearer certificates of deposit). The after-tax yield of deposits whose owner is identified has therefore been reduced from 37.5 percent to 36 percent on 1-year deposits and to 32 percent for 6-month deposits. There was thus a substantial narrowing in the rate differentials between sight and time deposits. 1/ As a result, the proportion of asset holdings constituted by sight deposits started to increase significantly in the first half of 1983. 1.12 At the end of 1982, excluding commercial sight deposits, the weighted average before-tax interest raLte paid on sight plus savings deposits was around 46 percent, the after-tax rate 34.5 percent. The reforms instituted in January 1, 1983 reduced the average rate on sight and time savings deposits to around 40 percent before taxes; and 32 percent after taxes. Hence, apart from slowing down the shift out of sight and into time deposits, it is unlikely that the new interest rate policy will have much impact on the demand for financial assets. Since inflgation is continuing to decline, the gradual increase in average real interest rates for deposits will maintain the growth of the demand for M2. 17 These rate differentials rere further narrowed in June 1983 when interest rates on time deposits were reduced by 5 percentage points.  -6- C. Increasing the Size of the Financial System 1.13 It was shown above that, despite the recovery in the last two years, the financial system in Turkey remains smaller than in most other countries with comparable per capita income and development. This suggests that there is potential to increase the size of the Turkish system. Furthermore, during 1982 there was a sharp liquidity squeeze in the corporate sector (examined in more detail in Chapter 5). It would also appear that, because of the adverse impact on corporate debtors of prior devaluations of the Turkish lira, there was a strong (though by 1982 lperhaps ill-founded) preference for domestic credits. It appears from these aspects that further financial deepening is desirable. 1.14 The most frequently suggested tool for increasing the mobilization of savings through the financial system is the payment of higher real interest rates. The liberalization leading to positive real interest rates proved that the demand for domestic financial assets in Turkey is very sensitive to yield. Given that before-tax rates on time deposits reached about 25 percent in real terms in the last quarter of 1982,, however, there was no scope for raising rates on time deposits further. On the contrary, the authorities and the leading banks were forced to take measures to reduce these rates, as mentioned above, although they have limited the impact of those measures on the after-tax yield to depositors by simultaneously reducing the witholding tax on interest income. In the long run, in which equilibrium can be reestablished in Turkey's financial markets, before-tax real interest rates on time deposits should not be more than 5 percent. 1.15 The inflation rate has declined quite rapidly in Turkey, from more than 100 percent in 1980 to 25 percent in 1982. There is therefore scope for new reductions of the ceilings on interest rates for time deposits. But the effectiveness of decisions of this type muwst be carefully assessed. In the last quarter of 1982, the previous higher ceilings for the interest rates on time deposits were often exceeded, because of the strong competition among the banks to attract new deposits with which to ease their serious liquidity problems. However, in view of the measures taken by the authorities, the compliance with the ceilings became more effective in the first half of 1983. In view of the decline in inflation, there seems to be no great danger that additional reductions of interest rates, if moderate and gradual, would have a considerably negative impact on the total demand for financial assets. But such reductions will be justified only if there are good reasons to expect that they will be reflected in reductions of about the same size in the lending rates. If the liquidity situation remains very tight, there is the danger that the interest rate ceilings will be circumvented by the practice of blocking a certain percentage of the amounts of the loans in commercial deposits. Moreover, in order to avoid risks of a substantial slow down in the growth of real M2, or even negative growth of that aggregate, the after-tax real interest rates for sight savings deposits should not be allowed to become lower than they are at present, and those for time deposits should not become lower than 5 percent after the withholding tax in the next few years.  1.16 Under conditions of rapidly changing prices, people's expectations often lag. It would appear that this may be true of depositors in Turkey, and that the real interest rate on deposits may not have been perceived to be as high ex-ante as it proved to be ex-post. Furthermore, although there is an appreciation among the public that inflation rates are declining, there is likely to be considerable uncertainty about the actual rates. This uncertainty may lead depositors to seek a premium to cover risk. The lagging expectations and the risk premium, together with the sharp competition for deposits among banks, may in principle be one of the reasons why interest rates on deposits appear to be so high in real terms. If this reason proved to be a strong one, it might be worthwhile to consider the possibility of allowing banks to offer savings deposits at interest rates indexed to the rate of inflation. As these deposits would both eliminate the need to predict: the rate of inflation and the risk of mistakes, the public might be willing to hold the same quantity of deposits at a lower real interest rate. However, if such deposits were offered, they would have to be matched with indexed loans, so that financial institutions would not be exposed to the risk of unexpected movements in prices. In addition, the indexation of interest rates on deposits might become a factor of rigidity creating difficulties for the reduction of inflation. Before any initiative is taken in this area it will be preferable to wait and confirm if, as seems to be the case, the pressures of bank competition and the behavior of savers are the more important obstacles to reducing the very high real interest rates that recently have been paid on time deposits. 1.17 The impact of the withholding tax on the demand for financial assets must also be considered. As mentioned above the rate of that tax was recently reduced from 25 percent to 20 percent on deposits whose owner is identified. The tax is being levied, however, not only on the real yield of time deposits but also on the part of the interest rate that represents the compensation for inflation. Thus, if the expected inflation rate were 20 percent, the effect of the 20 percent withholding tax on the interest rate of 40 percent paid to a 6-month time deposit would be to reduce the real annualised yield of the deposit from 20 percent to 12.1 percent. The incidence of the tax on thse real interest income would thus be about 40 percent. On the basis of this example it becomes apparent that, in principle, the withholding tax should be levied only on the real portion of interest revenues. But if inflation continues to decline rapidly in the coming years, and if nominal interest rates for time deposits decline even more rapidly, as would be desirable, this problem iwill become less important. Under such conditions, and in the face of the administrative difficulties of introducing a system of taxing only the real interest income, there may be some justification for maintaining the present system. However, as long as real interest rates and the inflation rate remain at high levels, the authorities should try to reduce further the rate of the withholding tax to 15 percent. The budgetary difficulties involved in such a reduction cannot be minimized. It is roughly estimated that the withholding tax on the interest from bank deposits corresponded in 1982 to about 5 percent of total tax revenues. An additional reduction of the withholding tax would deprive the Government of much needed revenue. It should, however, be mentioned that the very high expected rate of growth of time deposits gives the Treasury the possibility of avoiding big declines in the value of the revenues collected from the t:ransactions tax even if the tax rates are  -8- gradually reduced. Moreover, it should be stressed that the taxation of interest earned on time deposits, with its negative incidence on savings and the cost of credit, is less efficient from the point of view of the objective of economic development than other types of fiscal revenues which could be raised. 1.18 Another technique for increasing the mobilization of funds that has been attempted in other countries is to open new branches of banks to increase accessibility to depositors. This is not a feasible option in Turkey, however, where 6,235 branches are already in operation. As noted in Chapter 6, excessive branching and overmanning has raised administrative costs in Turkish banking to two to three times the level in other European countries, and a concerted effort must be made to reduce these costs. 1.19 Expansion of the Turkish banking system in nominal terms is now limited by the Government's policy to control inflation by restricting the growth of monetary and credit aggregates. Policy changes that make the holding of M2 more attractive must, however, be reflected in the nominal growth of the banking aggregates. The greater the demand for M2 as a percentage of GNP, the more rapid can be the expansion in domestic credit given the objectives set for inflation and for the level of net foreign assets held by the banking system. Thus, if the Government pursues policies to increase the public's desire to hold larger M2 balances, it must also allow the supply of money and credit to expand in line with the increased demand so that downward pressure will not be exerted on output. 1.20 The growth of Turkey's financial system should also be pursued by developing the activity of nonbank financial institutions. A special emphasis must be given to strengthening and developing the domestic capital market. Specific recommendations in this regard are made in Chapter 7.  - 9 - CHAPTER II - THE GROWTH AND ALLOCATION OF CREDITS FROM THE BANKING SYSTEM A. Credit Expansion and Monetary Policy 2.1 The fall in the demand for money in the second half of the 1970s was associated with a significant decline in real terms of bank credits from 19,77 to 1980. The decline was particularly noticeable in credit to the private sector, as shown in Table 2.1. Also, the share of total deposit money bank credit based on own resources fell sharply, from 82 percent in 1975 to 70 percent in 1980. This decline was partly related to the increase in reserve requirements imposed in 1977, but it also reflected difficulties in attracting deposits. These difficulties, together with the increased financing needs of the public sector, which is financed mainly by the Central Bank, also meant that the Central Bank assumed a much greater role in total credit expansion. Whereas the Central Bank share in the credit stock (in both direct credits and rediscounted credits) had been less than 30 percent in the beginning of the 1970s, by 1980 it had increased to almost 50 percent (Tab]Le 4 in the Statistical Annex). 2.2 Credit from the banking system continued to decline in real terms ifor 1980 as a whole, but there was a positive real growth in the second half of the year. During 1981, credits rose by nearly 15 percent in real terms, andl further real increases were recorded in 1982. The Central Bank's share in total credit started to decline in 1981 as the public sector demand for credit became subject to tighter controls. This decline was partly offset by an increase in Central Bank rediscounts of export credits. During 1982 there was a sharp reduction in the Central Bank's share in total credit as the demand from the public sector continued to decline, and the share of export credits rediscounted by the Central Bank fell markedly. Table 2.1: DIRECT CREDITS FROM THE FINANCIAL SYSTEM (Year end values: Billion TL) 1975 1976 1977 1978 1979 1980 1981 1982 Current Prices Total credits 219.2 317.1 441.1 562.7 844.8 1414.3 2208.6 2908.5 Credits to the public administration 27.0 38.6 70.9 90.6 135.0 259.0 390.7 441.2 Credits to SEEs 65.2 109.8 154.8 196.0 301.6 444.5 566.9 638.0 Credits to the private sector 127.0 168.7 215.4 276.1 408.2 710.8 1251.0 1829.3 Prices of 1975 Total credits 219.2 266.2 272.1 233.3 193.1 166.1 206.5 217.8 Credits to the public administration 27.0 32.4 43.7 37.5 30.9 30.4 36.5 33.0 Credits to SEEs 65.2 92.2 95.4 81.3 68.9 52.2 53.0 47.8 Credits to the private sector 127.0 141.6 132.9 114.5 93.3 83.5 117.0 137.0  - 10 - 2.3 The growth of bank credits since 1980 has been to a large extent determined by the ceilings on the expansion of domestic financial assets of the Central Bank, which were established in the context of the economic stabilization policy measures agreed with the International Monetary Fund. One of the main functions of these credit ceilings has been to ration the credit extended by the Central Bank to the public sector and the amount of preferential credit to the private sector financed by the Central Bank. In 1981 and 1982 total credit expanded by 56 and 32 percent respectively, significantly exceeding the rates of increase of wholesale prices (about 25 percent from the beginning to the end of the year, both in 1981 and in 1982). Despite this, and despite the steady increase in the share of total credits allocated to the private sector (see paragraph 2.6), banks and industrial corporations were facing difficult liquidity problems by the end of 1982 because of the high level of interest rates (as mentioned below in paragraph 2.8 and explained in more detail in Chapter 5). 2.4 Total credit from the banking system has been determined partly by the limits on the expansion of domestic financed assets of the Central Bank and partly by the money multiplier. Because of the different rates for reserve requirements for sight and time deposits and for funds used for preferential credits, and as a consequence of shifts in deposits and the distribution of credits, the multiplier has been very unstable up to the end of 1982. In addition, there has been widespread nonobservance by banks of their reserve requirement obligations. During the period of the stabilization program the general tendency has been for the multiplier to rise rapidly. Therefore, although the limits on credit from the Central Bank have generally been observed, money and total credit from the banking system have increased more than was planned. This increase may have slowed down the reduction of the inflation rate, but it was to a large extent offset by the increase of money demand at a faster rate than was forecast when the targets were set. A major problem in the implementation of monetary policy during the period of the stabilization program has thus been the insufficiency of policy instruments and the instability of the reserve money multiplier. An important step to stabilize the multiplier was taken with the decision to reorganize the system governing the cash reserve and liquidity requirements of the banking system from January 1, 1983. From that date, the highly differentiated cash reserve requirements have been replaced by an equivalent single rate of 25 percent. Also, the liquidity ratio--varying from 10 to 15 percent according to the size of the bank and monitored only once a month--has been replaced by a single rate of 10 percent to be monitored on an average daily basis. B. The Allocation of Credits 2.5 Table 2.2 presents the distribution of advances by the financial system to economic sectors. It is clear from the table that the Government budget has taken an increasing share of total credits up to 1980, but has subsequently declined. Its share of total advances increased from 12 percent in 1975 to 18 percent in 1980 but it started to decline in 1981 and it was only 15 percent by the end of 1982. The deteriorating finances of the State Economic Enterprises (SEEs) during the second half of the 1970s resulted in an increasing share of financial resources going to those enterprises, but their credit demands have been brought under much better control since 1980. From a  - 11 - peak of 36 percent of outstanding advances in 1979, the SEE share fell to 26 percent in 1981 and to around 20 percent by the end of 1982. Of the net increase in total credits only 15 percent went to SEEs in 1981 and only 1() percent in 1982. Table 2.2: DISTR:IBUTION OF FINANCIAL SECTOR ADVANCES (percent) Public Public Private Sector Breakdown Year adminis- enter- Private Agricul- Exports Const- Artisans Other tration prise sector ture ruction 1975 12 30 58 16 5 4 2 31 1976 12 35 53 12 4 4 2 31 1977 16 35 49 11 3 4 2 29 1978 16 35 49 9 4 4 2 30 1979 16 36 48 10 4 3 3 28 1980 18 32 50 10 6 3 3 28 1981 18 26 56 12 8 3 2 31 1982 15 22 63 12 8 3 2 31 Note: Advances include credits, bonds, and other forms of advance. 2.6 For the private sector, the percentage of funds available declinied from 58 percent in 1975 to only 48 percent in 1979. However, with the strengthening of the public finances, particularly the SEEs, the private sector share rebounded to 56 percent in 1981 and to 63 percent in 1982. Of net new credits in 1981, 68 percent went to the private sector. In 1982, that percentage rose to 82 percent. 2.7 Within the private sector, the Government has special credit programs for agriculture, exports trade, housing, artisans and small firms, and so on. Some of the credits listed in Table 2.2 under "Other" are also special credits included in one or another of the selective credit programs. Because of the ever-increasing credit demands of public administration in the late 1970s, less credit was available for the private sector, and the greatest restrictions fell on industry and agriculture. However, the sharp cut in the share of credit to the public enterprises in 1980 and 1981 made possible an increase in the share of credit to the private sector, particularly in general credits falling outside the Government credit schemes. 2.8 Although nominal credits were increasing rapidly during the entire period, credit in real terms was declining from 1977 through 1980. During those years the real value of advances to the private sector fell by more than one-third, from TL147 billion in 1975 prices in 1977 to TL94 billion in 1980. In nominal terms the rates of increase were 76 percent in 1981 and 46 percent in 1982. These rates correspond to real increases of 40 percent in 1981 and 17 percent in 1982 (using the wholesale price index as a deflator). In the second half of 1982, however, most of the real credit expansion was absorbed  - 12 - by the need to refinance part of the high real interest rates charged to private enterprises. This explains why during that period firms began to experience serious liquidity difficulties despite the increasing availability of credit in real terms. C. Finance of the Deficit of the Consolidated Budget of the Central Government 2.9 Central Government finances deteriorated significantly in the 1970s because the increase in expenditures, from 22.5 percent of GNP in the early 1970s to 27.0 percent in 1979, was not matched by a similar increase in the tax ratio (Table 6 in Statistical Annex). On balance, that ratio rose by only 2 percentage points. Correspondingly, the budget deficit widened from 1.5 percent of GNP in the early 1970s to about 4 percent in 1979. The main reason for the growth in expenditures was the rise in transfer payments, particularly to the SEEs. The relatively low elasticity of revenue with respect to GNP in this period of accelerating inflation was mainly the result of a loss of progressivity of the income tax, increased tax evasion, and a comparatively heavy reliance on individual taxes of a specific character. In 1980, despite a decline in the share of expenditure in GNP, the budget deficit increased further to more than 4.5 percent of GNP as the tax ratio continued to fall. In 1981, as a result of the tax reform, the ratio of revenue to GNP increased by about 2 percentage points, whereas expenditures, as a result of better expenditure control (particularly in transfers to SEEs), fell by 1.5 percent of GNP. 2.10 In fiscal year 1982, revenue fell considerably short of expectations primarily because of substantial deferrals of tax payments by the corporate sector, and revenue fell in relation to GNP. The authorities, however, responded to the shortfall by cancelling some planned expenditures, and expenditures were for the first time in a decade held below the original appropriations; in relation to GNP, expenditures fell to 21.5 percent. The budget balance remained unchanged, at 1.5 percent of GNP. The budget for 1983 implies a slight reduction in the budget deficit. 2.11 During the late 1970s a large part of the increasing deficit of the consolidated budget was financed by direct borrowing from the Central Bank. Such borrowing contributed substantially to monetary growth. In consequence of the budgetary developments mentioned above, the financing requirements of the public administration have been significantly reduced since 1980, as already mentioned in paragraph 2.5. In addition, the Central Bank's share in the financing of the public sector has been sharply curtailed. A more active sale of Treasury bills and bonds as a means of financing the Government deficit helped to bring about this reduction. 2.12 There has, therefore, been substantial progress in reducing the proportion of total credits from the banking system absorbed by the Treasury. Gross credits to the public administration absorbed only 7 percent of the growth of total domestic credits during 1982. Despite this, the Government could contribute substantially to improving the financial situation of the productive sector if it decreased the borrowing requirements from the banking system still further, and if it became, in marginal terms, a net supplier of funds to that system (which would imply a decrease in net nominal debts of the public administration to the banking system). Today in Turkey one of the most pressing issues is to reduce the excessively high real interest rates on  - 13 - nonpreferential credits to the productive sector. Several measures recommended in this report can contribute to that objective. But the most effective measure to reduce interest rates would be to increase the quantity of funds available for nonselective credits. Because a more rapid expansion of the supply of total domestic credit might negatively affect the economiic stabilization policy, more funds can be directed to general credits only by increasing the fraction of the total going to these credits. Such an increase would be easier if the Government reduced its indebtedness to the financial system. At the end of 1982, public administration absorbed about 15 percent of the total advances of the financial sector, as against 12 percent in the mid-1970s. Even the rechanneling of a few percentage points of total advances--from credits to the public administration into general credits to the productive sector--would help to reduce the very high real interest rates in that submarket. Rechanneling of advances can be achieved either by further reductions of the government budget deficit or more importantly by financing a substantial proportion of that deficit through the issue of bonds placed in the market. 2.13 The interest rates of the bills and bonds issued by the Treasury, in recent years have not been sufficiently attractive to productive savers. In consequence, the public debt securities have been subscribed mainly by banks that have needed them to fulfill their liquidity requirements. If the Treasury bills and bonds were offered at competitive interest rates and with maturities sufficiently adapted to demand, they might absorb some funds available for the productive sector, but they would also certainly have a positive effect on productive financial savings. They would therefore have a positive net contribution to the enlargement of the availability of finance to the productive sector. Moreover, the issue of Treasury bills at competitive interest rates would contribute to conducting open market operations and thereby influence the interest rate determination in a market-oriented manner. In addition, it would contribute to revitalizing the capital market as explained in Chapter 7. Treasury bills and bonds should be marketed more actively. Instead of being sold to the public only by the Central Bank and by the Agricultural Bank, government securities should also be distributed by bond dealers and by commercial banks. D. Finance of SEEs 2.14 An important element in the deterioration in the Government's accounts during the 1970s was a perceptible weakening in the financial position of the SEEs (see Table 8 in Statistical Annex). In the first half of the 1970s, the SEEs had small operational surpluses, but by 1978-79 deficits of the order of 4 percent of GNP were recorded. So-called duty losses--i.e., losses resulting from the imposition, by the Government, of price controls on the enterprises that must be reimbursed through budget transfers--rose particularly rapidly in the latter half of the 1970s. The steady growth in current transfers and direct borrowing from the Central Bank to finance a rapidly growing investment program contributed substantially to monetary growth. 2.15 With the burden of SEEs' financing on the budget approaching critical proportions, large corrective increases in SEE prices were announced in January 1980, and a restructuring of operational policies was initiated. The 1980 measures included a limitation of subsidies for duty losses to a list of  - 14 - a few basic items (coal, chemical fertilizers, electricity for metallurgy, services of the State Railways, and maritime transport) and large immediate price increases for many SEE goods and services. Further price increases were made later in the year. These measures caused a decline in the operational deficit of SEEs to 0.5 percent of GNP in 1980. A continuation of the pricing policies in 1981 and 1982, together with measures to reduce the problem of overstaffing, resulted in a further improvement in the operational position, and small surpluses were recorded in these two years. The new policies were formalized in a program established in 1982. First, it was decided to give priority to operational results; second, managers in certain enterprises were given greater autonomy to manage and to adjust prices; third, transfers to SEEs were limited and, in the case of nonpayment by SEEs of taxes, social security contributions, or amortization of foreign debt, such arrears were to be deducted before payments were made from the budget; finally, borrowing from the Central Bank was limited to two SEEs, the Soil Products Office, and the State Monopolies. 2.16 As mentioned, one of the main problems for the SEE sector was the ambitious investment program put up over the years. Total investment (including stocks) as a share of GNP rose from around 7 percent in the mid-1970s to more than 10 percent in 1980. Since then, the share has declined significantly, to around 7 percent of GNP in 1982. The decline was particularly sharp in 1982 after the introduction of the new procedures, according to which priority is given to operational results and the investment program is essentially a cash limit. In this new system, greater access to the Central Bank or to budget transfers higher than planned will not be permitted to compensate for shortfalls in self-financing or in financing from other sources. 2.17 Although there has been some improvement in the financial results of the SEEs since 1980, some important structural problems still remain before the economic efficiency of the SEEs can be strengthened. The monetary statistics show not only the reduction of the share of SEEs in total credits from the banking system but also the decline in the reliance on the Central Bank as a source of finance for SEEs. (See Table 8 in the Statistical Annex.) Since 1980, only the state monopolies, the Soil Products Office, and meat and fish industries increased their borrowing from the Central Bank; since 1981 the meat and fish industries have been excluded from access to the Central Bank. Credits from deposit money banks have risen only moderately in the period to 1981, although there were significant increases in 1979 and 1980. Credits from investment and development banks were an important source of finance, particularly in 1981. 2.18 An improvement in the system of financing the SEEs that is still required is the elimination of the special interest rate subsidies those enterprises are receiving. The Interest Rate Rebate Fund is granting a subsidy of 5 percent to SEEs on all the loans that they get from the banking sector and a subsidy of 1 percent to the banks that make such loans. Also, the State Investment Bank grants medium and long-term credits to the SEEs at highly subsidized rates. These interest rate subsidies to SEEs impose a burden on the Treasury and on nonpreferential credits to the private sector. At the same time, they distort the social profitability calculations of those enterprises. For these reasons, the special interest rate subsidies to SEEs  - 15 - should be eliminated. The SEEs should pay the same interest rates as the private enterprises. This means that they should have the same conditions of access to preferential credits. 2.19 Finally, the SEEs with a more solid financial situation and with better economic prospects should attempt to become more active in the bond market. They should try to cover part of their financial needs by issuing bonds with competitive terms, both as regards interest rates and maturities. This would alleviate their need for financial support from the Government budget, would reduce their absorption of scarce bank credits, and would invigorate the capital market.   - 16 - CHAPTER III - THE COST OF CREDIT A. The High Level of the Real. Costs of Credits 3.1 Up to June 1980, when the interest rates were regulated by the Government, the control on the lending rates effectively charged to borrowers could not always be very strict, because of the practice of commercial banks requiring that a certain percentage of the loan proceeds be blocked for compensating deposits at zero interest. Despite this practice, the lending interest rates, even in nonpreferential bank credits, largely remained negative in real terms, particularly when inflation accelerated sharply during 1977-80. 3.2 As mentioned earlier, the liberalization of interest rates in July 1980 was followed by a gentlemen's agreement among banks that fixed an upper limit to interest rates for nonpreferential credits. In addition, .a great variety of preferential credits remained under the control of public banks or continued to be influ,enced by the selective credit policies of the Central Bank. 3.3 Table 3.1 shows the great variety of lending rates applied in the second half of 1982. The information of the table is, however, incomplete because account is not taken cf the effect of loan proceeds blocked in commercial bank deposits. The percentages of proceeds so blocked, mainly required by commercial banks for short-term credits, vary widely from bank to bank, from client to client, and according to the fluctuations in the liquidity position of each particular bank. The percentages normally fall between 20 and 30 percent but they are sometimes higher. As a consequence, by the end of 1982 the cost of short-term nonpreferential credits was usually of the order of 60 to 70 percent, but in some cases it could rise to 80 percent. Similarily, the cost of short-term export credits for industrial products tended to be around 35 to 40 percent, rather than 29 percent as shown in Table 3.1. 3.4 Given this situation and the lack of correspondence between the statistics of credit allocation and the schedule of interest rates, it is practically impossible to present a distribution of credits according to-the interest rate levels and to calculate accurately an average of the interest cost paid by the borrowers. On the basis of very rough calculations, however, it is possible to conclude that the average interest costs paid by nonagricultural enterprises, excluding State Economic Enterprises (SEEs) and artisans and small traders, was of the order of 50 percent at the end of 1982. The same calculations suggest that it is likely that, for about half of the debts of the same enterprises, the interest costs were of the order of 60 percent or more. The inflation rate from the beginning to the end of 1982 was 25 percent, as measured by the wholesale price index. Thus it appears that in 1982 the average cost of credits to nonagricultural enterprises in real terms was of the order oi- 20 percent and may have exceeded 30 percent in many cases. 3.5 The negative impact of the high real interest costs on the after-tax profit of enterprises will be partially offset by the rules for the taxation of profits. For corporations whose profits are taxed at a rate of 40 percent, an interest rate of 70 percent: will correspond to an after-tax cost of  -17- Table 3.1: COST OF CREDITS TO THE BORROWER IN THE FIRST QUARTER OF 1983 (percent per annum) Interest Contribution Subsidity from Transaction Total Item, Rate Commission to the IRRF the IRRF tax cost Short-Term Credits Credits from the Central Bank to TMO for agricultural support purchases 10.0 - 1.0 - - 11.0 Agricultural credits 20.0 - - 20.0 Artisans and small traders 20.0 - - - - 20.0 Export Credits Industrial products 31.5 2.0 0 4.7 - 28.8 Other credits 31.5 2.0 1.6 0 - 35.1 General credits 36.0 2.0 3.6 - 5.7 47.3 Commercial bank credits to SEEs 36.0 2.0 0 4.0 5.7 42.3 Medium and Long-Term Credits Investment crediLs to SEEs, by the State Investment Bank 21.5 - - - - 21.5 Agricultural credits from the Agricultural Bank 22.0 - - - - 22.0 Artisans & small traders credit from the Halk Bank 22.0 - - - - 22.0 liousing credits from the housing bank construction savings system and social housing credit 16.0 - - - 2.4 18.4 Housing credit scheme & cooperative Society credit 22.0 - - - 3.3 25.3 Commercial housing 35.5 - 3.6 - 5.3 44.4 Rediscountable investment credits from commercial & development banks Export oriented in Underdeveloped regions 29.0 2.0 0 10.1 - 20.8 In other regions 29.0 2.0 0 5.8 - 25.2 Other loans in underdeveloped regions 38.0 2.0 0 11.4 - 28.6 Kediscountable loans for manufactures of investment goods 31.5 2.0 0 9.5 5.0 28.5 Rediscountable working capital loans In underdeveloped regions 38.0 2.0 0 5.7 6.0 40.3 In other regions 38.0 2.0 0 0 6.0 46.0 Other Medium and Long-Term Loans Investment Loans Export oriented In underdeveloped regions 45.0 - 0 15.75 - 29.25 In other regions 45.0 - 0 9.0 - 36.0 Other In underdeveioped regions 45.0 - 0 13.5 - 31.5 In other regions 45.0 - 0 6.75 - 38.25 Working Capital Loans In underdeveloped regions 45.0 - 0 6.75 6.75 45.0 In other regions 45.0 - 0 0 6.75 51.75 Note: The effects of blocking of a proportion of the amounts of certain loans in compensating bank deposits at zero interest is not taken into account. Several charges paid by the borrowers as stamp duty, communications, expenses, and the like are not included. Those charges tend to be of the order of 0.5X to 1X.  - 18 - 42 percent. In a situation in which the inflation rate is 25 percent, the real after-tax cost to enterprises of a 70 percent interest rate is thus of 14 percent instead of 36 percent before tax. The difference reflects the losses of the Treasury revenues from the taxation of profits--losses that occur because, under existing rules, the inflationary component of the interest rates is considered as a cost in the determination of taxable profits. But even with the tax effects taken into account, it will not be easy for many Turkish companies to get returns on their total assets high enough to enable them to bear after-tax interest costs in the range of 10 to 15 percent. Moreover, this argument is not valid for enterprises that havle losses or that benefit from tax reductions, (although the latter often receive preferential credits). In those cases, the tax mechanism does not provide any relief from the burden of high real interest rates, and the after-tax real cost of a 70 percent interest rate under 25 percent inflation will actualLy be 36 percent. B. The Need to Reduce the Real Cost of Credits 3.6 The description in the previous section shows that the cost of credits to nonagricultural private enterprises has risen in real terms to levels that cannot be sustained indefinitely. Few firms in Turkey can earn enough on operations to pay such high real rates on any substantial fraction of their debt. Until now the increase in real interest rates has to a large extent been offset by the decline in real wages, but this is a solution tlhiat cannot be maintained in the long run. Besides, many firms under the pressure of high interest rates are increasing their equity instead of increasing their debts. However, many firms will be forced to borrow from banks in order to finance part of their interest: payments. For such firms, there is the danger that the debt-equity ratio will rise progressively and that over an extended period they can be lead into bankruptcy. Hence, if the lending interest rates to firms are not substantially reduced, the present problem--which for many firms is already one of illiquidity (see Chapter 5)--may become one of insolvency. 3.7 To bring down the lending interest rates it will first be necessary to reduce the cost of the resources used in bank credits, the taxation of those credits and the costs oi- bank intermediation. The following sections analyze, therefore, the possibilities of reducing: (i) the average cost of bank deposits; (ii) the costs imposed on the banks by the liquidity and cash reserve requirements; (iii) the contributions to the Interest Rate Rebate Fund (IRRF); (iv) the transactions tax; and (v) the costs of bank intermediation. In addition, it will be necessary to look at the relationship between the supply and the demand for credits. The final section of the chapter presents some comments on that relationship. C. Reduction of the Cost of Bank Deposits 3.8 As explained in Chapter 1, the announced interest rates for time deposits rose substantially since their liberalization in June 1980. Taking into account the pattern of maturities, and compounding annually the interest rates for deposits of less than one year, it can be roughly estimated thaet the cost of time deposits to the banks has increased from about 15 percent in 1978 to approximately 55 percent in 1982. Actually, the average in 1982 may have been slightly higher because, as mentioned previously, many banks, under the pressure of competition, were paying to depositors more than the announced interest rates.  - 19 - 3.9 At the same time, the proportion of time deposits in total deposits has been increasing rapidly. Given the wide and growing differentials between the interest rates paid on time and on sight deposits, that proportion has grown from 17 percent in 1978 to about 55 percent by the end of 1982. The percentages would be even higher without the sight deposits kept by the public sector in specialized banks, particularly in the Agricultural Bank, and without the commercial deposits, which comprise about one-fourth of total deposits and are largely explained by the practice, mentioned above, of banks blocking substantial percentages of the proceeds of their credits. 3.10 The trend toward a higher share of time deposits, together with the increase in their interest rates, has resulted in a rapid increase of the average cost of total deposits. On the basis of the announced interest rates, that cost rose from about 4 percent at the end of 1978 to about 30 percent at the end of 1982. The marginal costs were even higher. Time deposits accounted for about two-thirds of the increase in total deposits of all deposit money banks from the beginning to the end of 1982. The cost of the additional deposits collected by the banks in that year was thus, on average, more than 35 percent on the basis of the announced interest rates. 3.11 Another factor contributing to the rising costs of bank credits, although less important, was the decline of the proportion of rediscounts and advances from the Central Bank in total liabilities (from 27 percent in 1978 to 15.5 percent in 1981) and the increase in the interest rates of such rediscounts and advances. Because the cost of the funds provided by the Central Bank has been lower than that of time deposits, they contributed to the increase of the marginal cost of the resources used in bank credits. 3.12 Recognizing the serious negative impact of the high costs of time deposits on the level of lending interest rates, the Government and the leading banks, as of January 1, 1983, introduced lower limits to the interest rates of time deposits and raised the rates on sight savings deposits (as explained in Chapter 1). As shown by the comparison of Table 3.2 and Table 3.3, it is expected that these measures will initially reduce the average cost of deposits to the banks by about 5 percent. Moreover, the new structure of interest rates will certainly slow down the shift of the demand for financial assets away from sight saving deposits into time deposits and thus will lead in the future to a lower average cost of deposits than would otherwise be the case. 3.13 In 1983 the Turkish authorities have projected that the inflation rate would be reduced to about 20 percent.!! With such a reduction of inflation, the real average cost of deposits, even after the reductions recently introduced, would still remain too high. Further reductions of that cost are thus called for. The authorities and the lending banks should take new initiatives to bring down the ceilings on the interest rates for time deposits, but they should do so only after it becomes clear that the compliance with the existing limits is satisfactory. Otherwise, there would be the risk that such official intervention would not produce effective results and would create significant distortions in the financial market. 1/ Inflation information to date indicates that this target would be exceeded by 5 to 10 percentage points.  -20- Table 3.2: THE INTEREST COST OF RESOURCES FROM DEPOSITS AVAILABLE FOR LENDING BY COMMERCIAL BANKS IN 1982 Resources Used in Resources Used in Nonpreferential Credits Export Credits Average of all Deposits and all Public and 6-month Public & 6-month Credits of Commerical Savings Time Commercial Time Commercial Deposits Deposits Deposits Deposits Deposits Banks 1. Interest paid 0.0 5.0 56.2 / 0.0 56.2 /a 30.9 a 2. Liquidity requirement 0.15 0.15 0.15 0.15 0.15 0.15 3. Reserve requirement 0.35 0.35 0.30 0.10 0.05 0.19 /c 4. Average yield of the liquidity requirement /d 20.0 20.0 20.0 20.0 20.0 20.0 5. Yield of the reserve requirement 12.6 12.6 19.6 12.6 19.6 16.4 /c 6. Earnings on the liquidity and reserve requirements /e 7.4 7.4 8.9 4.3 4.0 6.1 7. Total cost of funds available for lending Jf -7.4 -2.4 47.3 -4.3 52.2 24.8 8. Proportion of the funds from deposits available for lending /g 0.5 0.5 0.55 0.75 0.80 0.66 9. Average cost of the funds from deposits available for lending /b -14.8 -4.8 86.0 -5.7 65.3 33.3 /a Interest on 6-month time deposits compounded annually. /b Average based on the composition of total deposits in the last quarter of 1982: sight commercial and public deposit: 34%; sight savings deposits: 12%; time depositE: 54%. /c Average which takes into account the differences of the reserve ratios on time and on sight deposits and on funds used in preferential credits. /d Based on the assumption that the liquidity requirement is composed of 30% cash and deposits in the Central Bank, 20%. Treasury bonds with an interest rate of 20% andl 50% of Treasury bills with an interest rate of 32%. /e (Row (2) x row (4)) + (row (3) x row (5)). jf Row (1) - row (6). ji Row (9) - row (2) row (3). /h Ratio between row (7) and row (8).  -21- Table 3.3; INTEREST COST OF RESOURCES FROM DEPOSITS AVAILABLE FOR LENDING BY THE COMMERCIAL BANKS AFTER JANUARY 1, 1983 Resources used in All Credits Public and 6-month Average of All Commercial Savings Time Deposits and Deposits Deposits Deposit All Credits 1. Interest paid 0.0 20.0 44.0 /a 26.2 /b 2. Liquidity requirement 0.10 0.10 0.10 0.10 3. Reserve requirement 0.25 0.25 0.25 0.25 4. Average yield on the liquidity requirement /c 15.0 15.0 15.0 15.0 5. Yield on the reserve requirement 19.6 19.6 19.6 19.6 6. Earnings on the liquidity and reserve requirements /d 6.4 6.4 6.4 6.4 7. Total cost of funds available for lending /e -6.4 13.6 37.6 19.7 8. Proportion of the funds from deposits available for lending /f 0.65 0.65 0.65 0.65 9. Average cost of the funds from deposits available for lending /g -9.8 20.9 57.8 30.4 /a Interest rate on 6-month time deposits compounded annually. The interest rate of 45% on one-year time deposits is higher by only 1 percentage point. /b Average based on the composition of total deposits in the last quarter of 1982. /c Based on the assumption that the liquidity requirement is composed of 45% cash and deposits in the Central Bank, 20% Treasury bonds with an interest rate of 20% and 35% Treasury bills with an interest rate of 32%. /d (Row (2) x row (4)) + (row (3) + row (5)). /e Row (1) - row (6). /f Row (9) - row (2) - row (3) E Ratio between row (7) and row (8)  - 22 - D. Reducing the Burden Imposed by the Liquidity and Reserve Requirements 3.14 As already mentioned, the banks are obliged to place 10 percent of their total deposits, excluding interbank deposits, in a liquidity reserve. Before January 1, 1983, that ratio was 15 percent--except for very small banks, for which it was only 10 percent or 12 percent, depending on the size of their deposits. The liquiclity reserve can consist of cash, deposits in the Central Bank, and Treasury bills and bonds. On the basis of the recent experience regarding the proportions of the liquidity requirements held as Treasury bills and bonds and their respective interest rates, it can be roughly estimated that the average yield on the liquidity requirement was about 20 percent in 1982 and will be approximately 15 percent in 1983. 3.15 To mobilize resource; into the Central Bank for subsidizing the rediscounts of preferential credits and for providing low-cost loans to the Treasury, the authorities also impose a cash reserve requirement on banks' deposits. Until the end of 1982, the ratio for required cash reserves was 35 percent for sight deposits and 30 percent for time deposits. These ratios were reduced to much lower levels, however, when the funds from the deposits were used by the banks to finance preferential credits. For instance, the ratios were only 5 percent and 10 percent, respectively, on sight and time deposits used to finance export credits of industrial goods. As mentioned in Chapter 2, after January 1, 1983, all the different reserve requirement ratios were unified into a single ratio of 25 percent, which corresponds approximately to the total average of the cash reserve requirements previously held by deposit-money banks. In 1982 the required reserves from sight and time deposits earned interest rates of 1 percent and 1.5 percent a month respectively. Since the beginning of 1983 a single interest rate of 1.5 percent a month has been in force. 3.16 The burden imposed by the liquidity and reserve requirements on the cost of banks' resources available for nonpreferential credits was considerably alleviated as a result of the recent decisions of the authorities concerning the ratio of these requirements and their yields. The comparison of Tables 3.2 and 3.3 shows that those decisions, combined with the changes in deposit interest rates that were introduced at the same time, reduced the! average cost of banks' resources from deposits available for lending from around 37 percent to 30 percent. The effect on marginal costs was even more important, as suggested by the fact that the interest cost of resources from time deposits used for nonpreferential credits fell by 28 percentage points, from 86 percent to 58 percent. 3.17 The higher interest rates now being paid on the reserve requirements imposed on sight deposits increased the costs of the Central Bank. In order to compensate for the negative impact of such an increase on the profitability of the Central Bank and, indirectly, on the Government budget, there would be justification in increasing slightly the rediscount rates, especially for selective credits. 3.18 The difference between the average interest rates paid by the banks to depositors and the total costs of the resources from deposits availabLe for credits is now only about 4.2 percent, as against about 6.4 percent in 1982. That difference is still quite high and it should be further reduced. The  - 23 - reduction of the burden imposed by the liquidity and reserve requirements on the cost of credits can be achieved by two alternative methods: either by reducing the ratio of the liquidity and reserve requirements or by increasing the average interest rate paid on reserve requirements. 3.19 If the ratio of reserve requirements was lowered to 10 percent, the difference between the average interest rates paid by the banks to depositors and the total costs of the resources from deposits available for credits would fall to 0.7 percent. This would, however, imply a substantial reduction of the monetary base to maintain the same level of the money supply. For this to occur, it would be necessary to cut drastically the amount of credits outstanding of the Central Bank and to the public sector and also the amount of Central Bank rediscounts of credits extended by other financial institutions, consisting predominantly of selective credits. This means that if the solution of reducing the ratio of reserve requirements is envisaged: (a) it should, in principle, be implemented gradually (for instance in several steps, each one involving reductions of 5 percentage points in the reserve ratio), because of difficulty of an abrupt reduction of the monetary base; (b) it would imply considerable reductions in the amount of rediscounts of selective credits (especially agricultural and export credits) and in credits of the public sector (including in particular advances to the Treasury and credits to TMC and the State Monopolies). More important than increasing such rates, however, would be to reduce the great disparities which are found at present in the distribution of the burden of the reserve requirements between sight deposits and time deposits. The interest paid on the funds kept in liquidity and reserve requirements corresponding to public and commercial sight deposits provides some profit to the banks, although it is much lower than what the banks would get if they could use such funds for credits. By contrast, the liquidity and reserve requirements increase the cost of resources from 6-month deposits available for credits from 44 percent to 58 percent. Hence, commercial banks will still need to block about 25 percent of their non redisountable loans in order to get a marginal revenue equal to the marginal cost of the funds from time deposits available for credits. 1/ 1/ In nonpreferential loans the banks receive an interest rate of 36 percent plus a commission of 2 percent. With blocking of 25 percent of the proceeds of the loan, a credit of TL 100 increases the commercial deposits by TL 25. Of this amount, 35 percent, or TL 8.75 has to be placed as liquidity and reserve requirements. The remaining TL 16.25 can be used for additional loans. If blocking of 25 percent would be required on those additional loans, the process could continue indefinitely, so that the total amount of credit would expand-to 100 + 16.25 : 0.35=TL 146.4. The liquidity and the reserve requirement would rise by TL 5, producing a net yield of 25 x (0.286 x 15% + 0.714 x 19.6%)=TL 4.6. The total revenue of an initial loan of TL 100, with successive blocking of 25 percent, would thus be 146.4 x 38% + 4.6% = 60%.  - 24 - Table 3.4: EFFECTS OF THE PROPOSED CHANGES IN THE INTEREST RATES PAID ON REQUIRED RESERVES (percent) Public and Sight Average Commercial Savings Time of all Item Deposits Deposits Deposits Deposits 1. Composition of total deposits in the last quarter of 1982 34.0 12.0 54.0 100.0 2. Interest rate paid on the reserve requirements 2a. At present 19.6 19.6 19.6 19.6 2b. Proposed 0.0 10.0 36.0 21.8 3. Earnings on liquidity and reserve requirements - 3a. At present 6.4 6.4 6.4 6.4 3b. Proposed /a 1.5 4 10.5 6.7 4. Average cost of funds from deposits available for lending: 4a. At present -9.8 20.9 57.8 30.4 4b. Proposed /b -2.3 24.6 51.5 30.0 5. Differences between the cost of funds from deposits available for lending and the interest rate paid to depositors 5a. At present -9.8 0.9 13.8 4.2 5b. Proposed /c) -2.3 4.6 7.5 3.8 /a 0.10 x 15% + 0.25 x row (2b) of this table. /b Line (1) of Table 3.3 - line (3b) of this table):0.65. /c Line (4b) of this Table - line (1) of Table 3.3.  - 25 - 3.21 To correct this situation, the interest rate paid on the reserve requirements from time deposits should be higher. In order to illustrate this proposal, Table 3.4 shows, on the basis of the composition of deposits in the last quarter of 1982, what would be the effect of raising the interest rate on the reserve requirements corresponding to time deposits from 19.6 percent to 36 percent. This would increase only slightly the total costs of the Central Bank if, at the same time, the rates on the reserves required from sight deposits were reduced from their present level of 19.6 percent to zero in the case of public and commercial deposits, and to 10 percent, in the case of savings deposits. With such changes, the difference between what the banks are paying for deposits and the costs of the funds available for lending would fall slightly from 4e2 percent to 3.8 percent on average terms but the changes would be substantial on the different types of deposits; the difference between the cost of loanable funds and deposit yields would be reduced from 13.8 percent to 7.5 percent in the case time deposits, and it would increase from -9.8 percent to -2.3 percent in the case of commercial and public sight deposits, and from 0.9 percent to 4.6 percent in the case of sight savings deposits. The marginal costs of funds from time deposits used in credits would be reduced by 6.3 percentage points, and there would be much less disequilibrium than now between the costs of funds from the different types of deposits. E. Contributions to the Interest Rate Rebate Fund 3.22 A contribution of 10 percent to the IRRF is levied on the interest paid on non-selective credits. Export credits of non-agricultural goods pay 5 percent to the same Fund. The proceeds of those contributions are used to finance interest rate subsidies in favor of different types of selective credits (see Tables 11 and 12 in the Statistical Annex). In 1982 the expenditures of the IRRF exceeded its revenues. 3.23 The levy to the IRRF increases the cost of nonpreferential credits by 3.8 percentage points. The elimination of that levy could provide a significant contribution to lowering the interest costs paid by many borrowers. It will not be possible, however, to cut the revenues of the IRRF unless its expenditures are reduced by the same extent. That would require cutting the interest rate subsidies of selective credits, as recommended in Chapter 4. However, if selective credit policies are not going to be completely abolished, it will be preferable to maintain the contributions to and the subsidies from the IRRF--rather than, for instance, to subsidize the selective credits by means of preferential rediscount facilities or to lower the required reserve ratios, which impose the same cost burden on nonpreferential credits and tend to create more disturbances for the conduct of monetary policy. 3.24 In such conditions, the possibilities of reducing the rates of the contributions to the IRRF are very limited, particularly so because by the end of 1982 the accounts of the IRRF were starting to show a deficit. Thus, even if significant efforts are made to reduce the range of preferential credits and the degree of subsidization of their interest rates, it would not be realistic to envisage the elimination or the reduction of the contributions to the IRRF.  - 26 - F. Elimination of the Financial Transaction Tax 3.25 The financial transaction tax is levied on the interest earnings of credits and other revenues received by the banks. Export and agricultural credits, however, are exempt. The rate of the tax at present is 15 percent. Its effects on the cost of credits are substantial: it increases the cost of nonpreferential credit from 38 percent (including the bank commission of 2 percent) to 43.7 percent. If blocking of 25 percent of the credit is required, the cost with commission will increase from 50.7 percent before the tax to 58.3 percent after it. If the inflation rate in 1983 comes down to 20 percent, this means that the t:ransaction tax increases the real interest cost of credits with blockage of 25 percent from about 25 percent to approximately 32 percent. 3.26 The revenues provided to the Treasury by the transaction tax are quite important. In 1982 they accounted for about 4.5 percent of total tax revenues. The elimination of the tax will thus pose considerable difficulties for the reduction of the deficit of the public sector budget. However, there is an urgent need of correcting the detrimental effects of the tax on financial intermediation, on investment, and on the survival of firms that are being penalized by excessively high interest rates on their debts. For this reason, the authorities should seek fiscal revenues with a less negative effect on economic growth and should, as a matter of great priority, eliminate the transaction tax. If the elimination in one step is not feasible because of the budgetary constraints, it should be made in two or three steps. These steps should be implemented within a predetermined period--for instance, two years. G. Costs of Bank Intermediation 3.27 As explained in more detail in Chapter 6, the costs of operation plus the profits of the Turkish deposit money banks corresponded in 1980 and 1981 to about 11 percent of the toi:al value of their assets and in 1982 to about 8 percent of the same total. These very high intermediation margins are covered in substantial portion by the high commissions that the banks are charging for their services, particularly in foreign exchange operations and in bank guarantees. At the same time, the high intermediation cost of the banking system is one of the important reasons that lending interest rates are so expensive in Turkey. In fact, these costs imply that, even if the transaction tax and the contributions to the IRRF were abolished and if the liquidity and reserve requirements did not impose any net cost on the banks, borrowers would have to pay for bank credits about 6 percent more than the average interest rate paid to depositors, on the assumption that commissions from bank services would cove:r about one-third of the total intermediation margin. If the Turkish banking system were as efficient as in most other OECD countries, that difference would be only 2 to 3 percent, and the commissions charged for bank services couLd be one-third of what they are. 3.28 In spite of the progress achieved in 1982, the reduction of the operating costs of the banks is thus one of the most needed improvements in the Turkish financial sector. The measures that must be implemented to achieve such an improvement are analyzed in Chapter 6. Unfortunately, the problem is of a structural nature, and cannot be solved quickly.  - 27 - H. Lending Interest Rates and the Supply of Credit 3.29 It was shown above, in the comparison of Tables 3.2 and 3.3, that the new ceilings on interest rates and the new regulations on liquidity and reserve requirements in force since January 1, 1983, have reduced the average cost of resources available for credits by about 7 percentage points on average and by around 28 percentage points on funds from time deposits used in non-selective credits. It is roughly estimated that for each additional reduction of 1 percent in the interest rate for time deposits that may be introduced in the future, the average cost of the funds used in credits will drop by about 0.5 percent. Moreover, the elimination of the transactions tax, which is strongly recommended, would reduce the cost of general short-term loans by a further 5.7 percentage points. 3.30 It is not certain, however, that all these cost reductions would be transferred to the borrowers. Lending rates are determined not only by the cost of funds used in bank credits but also by the relationship between demand and supply of those credits. As explained in Chapter 5, it seems that many of the Turkish firms find themselves in a vicious circle. Because their operational profits are not sufficient to cover the high real interest costs charged on their debts, firms have to increase their borrowing to meet the interest payments. This creates great liquidity pressure on the banks and, as mentioned before, has led many of them to pay interest rates above the ceilings of the gentlemen's agreement in order to attract additional deposits. 3.31 If this situation continues, there is the risk that the attempts to reduce the cost of credit by introducing progressively lower ceilings to the interest rates paid on deposits will not produce effective results. For the same reason, the reductions in the costs of the reserve requirements already in force and the elimination of the transaction tax may be reflected in higher intermediation margins for the banks or in higher interest rates for the depositors, above the announced ceilings, rather than in lower interest rates for the borrowers, as desirable. This problem cannot be easily solved by imposing strict ceilings on lending rates because the banks can easily circumvent those ceilings by blocking part of the proceeds of the loans, as they do already. 3.32 In such conditions, it will be necessary to follow closely the. effects that the measures introduced in January 1, 1983, will have on the level of lending rates and on the average proportion of blocking. By the middle of 1983, it was apparent that these measures were contributing to improving the profitability of banks, which in many cases reached unsatisfactory levels in 1982, but they were not yet clearly reflected in lower lending rates. The behavior of the lending rates in the second half of 1983 will provide important guidelines for future action. If it becomes apparent that the liquidity situation prevents the interest rates from falling to more moderate levels, then this will suggest that the total supply of credit is a constraining factor. 3.33 The problem is that a faster increase in the supply of credit would certainly have a negative impact on the objective of reducing inflation. However, if the experience of the next few months proves that real interest rates cannot be brought down as much as is necessary by measures such as those implemented at the beginning of 1983, then the Turkish authorities will have a difficult dilemma.  - 28 - In order to maintain their targets for a rapid reduction of inflation, the economy will have to suffer the negative effects of abnormally high real interest rates. But if they try to reduce those interest rates substantially by a faster expansion of domestic credit, then they run the risk of rekindling inflation or at least of not being able to achieve the desired improvements in the inflation rate. Since the objective of fighting inflation continues to deserve a high priority, it is very probable that the level of real interest rates cannot be brought down very quickly in the near future. 3.34 Several measures recommended in this report may help in resolving the dilemma mentioned in the preceding paragraph, but their contribution will only be partial. Such measures include: (i) further efforts 1:o reduce the borrowing requirements of the administrative public sector from the banking system, in order to leave a greater share of bank credits to the productive sector (as recommended in Chapter 2); (ii) the frequent adjustment of the overall credit ceilings to the increase in demand for M2, which, as mentioned in Chapter 1, is likely to continue even if real interest rates on time deposits would fall to more moderate positive levels; (iii) an increase in the availability of general credits by reducing the share of total resources at present allocated to selective credits (as recommended in Chapter 4); (iv) the elimination of the transactions tax; (v) the reduction of the ratio of reserve requirements or the payment of higher interest rates for the reserves required on time deposits (as recommended in the preceding paragraphs of the present chapter); and (vi) encouraging the rapid development of capital markets, and in particular a revitalizing of the bond market (as recommended in Chapter 7) with the objective of creating competition to the banking system, of forcing the reduction of the high intermediation costs of that system and of contributing to the increase of the demand for financial assets and, consequently, to the expansion of the real amount of finance available for firms.   - 29 - CHAPTER IV - SELECTIVE CREDIT POLICIES A. Controlling Credit Allocation Introduction 4.1 In Turkey, as in most developing countries, the Government intervenes in the credit markets to influence the allocation of credit among sectors. Wbat is different in Turkey is the extent of the system and its administraltive complexity. Roughly three-quarters of all advances from the financial system are made at Government directive, at interest rates determined by the Government, or both. The list of those receiving special credits at preferred interest rates includes public administration, public enterprises, agriculture, exports, artisans and small firms, housing, industrial investment, outlays in backward regions, and so on. Within agriculture thiere are subprograms for sales cooperatives, credit cooperatives, and others. Other sectoral programs are also subdivided--for example, by loan maturity--witb different interest rates and different conditions applying to each subprogram. Hence, on a sectoral basis the credit markets of Turkey are extensively fragmented into a plethora of noncompeting segments with wideLy differing interest rates applying to loans in the various submarkets. 4.2 A vast array of controls have been developed to implement this system. On all but the loans classified as general (constituting about one-quarter of the total), the Government specifies the interest rate to be charged. To help the banks defray the expenses of low-interest loans, the Government has established an Interest Rate Rebate Fund (IRRF), which provides interest rate subsidies to preferential credits (as indicated in Table 12 of the Statistical Annex). Non-preferential credits must pay a surcharge of 10 percent of the interest paid into the Fund. In addition to the subsidies from the IRRF, there are special discount rates available from the Central Bank for loans made to priorit:y sectors (as indicated in Table 10 of the Statistical Annex). Until the end of 1982, wben banks used their funds from deposits for certain preferent:ial loans there was a corresponding reduction in the reserve requirements, whichb enhanced the profitability of such loans for the banks. In addition to these basic system8, the State Investment Bank, wbicb lends to public enterprises, bas received special loans from the . Treasury and from the Social 'Security System at favorable interest rates. In the past many public enterprises bad access to low-interest loans from the Central Bank; at present such loans are limited to the Soil Products Office and the State Monopolies. The Government keeps substantial non-interest-bearing deposits at the Agriculture Bank; this in turn enables the Bank to make low cost loans to agricultural sales cooperatives and to farners. Finally, the Government uses its influence to arrange foreign loans at favorable rates for the development banks, for industry and for agriculture. 4.3 Special credit programs for the private sector have two basic objectives. Paramount is the desire to increase credit flows to particular sectors in order to enhance economic activity in priority areas. The second objective is to transfer income to recipients of special credits through low interest rates. It is clear t:hat in Turkey, whatever the original intention, the second objective has now taken precedence over the first. There bas been much wider dispersion of interest rates, but at the same time the supply of  - 30 - preferential credits has been severely restricted, and their share in total credits has fallen. With the system of contributions to the IRRF and preferential rediscounts, the major costs of low-interest loans is borne by general borrowers. Investigations of credit subsidy schemes in other countries indicate that not all the benefits go to the intended final uses. There are often cases of misuse of selective credits, and often selective credits displace other forms of finance that could otherwise be initiated. The effects of those credits on the real allocation of resources may consequently be more limited than it would appear at first sight. 4.4 As now implemented, interest rates are set so low in the special credit programs that for each type of credit there is excess demand. Hence, it is necessary to have borrowers of credits to agriculture, housing, artisans, exports, and the like wait in a queue for loans or to ration credit by other means. For borrowers, the availability and timing of credit is often more important than the price. Rationing can produce rather arbitrary allocations, and delays can have an adverse effect on the borrower's situation. 4.5 It is the function of financial markets and institutions to assess applications for funds and to select projects for financing that have the greatest chance of success. Better allocation and enhanced productivity of capital is achieved if credit is correctly allocated. In Turkey, the numerous special credit programs fragment the markets and prevent the financial system from performing the allocative function. The evidence suggests that this has adversely affected economic efficiency. 4.6 Another serious difficulty raised by selective credit policies is their negative impact on the cost of nonpreferential credits. As explained in Chapter 3, that cost is increased considerably by the contributions to the IRRF and by the reserve requirements imposed on deposits with the purpose of mobilizing resources for rediscounting selective credits to the private sector and for granting Central Bank credits at low cost to the public sector. The reduction of the cost of nonpreferential credits, which may be crucial for the survival of many firms (see Chapter 5), will not be possible if the interest rate subsidies to preferential credits remain too high. Improvements in the System of Selective Credits 4.7 In face of all these reasons, the lowest interest rates in preferential credits should be significantly increased. The availability of such credits is certainly more important than their cost. Because severe rationing is unavoidable in view of the constraints on the overall supply of credit, higher interest rates will also be preferable from the point of view of equity in the distribution of special benefits. As explained in Chapter 3, the main priority of financial sector policy in Turkey in the present conditions should be to reduce the excessive costs of nonpreferential credits. Although that objective should be pursued by the adoption of several measures analyzed in other chapters of this report, an important contribution should also come from the reduction of the amount of subsidies paid to preferential credits. 4.8 At a minimum, the interest rates charged on all special credits should be equal to the rate of inflation. With the prospect of declining inflation in the coming years, this principle will not be far from being  - 31 - observed even with the most subsidized credits. Since the projected inflation rate for 1983 is 20 percent, even the interest rates on the most subsidized credits to housing, agriculture and artisans are now close to zero in real terms.!' But in the present difficult situation, this will hardly be sufficient. Even the lowest interest rates sbould be significantly positive in real terms. As long as the real interest rates on nonpreferential credits exceed 20 percent, it will be difficult to justify interest rates in preferential credits of less than 5 percent. At present, a real rate of 10 percent in preferential credits would still imply a differential of at least 15 percent compared with the cost of general credits. 4.9 Furthermore, the plethora of different rates is unnecessary. Soae of the interest rate differentials are witbout significant rationale. The number of special rates in short-ter. credits, should be reduced. 4.10 The wide variety of special credit programs in Turkey creates a great fragmentation of the financial markets. To give priority to sucb a wide variety of activities is much the same as having no priorities. The Government should define priority activities in need of special credit allocations quite narrowly. Exclusion from special allocations does not mean that these sectors will not get credit. With less fragmentation into submarkets, the financial system will do a better job of credit allocation. l 4.11 In the second half of 1982, there was a substantial shift in deptosits among banks, in particular into deposits in the Agriculture Bank. Hence, there bas been a large increase in the resources that bank has available for lending. The Government must monitor such deposit sbifts not only because of their impact on liquidity but also for their impact on credit allocation. If such shifts were to lead to an excessively rapid buildup in credits to agriculture at the expense of loans to other sectors, the autborities might offset them by reducing the rediscounts of agricultural credits. B. Agricultural Credit The Agricultural Credit System 4.12 The agricultural credit system includes credits to farmers, to credit cooperatives and to sales cooperatives. Between 1975 and 1980 the real level of agricultural credit 2/ was reduced by almost one-balf, and as a share of agriculture GDP (value added in crops and livestock) it fell from 27 to 16 percent (Table 4.1). The trend is similar for agricultural credit as a sbare of total bank credit, a decline from 19 to 20 percent in 1974-75 to 15 percent in 1980. In 1981, bowever, there was a reversal. Agricultural credit grew by almost 16 percent in real terms over 1980; as a sbare of agricultural GDP (value-added), which grew by less than 1 percent, it increased to 21 percent. 1/ Recent information indicaltes that inflation is likely to exceed this target by 5 to 10 percentage points. 2/ Funds extended by the Cenitral Bank to TMO and otber agricultural SEEs are not included bere in agricultural credits since they are used for purchases of commodities from farmers and not for credit. In many cases, the farmer initially only receives partial payment for crops, witb the balance paid after the SEE has sold the crops. Credit is thus extended "in kind" from the farmers to the SEE.  - 32 - Table 4.1: THE SUPPLY OF AGRICULTURAL CREDIT (year-end figures) Item 1975 1976 1977 1978 1979 1980 1981 1982 Total Outstanding Agricul- tural Credit (TL billions) Current prices 35.1 39.5 49.1 52.9 82.6 145.8 266.0 356.4 1975 prices 35.1 33.1 30.2 21.9 18.9 17.1 24.8 26.7 Agricultural Credit as percentage of: Total credit 18.6 15.1 14.7 12.9 14.0 15.1 17.3 13.3 Value added in crops and livestock 26.5 22.9 23.3 18.2 18.4 16.3 20.7 21.9 Source: Central Bank, Quarterly Bulletin. 4.13 More than 95 percent of the institutional 1/ supply of agricultural credit is provided by the Agricultural Bank, TCZB. The share provided by commercial banks is negligible (0.6 percent), and other sources include two agricultural SEEs, Sekerbank (Sugar Bank) and EBK (Meat and Fish Authority), which together supply about 4 percent of the total. The sales cooperatives, which purchase and sell agricultural commodities whose prices are supported by the Government, have received an increasing share of TCZB credit during the 1970s; currently they account for about one-half of the total (Table 4.2). Credit to such cooperatives is essentially trade credit related to agricultural products rather than credit to farmers, who often have to wait several months for the payment of the products delivered to sales cooperatives. The credit cooperatives, which extend credit mainly for short-term input financing, particularly for fertilizer, have been allocated between 10 to 15 percent of total TCZB credit. The balance has gone primarily to individual farmers. 1/ That is, credit which is not self-financed or financed in informal markets.  - 33 - Table 4.2: AGRICULTURAL CREDIT BY RECIPIENT Item 1975 1976 1977 1978 1979 1980 1981 1982 Outstanding Agricultural Credit (TL billion) Sales Coops 21.8 22.5 27.9 24.0 44.7 72.9 141.4 151.6 Credit Coops 3.2 4.0 5.0 7.4 10.0 21.9 34.7 50.3 Other 10.0 12.8 16.0 20.9 27.8 51.0 89.8 154 5 Total 35.1 39.5 49.1 52.9 82.6 145.8 266.0 356.4 Percentage Share Sales Coops 62.1 57.1 56.8 46.3 54.1 50.0 53.1 42.5 Credit Coops 9.2 10.5 10.6 14.1 12.4 15.0 13.0 14.1 Other 28.5 32.4 32.6 39.6 33.2 35.0 33.8 43.3 Source: Central Bank of Turkey, Quarterly Bulletin. 1/ That is, credit which is not self-financed nor financed in informal markets. 4.14 The growing dependence over the last decade on the Central Bank as a source of funds for agricultural credit is striking. In 1971, the Central Bank financed 30 percent of total agricultural credit, and by 1981 this s'hare had grown to more than 80 percent. This development primarily reflects the increasing volume of funds channeled to the cooperatives; equity contributions to the sales and credit cooperatives equivalent to 10 percent of their lending provide them with very limited funds of their own, insufficient even to cover their operating expenses. The sales cooperatives have absorbed the largest share of Central Bank funds throughout the decade, but the credit cooperatives have needed a growing percentage to cover fertilizer costs following the substantial price increases in 1980 and 1981. The fivefold increase in fertilizer price in 1980 was reflected in a threefold increase in Central Bank rediscounts with the credit cooperatives in that year. In view of the government commitment under World Bank Structural Adjustment Loan II to phase out fertilizer subsidies over the next five years, the financing needs of the credit cooperatives are likely to continue to grow. 4.15 As a result of this huge drain on Central Bank funds for the short-term credit needs of inventory financing (sales cooperatives), as well as for the operating needs of agricultural SEEs (such as TMO, Meat and Fish Authority, Sugar Factories, T,ea Authority, and the like) and for input purchase (credit cooperatives), very little funds have remained for investment 1/. In fact, Central Bank rediscounting of medium-term agricultural loans has been negligible. Medium-term agricultural credit 1/ As of end-1981, 40 percent of outstanding and direct Central Bank credit to the public sector was to agricultural SEEs.  - 34 - (excluding credit to TZDK, the Turkish Agricultural Equipment Agency, for fertilizer) has been between 10 to 15 percent of TCZB's portfolio, around 10 percent of which has been term credit for agro-industries. This reflects the lack of incentive for TCZB to extend medium and long-term credits. Interest Rate Policy in Agricultural Credits 4.16 All institutional agricultural credit has been distributed at highly subsidized rates. The interest rate on short-term lending is currently 20 percent, and on medium- and long-term lending it is 22 percent in nominal terms. On the basis of the official forecast of the inflation rate for 1983 of 20 percent, this corresponds to rates close to zero in real terms. As mentioned above, more than 80 percent of agricultural credits are financed through the Central Bank. They receive interest rate subsidies of 20 to 15 percent, based on a preferential rediscount rate from the Central Bank. The high cost of these subsidies is supported either by the Treasury, which maintains large deposits at zero interest rate with TCZB, or by nonpreferential credits which are affected by the contribution to the IRRF and by the large reserves required on deposits. Such a large subsidy has led to an excess demand for agricultural credits and has created incentives for their misuse and diversion. As long as the subsidization is based on Central Bank rediscounts, the overall supply is constrained by the ceiling on the expansion of domestic financial assets of the CentraL Bank. A process of rationing of agricultural credit has therefore emerged, with priority given to the financing of the cooperatives engaged in price support and input purchases. Credit allocated directly to the farmer by TCZB is thus a residual, and it has tended to be insufficient for the farmer's needs, late in being disbursed, and often given in installments. The share of credit allocated directly to the small and medium-scale farmer seems to be particularly small. 4.17 The interest rate policy should be reconsidered. While there is argument for preferential treatment to be given to agriculture through a selective credit policy, the current size of the subsidy is very high and raises serious questions about the cost effectiveness of resource transfers of such a magnitude in achieving either production or equity objectives. At a minimum, agricultural interest rates should not be allowed to become again negative in real terms. In accordance with what is recommended in para. 4.18, the authorities should even consider the possibility of raising those rates to at least 5 percent in real terms if the real costs of other credits continue to exceed 15 percent for a long period. The objective should be to reduce the extremely wide differential between the interest rates in agricultural credits and the nonpreferential credits. 4.18 With respect to medium and long-term lending, the interest rates should also be raised gradually to reduce the disparity with regard to the interest rate on credits to other sectors. A system of variable interest rates as described in paras. 4.63 to 4.72 should apply to these loans. When the reference interest rate is changed, the new interest rates would apply to all outstanding loans; at present, outstanding medium and long-term loans are affected by changes in agricultural interest rates, but interest rates are not revised on a regular basis.  - 35 - Institutional Problems 4.19 In common with other financial institutions in Turkey, the intermediary institutions involved in agricultural credit have very high intermediation costs. They enjoy a monopoly position in agricultural lending, and, although they have an extensive branch network, administrative procedures are cumbersome and result in an inflexible lending policy and delays in processing. There are serious institutional problems with the sales cooperatives. As with TCZB, they suffer from overstaffing and also from a lack of accountability either to their members or to TCZB. This has led to an unwarranted expansion of their staff, with all operating expenses covered by cheap Central Bank credit, and to an unchecked expansion of their manufacturing activities. Given the current shortage of agricultural credit, particularly production credit, as well as the large projected requirements over the medium-term 1/, resulting in part from further reductions in input subsidies, major institutional and policy reform is needed if this waste is to be curtailed and requirements met. 4.20 Significant improvements in TCZB's operational efficiency should be sought, among other means, by encouraging more competition in the agricultural credit market. TCZB's monopoly position in agricultural lending should be contested by encouraging other financial institutions to participate more actively. Given the overall shortage of loanable funds and the excess demand for credit from more lucrative commercial and industrial lending, it is unlikely that commercial banks would enter the agricultural credit market in a big way. Commercial banks could, however, become interested in extending credit to agro-industrial enterprises and larger farmers if interest rates were higher and if the banks were given equal access to rediscount facilities for agricultural loans at the same terms given to TCZB. This could release some of TCZB's funds for additional credit to small and medium-scale farmers. Also, the privately-owned investment and development banks, TSKB and SYKB, could be a source of investmenLt credit for agro-industries. More competit:ion to TCZB in lending to farmers can also come from agricultural credit cooperatives if their role is strengthened. These cooperatives have already an extensive branch network and are active at the grass-roots level. Differentiation in the Types of Agricultural Credits 4.21 There should be a clearer differentiation in the types of credit supplied, in conditions of leniding and the institution involved. In particular, a distinction should be made between: (i) crop purchase activities; (ii) input financing; (iii) medium and long-term credit; and (iv) small farmer credit. 1/ The growth targets in the recent World Bank Agricultural Sector Survey imply a doubling in the provision of agricultural credit in real 1981 terms during the 1980s, and this could easily become larger if supplemented by a strong farmer demand for investment credit. Agriculture Development Alternatives iFor Growth with Exports, World Bank, (Report No. 4204-TU) Grey Cover, June 30, 1983.  - 36 - Crop purchase activities. 4.22 Sales cooperatives have the network and expertise to undertake the crop-marketing operations and should continue to do so. The financing of such cooperatives should, however, be re-examined. First, the role of TCZB and the Central Bank should be more clearly defined. At present TCZB is acting as a mere agent in providing credits to the sales cooperatives on behalf of the Central Bank. TCZB has suffered, however, from delays in the reimbursement of the losses and arrears from such credits. Since there are only 15 sales cooperatives, it would be easy for the Central Bank to finance them directly, without using TCZB as an agent. On the other hand, given the constraints on the expansion of Central Bank credits, it may be preferable for the conduct of monetary policy that part of the loans needed by the sales cooperatives be provided by TCZB rather than by the Central Bank alone. However, if the present system is maintained, it should at least be clearly established that the credit risks are the responsibility of the Central Bank or of the Treasury. In order to limit such risks to moderate levels there should be strict supervision of cooperatives with arrears. These cooperatives should be penalized by refusal of further loans and by interest rate surcharges on their arrears. Second, sales cooperatives should function strictly as a marketing agency and should not continue with their manufacturing activities unless they have the support of their members and are subject to appropriate supervision. Third, their operating expenses should be limited, and the profits should be returned to the farmers unless retained to finance investments that are approved by cooperative members. Fourth, they should have their accounts audited by procedures at least as stringent as for SEEs. Finally, as discussed above, the interest rate subsidy for crop purchasing should be substantially reduced. Input financing. 4.23 The present system appears to be working reasonably well through the credit cooperatives except for the general shortage of loanable funds. The credit cooperatives provide funds primarily to finance small and medium-scale farmers. The activities of the individual cooperatives are monitored by regional unions, which also provide technical assistance to the cooperatives. The level of subsidy, however, is excessive, and, as noted above, the interest rate on these short-term credits should be increased. It is also important that the resource mobilization efforts, particularly of the credit cooperatives, be improved. This is important in view of the significant increase expected over the medium-term in the financing requirements for inputs, as mentioned in paragraph 4.19. Medium and long-term credit. 4.24 Currently, less than 20 percent of agricultural credit is medium- to long-term. In the first half of the 1970s there was substantial investment in agricultural vehicles and machinery; this investment peaked in 1976, and there has since been a steady decline in total fixed investment in agriculture. It is unlikely that there would be a major demand for investment credit for heavy vehicles and machinery until the end of the 1980s, although demand for smaller equipment items could develop earlier. Substantial investment credit would, however, be needed for orchards and for the development of newly irrigated land. Investment credit is also needed for rehabilitation of buildings.  - 37 - 4.25 To increase the share of TCZB funds available for medium and long-term credit, there should be: (i) greater efficiency in the operations of the intermediaries involved in agricultural credit, particularly TCZB and the sales cooperatives; (ii) larger sbares of the agricultural credits from the Central Bank and from TCZB allocated to medium and long-term credits and improved efforts of resource mobilization in the rural areas; (iii) variable rates oi interest on medium and long-term agricultural credit, based on the solutions mentioned in paras. 4.63 to 4.72; and (iv) improved project appraisal by TCZB, and speedier approval and disbursement of credit. Small-farmer credit. 4.26 While it is recognized that further fragmentation of the Turkish financial system is ill-advised, the fact remains that very little credit is going to small-scale farmers and that some provision should be made to increase the share of credit t:o this group. This could be done by requiring TCZB to extend a minimum share of its loans to small-scale farmers or by increasing the lending to sma]Ll-scale farmers by cooperatives, possibly witb World Bank assistance. The objective of increasing the share of agricultural credits going to small farmeri; will, however, be of difficult implementation, because of the problems of controlling the ultimate uses of the credits, of defining small farmers, of avoiding the excessive accumulation of arrears, etc. In such circumstances, ithe improvement of the system of credits to small farmers should be preceded by careful study. Such a system would be greatly assisted by a credit risk guairantee scheme, which should be available to all financial institutions so as ito encourage maximum participation. C. Export Credits The Growth of Eport Credits 4.27 Export credits have 'been crucial to Turkey's achieving the rapid growth of exports in recent years. The interest rate charged for export credits has been substantially below the rate for nonpreferential credits although the gap between the two bas narrowed relatively over time. The subsidized interest rate is justified on the basis of the bias against exports stemming from the protection of the domestic market. Also, the share of export credits in total concessional loans extended by the banking system has steadily increased in recent years. This has been a critical factor in stimulating export activities, in a situation in which credit availability has been severely constrained. Export credits refinanced by the Central Bank alone represented 32 percent of total exports in 1979, 22 percent in 198CI, 21 percent in 1981 and 11 percent in 1980. With commercial bank contributions out of own resources included, the shares would increase to 41 percent inI 1979, 29 percent in 1980, 36 percent in 1981 and 40 percent in 1983 (Table 4.3). In otber words, around 30 to 40 percent of exports bave been  - 38 - financed one way or another at preferential terms. In 1982, the amount of export credits financed by the Central Bank declined by about 11 percent in nominal terms, in the face of about 80 percent increase in the Turkish Lira value of exports. That decline has created a long queue at the Central Bank rediscount window. It appears, however, that, in spite of lower rediscounts from the Central Bank, deposit money banks continued to expand their export credits rapidly. Statistics on export credits extended by these banks have only been compiled on a comprehensive basis since the beginning of 1982. Those statistics show that between end January 1982 and end December 1982, such credits increased by TL 176 billion, from TL 202 billion to TL 378 billion or by 87 percent. However, deposit money banks have been complaining about the serious liquidity difficulties created for them by the reduction of Central Bank rediscouits of export credits. The Export Credit Scheme 4.28 Exporters have access to export credits through two main channels. First, firms with a letter of credit (L/C) can directly apply to a commercial bank for export credit. Second, firms who do not hold an L/C may obtain from the State Planning Organization an Export Promotion Certificate upon making an export pledge and then apply to a commercial bank. Commercial banks can rediscount up to 45 percent of the credit at the Central Bank. Table 4.3 shows that more than two-thirds of the Central Bank export credits fall into these categories. The other important ones are tobacco export financing and export promotion fund credits earmarked for exports of fresh fruits and vegetables. Table 4.3: EXPORT CREDITS EXTENDED (TL billions) Item 1979 1980 1981 1982 1. Central Bank credits 382.1 655.2 925.5 910.5 l.a. Export credits 24.3 48.1 113.2 100.6 (% share) /a (6.4) (7.3) (12.2) (11.5) l.a.l. Without certificate - 5.6 46.5 23.7 l.a.2. Realized sales 4.9 5.6 - - 1.a.3. Tobacco 3.2 5.0 8.8 3.0 1.a.4. Bills with documents 0.3 0.4 0.5 0.5 l.a.5. Export promotion fund 2.2 5.7 17.9 0.6 l.a.6. With certificate 13.7 25.8 39.5 72.8 2. Total export credits 31.0 /b 64.0 /b 189.0 /b 377.5 3. Exports 75.7 221.5 530.7 934.0 (% share) /c (41) (29) (36) (40) Note: 1979-81 figures are at end-year; 1982 figures are at end-October. Sources: Central Bank;. mission estimates. /a Row (1)/row (l.a). /b Including commercial bank contribution; rough estimates based on average proportion of commercial bank credits rediscountable at the Central Bank. /c Row (2)/row (3).  - 39 - 4.29 The credit limit of export financing varies, but normally it never exceeds 80 percent of the value of exports. The term is up to 8 months, with the possibility of an extension to 12 months. The extension is granted in most cases in which 60 percent or more of the export pledge has been fulfilled. In case of defaults, the interest rate would be raised, and stamp and other duties would be imposed. 4.30 The interest rate charged for export credits is 31.5 percent, as opposed to 36.0 percent for nonipreferential credits. The interest is exempt from the financial transaction itax levied on other credits. In addition, credits for the exports of industrial products are exempt from the contribution to the IRRF and re,ceive from that Fund a subsidy of 15 percent. Consequently, final borrower cost is at least 18 percent lower for export credits of industrial goods than for nonpreferential credits (excluding the differential cost of blocking, on average 25 percent of credits, as compensatory balance for nonpreferential credits--as opposed to between 15 and 20 percent in the name of guarantee deposit for export credits). If the effect of blocking is considere,d, the real cost of export credits, measured in relation to the increase of the wholesale price index, was around 8 percent in 1982. 4.31 The interest differential has been substantially reduced since early, 1981 (Table 4.4), but it has increased for other types of concessional loans (e.g., credits for agriculture, housing, artisans, and small traders). With a view to encouraging the coimnercial banks to grant export credits from their own resources, however, the IRRF provides an interest rebate of 10 percentage points since August 1981. Similarly, the contribution of the Central Bank iTn rediscount credits for exports was adjusted downward from 75 percent to 55 percent in November 1981 and to 45 percent in March 1982 to reduce the pressure on Central Bank credits. Table 4.4: FINAL INTEREST COST TO BORROWERS AND SUBSIDY TO BANKS 1/ (percent) Nonpreferential Export credits credits /a (end-year) - 1979 - Feb. 1980 7.9 24.4 Mar. 1980 - Jun. 1980 16.7 34.0 Jul. 1980 - Jan. 1981 17.4 49.0 Feb. 1981 - Dec. 1981 22.2 49.1 Jan. 1982 - Apr. 1982 23.6 47.3 From May. 1982 28.8 47.3 Memo: Subsidy from the IRRF to banks for extending export credits out of own resources. Prior to Feb. 1981 : none Mar, 1981 - Aug 25, 1981 : 6 percentage points From Aug. 26, 1981 : 10 percentage points /a For industrial exports (excluding food, beverages and tobacco). Sources: Central Bank; mission estimates. Note: Excluding the effect of the blocking of compensatory balances, which may increase costs by about 25 to 30 percent in nonpreferential credits and 15 to 20 percent in export credits.  - 40 - Improvement of the System of Export Credits 4.32 The current level of subsidies to export credits seems to be justified on grounds of the need for sustaining the momentum of the export drive, the existence of a bias against exports, and the reduced spread between export credits and nonpreferential credits since March 1982. However, in view of the fact that the availability of export credits is probably more important for exporters than the interest rate subsidy, the main objective in this area should be to ensure an adequate priority to the supply of these credits, within the constraints imposed by the monetary policy. To deal with the difficulties currently being faced in the area of export credits, the following recommendations should be considered: (i) investigate the possibilities of enlarging the share of export refinancing in the total credit of the Central Bank; (ii) since the refinancing of export credits will in any case continue to be subject to severe constraints, devise a more equitable allocation mechanism. One way of achieving this would be to lower the rediscountable portion of the credit at the Central Bank. Another way would be to shorten the maturity of credits to achieve a faster turnover; (iii) grant export credits on the basis of value added rather than on the value of exports. In light of the working-capital financing nature of export credits and the principle of equal treatment of imported and domestically produced inputs in a free trade regime, the existing scheme would be appropriate if interest rates were not subsidized. But, given that the margin of interest rate subsidization should in principle be related to value added and that there is extreme scarcity of resources available for export credits, exports with lower value added should benefit less from export credits on favorable terms. At present, exports with less than 20 percent of domestic value added are excluded from the benefits of preferential export credits, but this percentage could be increased to 25 or 30 percent; (iv) develop a medium-term credit facility, to facilitate the evolution of the structure of Turkish exports to have a higher portion of capital goods than at present; (v) establish an export credit insurance scheme to provide exporters with necessary protection against a wide range of risks. The scheme should cover nonpayment of contract arising from commercial and noncommercial causes (i.e. political, catastrophic, and balance of payments risks, which generally are not covered by commercial insurance). As usual in other countries, noncommercial risks should be covered by State guarantees;  - 41 - (vi) ensure equal treatment between direct and indirect exports in providing automatic access to export credits. Currently, only final exporters are eligible for credits. The present system would not promote the backward linkage effect of exports and thus would hamper the country's export and industrialization drive in the long run. Tlerefore, a back-to-back domestic L/C system should be carefully studied for implementation. Through the domestic L/C issuaed by an exporter's bank on the basis of a direct export order, the indirect exporters at all stages can qualify to receive equal access to basic export incentives f'or the part of their value added that is exported. In Korea, this system has been successful in drawing small and medium-size firms into export production and in inducing them to produce goods of international standard. 1/ D. Housing Credits The Availability of Housing Credits 4.33 The housing sector in Turkey has been most severely affected in the last few years by the need for stabilization efforts in the economy. Investment in housing declined from 6.4 percent of GNP in 1979 to 4.4 percent in 1980 and further to 3.9 percent in 1981. The number of dwelling units completed according to the number of occupancy permits issued, however, increased from 124,297 in 1979 to 139,207 in 1980 and then dropped sharply to 113,511 in 1981. Based on the first eight-month figure of 65,319, it appears that less than 100,000 new hames will be coming into use in 1982. As a result, the shortage of affordable housing has been aggravated, especially in the urban areas where population has increased at 4.6 percent per annum (twice as fast as the total population increase of 2.3 percent, or approximately 1 million people per annum during the last decade or so). In view of the fact that the stabilization efforts should and will continue in the next couple of years, the housing crunch is not expected to be significantly alleviated in the immediate future. Nevert'heless, an adequate supply of dwelling units should not be postponed too long in light of the long-term goals of socio-economic development in Turkey. 4.34 The availability of 'housing credits is a crucial factor in the demand and supply of housing units. Middle-income housing, mainly serving the second quintile of the income scale, is especially dependent on concessionary financing from various public sector financial institutions. In Turkey housing credit facilities are fairly recent and have not fully developed into major components of the domestic banking system. Throughout the 1970s, the share of housing credit in total deposit money banks' credits averaged about 2 percent. Housing credits financed only about 7 percent of housing investment. The limited role of institutional financing for housing in Ttrkey contrasts with the experience of other countries with lower per capita 1/ For detailed information on how the Korean system works, see Yung W. Rhee "Administrative Arrangements for Korea Export Promotion." (Washington, D.C.: The World Bank, 1981).  - 42 - incomes. For example, in Tunisia the share of loans for housing averaged about 11 percent of total housing investment during the period 1975-80; in Thailand institutional financing accounted for 33 percent of total housing investment in 1981; and in the Philippines 26 percent of housing investment was financed by various financial institutions in 1977. The Housing Credit Scheme 4.35 Households in Turkey may get loans for dwellings from several sources: the Real Estate Credit Bank (TEKB), the Social Insurance Organizations for the armed forces and the self-employed, the Istanbul Security Fund and finally the building contractors. Among these institutions TEKB stands out for both the number of beneficiaries and the amount of loans involved. TEKB alone accounts for around 95 percent of total institutional financing. Other institutions provide loans only to their members. 4.36 The interest rate charged by TEKB varies between 12 percent and 22 percent, with a maturity ranging between 10 years and 20 years depending on the type of credit scheme. For the bulk of housing credits the rate is typically 16 percent, with a maturity of 15 years. TEKB does not charge any commission for housing credits, and it receives from the IRRF an interest subsidy of 14 percent. Therefore, the final borrower cost of a typical housing loan is only 18.0 percent as opposed to around 50 percent for nonpreferential medium-term credits. 4.37 The credit amount of housing financing is normally three times the contribution of own savings by the borrower, with an upper limit of TL 1,500,000. The waiting period for the loan is one year. Since the average cost of middle-income housing normally exceeds TL 2.0 million, the scheme effectively does not finance more than about half the cost of housing. However, under a special scheme, emigrant workers can borrow up to TL 3 million; the waiting period for such loans is only six months. Currently the total amount of credit involved under this scheme is meagre, representing far less than 1 percent of total housing credits extended by TEKB. 4.38 Besides providing housing credits out of its own resources, TEKB plays an intermediary role in two important schemes for which the Government channels funds out of the budget: (i) credits extended under the auspi.ces of the Natural Calamities Law; and (ii) credits out of the Squatter Housing Fund. The size of these credits has exceeded over 40 percent of total TEKB housing loans in the past five years. Improvements of the System of Housing Credits 4.39 In the context of the scarcity of long-term capital in Turkey, housing finance has experienced particularly severe problems. Primarily because of uncertainty about future inflation, and in part because of the lack of a liquidity mechanism for long-term paper, long-term resource mobilization has virtually come to a halt in recent years. As a result, TEKB has relied heavily on retained earnings and transfers from the Government budget. On a limited scale, TEKB has floated bonds with a relatively short maturity ranging from one to five years to relieve resource constraints in making housing loans. A shortage of resources has persisted, however, and will remain the core of housing finance problems for Turkey in the foreseeable future.  - 43 - 4.40 The limited supply ofE housing credits has created excess demand, which is aggravated by heavily subsidized interest rates. The interest subsidies are quite high, but comparatively few borrowers benefit from them. This is undesirable for reasons of equity, especially in the face of the scarcity of resources for subsiidization. To resolve these housing finance issues in Turkey, the following recommendations should be consideredt (i) the interest rate on housing credits should be increased substantially, al: least to a moderately positive level in real terms. The resuLting risks on the user side arising from higher interest rates should be reduced by adopting a floating mortgage interest rate scheme and by the partial refinancing of interest payments, in accordance with the recommendations in paras. 4.63 to 4.72, 4.82 and 4.83; (ii) in an environmeni: with a high and variable inflation, variable rate bonds (described in Chapter 7) may be a viable financial instrument in mobilizing medium and long-term resources; and (iii) the average size of housing loans now is far from adequate to cover the cost of housing. As the availability of resources improves, the upper limit of housing credits should be gradually increased. E. Medium and Long-Term Credits Introduction 4.41 Medium and long-term bank credits are granted both by deposit money banks and by development and investment banks. Deposit money banks are required by the existing regulations to place 20 percent of their deposits in medium and long-term credits. Investment and development banks concentrate their activity practically enitirely in those credits. 4.42 At the end of 1981, about 60 percent of medium and long-term credits outstanding had been distributed by deposit money banks. It seems, however, that most of the banks were fulfilling the portfolio requirement concerning medium and long-term credits lby extending those credits mainly to enterprises belonging to their own economic group. Although only about one-fifth of total credits granted by the banks has been formally classified as having medium and long-term maturity, a large proportion of the short-term credits is successively renewed and thus corresponds in fact to medium and long-term credits. 4.43 Five banks are classified as investment and development banks. At the end of 1981 the State Investment Bank (DYB) which concentrates its activity in financing investment of State Economic Enterprises (SEEs) accounted for 58 percent of the total assets of that group of banks. The private manufacturing sector was served principally by Turkiye Sinai Kalkinma Bankasi (TSKB), Sinai Yatirim ve Kredi (SYKB), and Devlet Sanayi ve Isci Yatirim Bankasi (Desiyab). The Halk Bank also finances small-scale industry,  - 44 - principally small tradesmen and artisans (often organized in cooperatives), through a comparativly large network of branches. In this report, however, the Halk Bank is not considered as a development bank. The share of credits extended by investment and development banks in total credits, excluding credits to the administrative public sector, has been declining rapidly--from 25 percent in 1978 to 21 percent in 1979, to 16 percent in 1980, and to 14 percent in 1981. This decline may reflect in part the crisis in investment, but it is mainly explained by the difficulties of the development and investment banks in mobilizing domestic currency resources. The State Investment Bank 4.44 The State Investment Bank (DYB) concentrates its activity only in granting long-term credits to SEEs. Although the DYB is independent in its own decisions about the quality of the projects to be financed, such projects must necessarily be included in the annual program approved by the Government. The problem of submitting the public sector to the discipline of economic and financial criteria normally associated with development banking continues to be an important one. The World Bank has maintained a dialogue with the Government on this aspect, in the context of its Structural Adjustment Lending (SAL) program. Under the State Industrial Enterprise Project (Loan No.1998-TU) the Government is to carry out a study of DYB's future role. The study is to address, in particular, DYB's role in the selection, financing and supervision of investment projects of SEEs, DYB's relationship with the Treasury and SPO, and its organizational, financial and staffing requirements to carry out its responsibilities. Commencement of the study, which was held up for the passage of the SEE decree, is expected shortly. 4.45 The maturities of the credits of the DYB are usually about 10 years for extensions of existing projects and around 18 years for new projects, but in some cases they may go up to 30 years. The loans in Turkish Lira are made in annual tranches. The interest rate is fixed for the entire life of each tranche and has recently been 21.5 percent. Loans in foreign currency carry interest rates depending on the interest rate of the source of the funds. 4.46 Until a few years ago, the resources of the DYB were obtained mainly from the issue of bonds with maturities of 20 years and a fixed interes.t rate of 20 percent, which were subscribed by Social Security Institutions and by Retirement and Pension Funds. With the rise of inflation, the interest rate of such bonds became highly negative in real terms and the financial situation of the Social Security Institutions deteriorated sharply. These interest rates were raised in 1981 to 30 percent, on the basis of subsidies granted by the Treasury, but they continued to be insufficient. The DYB has also used some foreign credits, mainly from the World Bank and the European Investment Bank. The exchange risks are borne in full by the SEEs to which the funds are lent. 4.47 The main problem of the DYB is the scarcity of additional financial resources to satisfy the demand for new investment credits from the SEEs. The Social Security System and the Pension and Retirement Funds, after the losses they have suffered, are not interested and have not the capability of increasing their portfolio of bonds from DYB. At the same time, the access to foreign credits has been limited. To provide DYB with funds for new operations, the Treasury has granted it two loans of TL 40 billion each, in 1981 and 1982 respectively, with maturities of 30 years and an interest rate  - 45 - of 10 percent. This reflected the Government's policy of channeling all investment funds to SEEs through DYB, rather than lending them directly, in order to improve project evaluation and execution. 4.48 The highly subsidized interest rates of DYB resources in Turkish Lira represent quite a heavy burden for the budget and for the financial system. To reduce the existing disparities between the conditions of access of the SEEs and of private enterprises to medium-term credits, it would be advisable to establish for DYB the same conditions for the mobilization of domestic or foreign currency resources that apply for the other investment and development banks. That would imply that the special privileges of DYB in the access to low interest loans from the Treasury and from the Social Security Institutions should be eliminated. The Government could, however, assist DYB by providing guarantees to the bond issues it may place in the market, if this could contribute to lower significantly the interest rates of these bonds. The same type of assistance should, however, also be given to other investment and development banks as proposed Fbelow in para. 4.60. 4.49 A corollary of this recommendation would be that the loans of DYE to SEEs should be sUDbject to the same rules as the loans of the private investment banks. They should be based on the same reference interest rate described below in paras. 4.68 to 4.72 and they should receive the same subsidies from the IRRF. TSKB and SYKB 4.50 TSKB is by far the largest of the development banks serving the private industrial sector. It accounts for about two-thirds of the total assets of investment and development banks, excluding DYB. TSKB's assistaLnce has traditionally been financing the import component of investment financing of the private manufacturing sector. It is considered primarily as a foreign exchange investment bank. SYKB has mainly given assistance to the same sector in the form of medium-term local currency loans. During the last three years, however, SYKB has started to extend long-term financing, including loans in foreign exchange, in direct competition with TSKB and foreign currency lozns now comprise 80 percent of its portfolio. Desiyab is a more recent institution whose primary role has been in channeling resources of workers abroad into productive investments in Turkey. 4.51 Development banks are not allowed to compete with deposit money banks for deposits. They can raise local resources only through their equity, bonds, and limited rediscount facilities with the Central Bank. This limitation has been particularly binding in the recent past. The foreign exchange resources of the development banks, the bulk of which has been granted by the World Bank Group, are of a long-term nature and, therefore, cannot be readily used for short-term and working capital financing. Besiides, their clients borrowing foreign loans have to bear the foreign exchange risk. This risk has been extremely high over the past three years, and despite the fact that it is now much lower, the risk has deterred many potential borrowers. 4.52 The role of TSKB and SYKB in the economy should be strengthened. Their cost of intermediation is considerably lower than that of commercia1l banks. The appraisal and technical assistance from development banks are recognized as an important guarantee of the quality of the project and its implementation. In addition, few commercial banks can meet the appraisal criteria of major international lenders.  - 46 - 4.53 The basic requirement for strengthening the role of TSKB and SYKB will be to provide them with a better possibility of access to domestic currency resources, a point elaborated on in paras. 4.56 to 4.61. Moreover TSKB, as the largest and best equipped of the private development banks, should extend the scope of its activities--by going beyond the appraisal of investment projects and of the financial soundness of borrowing firms and by becoming more involved in actions of technical assistance and advisory services, during and after the appraisal of projects. It might also be desirable to extend the activities of TSKB to nonmanufacturing sectors related to industry, such as agro-industry, mining, shipping and leasing. 4.54 Because there might be a danger in giving TSKB a kind of monopoly in development banking, efforts must be continued to ensure a more rapid growth of SYKB. The question of upgrading the Halk Bank to the level of a development bank should also be addressed. Main Problems in Medium and Long-Term Financing 4.55 The main problems to be faced in the area of medium and long-term credits are the following: (i) mobilization of resources in TL for the investment and development banks; (ii) use of fixed or floating interest rates in medium and long-term loans; (iii) subsidies to medium and long-term credits; (iv) exchange risks in credits expressed in foreign currency; (v) liquidity problems created for the borrowers by high nominal interest rates in medium and long-term credits. The analysis of these problems and the recommendations for their solution are presented in the following sections. Mobilization of Resources in Domestic Currency for the Investment and Development Banks 4.56 As mentioned above, the development and investment banks have no possibility of competing with deposit money banks in attracting deposits from companies and households. In addition, they cannot rely as much as in the past on funds provided by the Social Security Institutions, the Treasury or the Central Bank. 4.57 The creation of an interbank market might improve the possibility for the development banks to obtain short-term resources to solve their occasional liquidity problems. It does not seem likely, however, that the access to an interbank market would be sufficient to solve the difficulties arising out of the lack of domestic currency resources for those banks. 4.58 One possible way for the investment and development banks to mobilize more domestic currency resources would be through the issue of bonds to be placed in the market. Such bond issue could be an important element in the revitalization of the capital market. 4.59 Ideally, the bonds should have comparatively long maturities, but that would not be feasible without a system of floating interest rates of the type mentioned in Chapter 7. If the issue of floating interest rate bonds does not prove to be successful, the investment and development banks might try to issue bonds with fixed interest rates and comparatively short maturities (one to four years) and make the term transformation of the resources thus obtained into medium and long-term loans.  - 47 - 4.60 Under current market conditions, the funds collected from the issue of bonds, either with floating or fixed interest rates, would be very expensive. The Government might consider providing guarantees to such bonds, if that could significantly retduce their interest rates. In any case, the cost of such bonds would tend to exceed the interest rate of time deposits and they could not be mixed with lower cost funds from sight deposits, as in the case of commercial banks. However, given the importance of the role of investment and development banks in the field of medium and long-term lending, which cannot be easily filled by other institutions, the lending interest rates of such banks should not: be higher than those of commercial banks. It is therefore proposed below (para. 4.70) that the investment and development banks should receive a subsidy from the IRRF to cover the difference between the average cost of their resources in TL and that of the commercial banks, taking into account the differences in intermediation costs. 4.61 Credits from the Central Bank can also be a source of funds in TL for investment and development banks. The relative importance of the contribution of such credits declined sharply in recent years. The liabilities of the investment and development banks to the Central Bank, consisting mostly of liabilities of DYB, dropped from 38 percent of the total domestic currency resources of these banks in 15177 to about 25 percent in 1981. The objectives of economic stablization require the Central Bank to keep the expansion of its domestic assets under tight control. A rapid growth of Central Bank rediscounts of medium and long-term credits cannot, therefore, be realistically envisaged. If t:he ratio of reserve requirements was reduced (as mentioned in paras. 3.18 and 3.19), the rediscounts of medium and long-term credits would become even more difficult. 4.62 Some improvements in this area might, however, be envisaged. First, the possibilities of enlarging the share of rediscounts of medium and long-term credits in the total. credit of the Central Bank should be analyzed. Although some types of rediscounts (for instance, export credits) deserve a high priority, the Central Bank would have more possibilities of refinancing medium and long-term credits if more progress were achieved in reducing the Treasury's borrowing requirements from the Central Bank and if credits to TMO, to the State Monoplies and to agricultural sales cooperatives became less dependent on rediscounts. Second, the credits of the Central Bank to investment and development banks should be equitably distributed among those banks and should not be concentrated to such a large extent in DYB. Third, the Central Bank rediscounts of medium and long-term credits by deposit money banks should depend more strictly upon the quality of the evaluation of the projects being financed. At present, the Central Bank rediscounts only medium and long-term credits linked t-o projects which have received an investment incentive certificate from the State Planning Office. Given the limits on Central Bank credits, it may be necessary to establish more exacting standards as regards the quality of the projects which may be refinanced by the Central Bank. One way of improving the quality of the evaluation of investment projects receiving medium and long-term credits from commercial banks would be to give priority in rediscounl:ing those credits to operations involving cofinancing by development and investment banks, which have good expertise on project evaluation, or international financial institutions.  - 48 - Fixed or Floating Interest Rates in Medium and Long-Term Loans 4.63 When inflation accelerated in Turkey, the real interest costs of medium and long-term loans that had been granted at fixed interest rates became strongly negative in real terms. The benefits thus obtained by the borrowers were determined on a purely arbitrary basis by the changes in the inflation rate. If inflation continues to decline gradually in the coming years, there is the opposite risk that medium and long-term loans granted now at fixed interest rates will become too expensive in real terms. For instance, if inflation comes down to 10 percent within the next few years, even the loans of DYB granted to SEEs at the highly subsidized interest rate of 21.5 percent would become too costly. 4.64 Because of the uncertainties associated with high and unpredictable inflation, the widespread use of floating interest rates on medium and long-term loans is recommended. Such rates are now applied to a certain extent, but there is a great diversity of regimes. In some cases the interest rates are fixed; on most of the domestic currency loans of the private development banks they are adjusted every six months in line with changes in thg average cost of liabilities of those banks; and on medium and long-term credits granted by commercial banks the interest rates have been adjusted in accordance with changes in the rediscount rates of the Central Bank or changes in the limits fixed by the authorities for the interest rates of preferential credits. As a rule, the nominal interest rates of medium and long-term credits, even when they are subject to some adjustments, have not followed the fluctuations in the rate of inflation or in short-term interest rates. 4.65 In such conditions, it is clearly desirable to introduce a more clearly defined system of floating interest rates for medium and long-term loans. For that purpose, it would be necessary to define a reference interest rate. The interest rate on medium and long-term loans would then be equal to the reference rate plus a positive or negative spread, resulting from market conditions or fixed by the authorities. IThe interest rate would be adjusted periodically--say, every six months--in line with changes in the reference interest rate. The main difficulty is to choose a reliable reference interest rate, for in Turkey there is hardly a rate that satisfactorily reflects a situation of equilibrium in the financial markets. 4.66 Several solutions can be envisaged as far as the choice of a reference interest rate is concerned. It would in principle be possible to use as reference interest rate: (i) the annual rate of increase of a price index (for instance, the wholesale price index); (ii) a reference rate fixed by the Central Bank (for instance, the rediscount rate for general credits); (iii) the interest rate announced by leading banks for time deposits (for instance, on six-month time deposits); and (iv) an index of the cost of resources of the financial institutions. 4.67 The last of these solutions is the preferred one. The system of indexation to the price level is only feasible if the floating interest rate credits are matched in the balance sheets of the lending institutions by equal amounts of liabilities (time deposits or bonds) with interest rates linked to the same price index. The introduction of such liabilities, however, might create risks of disturbances in the market--resulting, for instance, in sudden  - 49 - shifts of funds between non-indexed financial instruments and those whose interest rates would be based on the price level. The use of a reference interest rate fixed by the CerLtral Bank would also not be a satisfactory solution. The experience of the recent past shows that changes in the rediscount rates of the Central Bank have not closely followed the fluctuations of market interest rates. Finally, the use of the interest rate announced by the leading banks for time deposits as a reference rate would lead, in the present conditiorLs, to excessively high lending rates in real terms. 4.68 The use recommended here of the average cost of resources of the lending institutions as the reference interest rate for medium and long-term loans is a solution already adopted by investment and development banks. That solution offers some advantages over the use of the interest rate on time deposits as the reference rate. First, it would lead to reducing the reference rate because of the moderating influence of the lower interests paid on sight deposits. Second, it: would be better adapted to the requirements of financial equilibrium of the banking system, because it would reflect the consequences upon the cost of resources of the changing composition of sight deposits, time deposits and other liabilities. 4.69 An important difficulty is that the average cost of resources is not the same for the different types of banks. That cost will tend to be especially low in the Agricull:ural Bank that receives large deposits from the Treasury at zero interest rates. It will also, as a rule, be lower in commercial banks than in investment and development banks because of the commercial banks' access to lower cost deposits. In spite of these differences in the cost of resources, the reference rate for medium and long-term loans of investment and development banks, corrected for the differences in operation costs, should be the same as for similar loans of commercial banks. In fact, it will be desirable to avoid the distortion of the present situation, under which the interest rates of loans from TSKB and from SYKB are different from those of commercial banks. 4.70 The common reference rate should thus be the average cost of resources for the commercial 'banks. Since the average cost of resources plus the intermediation margin will tend to be higher in investment and development banks, those banks should receive a subsidy from the IRRF to offset the part of the difference that is not compensated by their lower operation costs. This means that if at a certain moment the average cost of resources is 30 percent for commercial banks and 40 percent for investment and development banks, and if the average operation costs of the former exceed those of the latter by 6 percentage points, investment and development banks would receive a subsidy of 4 percent 1/ on their medium and long-term loans. At presen,t, the IRRF is already providing subsidies to banks engaged in preferential credits, but on different grounds. With the proposed solution, both commercial banks and development and investment banks, would be able to charge the same interest rates. The subsidy to investment and development banks: would be justified by the higher cost of their resources and especially by the role of those banks in project selection and evaluation and in improving the efficiency of the allocation of investment credits. 1/ 40%-30%-6%=4%  - 50 - 4.71 The implementation of this solution would require the establishment of a formula for the determination of the average cost of resources for the group of commercial banks and for the group of investment and development banks. The weights of the different types of deposits and of other liabilities in the formula should be revised periodically on the basis of the available statistics. The average cost of resources would be determined every six months on the basis of data collected by the Central Bank. 4.72 If the financial market were not fragmented the choice of a reference interest rate for medium and long-term credits would be much easier. In the absence of legal barriers to competition, the financial institutions of different types would have access to the same pool of financial resources and there would be a single market rate that would provide a good basis for the indexation of the floating interest rates of all medium and long-term loans. The solution analyzed above, with special subsidies to compensate the investment and development banks for the higher cost of resources, although less simple than would be possible if the market was fully integrated, would correct some of the distortions of the existing systems. Subsidization of the Interest Rates of Medium and Long-Term Credits 4.73 Most of the medium and long-term credits benefit from highly subsidized interest rates. As shown in Table 3.1, the cost of medium and long-term credits in domestic currency goes from about 21 percent in loans to finance export-oriented investments in underdeveloped regions, to about 40 percent in general investment loans, and to between 40 percent to 50 percent in working capital loans. Differences of that size are hard to justify and they should be considerably narrowed. 4.74 In present conditions, the average cost of resources of the commercial banks, including a reasonable margin for their operation costs is close to 35 percent. The recommendation presented in the previous paragraphs will therefore lead to nominal interest rates in medium and long-term loans exceeding 35 percent. In real terms such rates would be too high, given the present expected inflation rate. This may be unavoidable if the overall scarcity of credit maintains the deposit interest rates at very high levels. The cost paid by the borrowers can, however, be reduced to some extent by means of subsidies such as those already granted by the IRRF. 4.75 At present the subsidies of the IRRF in favor of borrowers of medium and long-term loans range from 10 percent to 35 percent of the interest charged by the banks. In view of the scarcity of resources for subsidization and of the need to alleviate the burden imposed by subsidies on non-selective credits, the level of subsidization should be reduced and the disparities in the subsidies granted to the different types of medium and long-term loans should be considerably narrowed. 4.76 To illustrate how the system would work, suppose that the average cost of resources of the commercial banks and of the investment and development banks in a given semester were 30 percent and 40 percent, respectively. The reference rate would then be 30 percent. If the average intermediation costs of development banks were 6 percent below those of commercial banks, the former would receive a subsidy of 4 percentage points  - 51 - from the IRRF. Suppose now that a given loan were granted at the reference rate plus a spread of 6 perceni: and that, given the nature of the project to be financed, the loan would be entitled to receive a subsidy of 20 percent from the IRRF. If no transactions tax were paid, the cost to the borrower would be; Interest rate charged by the bank 36.0 percent Contribution from the IRRF -7.2 percent 28.8 percent A similar calculation shows that if in the next semester the reference rate would drop to 26 percent, the cost to the borrower would decline to 25.6 percent. The Coverage of Exchange Risks in Medium and Long-Term Loans 4.77 The exchange risks of foreign currency loans granted by TSKB, SYKB and other institutions have been borne in most cases by the borrowers. In consequence of the sharp devaluations of the Turkish Lira that took place in 1980 and 1981, many borrowers suffered enormous losses in their foreign currency loans, and some of them had to be assisted by the State. In view of such losses, borrowers have been extremely reluctant in recent months to accept the exchange risk of foreign currency loans. In consequence, TSKB and other institutions have faced great difficulties in disbursing the foreign currency resources provided by the World Bank, the European Investment Bank, and other financial institutions. 4.78 In the present circumstances, the reluctance of the borrowers to accept foreign currency loans is not well founded. The exchange rate of the Turkish Lira is now much closer to equilibrium than in the past and with the present policy of gradual depreciation, the risks of sudden and substantial devaluations of the currency are now much lower. If the expected rate of devaluation of the Turkish Lira against a given foreign currency during a year is e, and if the interest rate for loans in that currency is f, the total cost of such loans expressed in TL is (1 + e)(l + f). At present the domestic interest rate d is so high that it can be taken for granted that (1 + e)(l + f) is lower than (1 + d). 4.79 Borrowers should, therefore, prefer loans in foreign currency under present circumstances. However, given their psychological hesitation resulting from recent experiences and given the volatility of the exchange rates of some key currencies, it would be advantageous to introduce a system for the coverage of foreign exchange risks. A special Exchange Risk Fund should thus be created in the Central Bank. The system should, however, lbe established in such a way that the Exchange Risk Fund would not be likely to run into losses, as happened in the late 1970s with the Central Bank's exchange rate guarantees on convertible Turkish Lira deposits. 4.80 The borrowers seeking the coverage of exchange risks on medium and long-term loans denominated irn foreign currencies should, therefore, pay commissions to the Exchange Risk Fund high enough to ensure that the  - 52 -- probability that the Fund would suffer significant losses would be small. An appropriate solution would be to collect as commission the difference between the reference interest rate used in medium and long-term loans in domestic currency (plus the average spread for such loans) and the interest rates paid on foreign currency loans. The borrowers should receive for such loans the same subsidies from the IRRF as those which are granted to similar medium and long-term credits in Turkish Lira. The borrowers would, therefore, pay the same for loans in foreign currency as for loans in domestic currency. 4.81 If the exchange rates of the Turkish Lira and of other currencies were changing in accordance with their purchasing power parity, and if the real rates of interest were the same in Turkey and in other countries, the revenues of the Exchange Risk Fund from commissions would tend to equal its expenses. In practice, these conditions are not fulfilled. But if the exchange rate policy of Turkey continues to be based on purchasing power parity guidelines, as it has happened recently, and if the interest rates remain higher in Turkey than in other countries, it is more likely that the Exchange Risk Fund would accumulate profits rather than suffer losses. The Liquidity Problems Created by High Nominal Interest Rates on Medium and Long-Term Loans 4.82 Few borrowers will be able to generate enough cash flow to pay interest rates on medium and long-term loans of the order of 30 to 50 percent, such as those which have recently prevailed in Turkey. One solution would be to refinance part of the required interest payments. If the refinancing does not exceed the component of the interest rate that compensates inflation, the capitalization of part of the interest payments will not imply the growth of the real value of the debt outstanding. 4.83 In such conditions, one way of assisting the borrowers of medium and long-term loans would be to capitalize part of the interest payments corresponding, for instance, to the inflation of the previous year less a safety margin of say, 5 percentage points, to guard against over-compensation for inflation. Annex A presents an example that illustrates how the proposed system would work. For medium-term loans, approximately the same results would be obtained if in the beginning the Bank increased the amount of the loan by, say, 20 percent and retained that increase, which would later be used in payment of part of the interest rates of the first two years. That would be an easier solution, but for long-term loans it would not satisfactorily replace the system of capitalization of interest exemplified in Annex A.  - 53 - CHAPTER V - CORPORATE FINANCE A. The Financial Situation of Industrial Firms 5.1 Prior to the initiation of the stabilization measures in January 1980, industry suffered from a shortage of foreign exchange for imported raw materials, and low capacity utilization associated with civil unrest and strikes. Accelerating inflation, however, fueled local demand so that the private sector on the whole was profitable. 5.2 The 1980 stabilization measures involved a substantial devaluation of the Turkish Lira, increased interest rates and contracted domestic demand. These measures decreased profitability and caused liquidity problems in the private industrial sector. Industrial production in 1981 was at the same level as in 1977, and it increased only marginally in 1982. Production for foreign markets has thus become more attractive, and that for the domestic market less so. But many firms do not have the appropriate product mix, machinery, or knowledge to be able to export. Other industrial problems have to do with location in relation to markets, with excessive capital equipment purchased when interest rates were low, and with labor laws and practices; that make it difficult for firms to lay off redundant labor and make efficient: use of the work force. 5.3 These "real" problens have been compounded by financial difficulties. Stagnant output, redundant labor, excess equipment, and the like have resulted in low operating profits, and for some firms, operatirng losses. High rates of inflation have made it necessary to increase working capital in nominal terms to maintain the level of output. Also, interest: costs which have risen substantially in the last two years, must be subtracted from operating profits, and this has brought net losses even to firms that had operating profits. Furthermore, those firms with debt in foreign currencies have had to write up these debts severalfold as the currency was devaluecl in recent years. By 1982 a substantial number of firms in Turkey could not pay the interest due on their loans, let alone cover principal repayments. 5.4 Because aggregate data on the operations and financial structure of the sector were not available, a data base consisting of the annual accoxnts of 127 firms for the four years 1978-81 was established for the purpose of quantifying recent operating trends and financial structure of the sector. 5.5 During the four-year period, the combined net sales of the 127 fEirms increased substantially in nominal terms but only marginally in real terns (as deflated by the wholesale price index). Profit before interest and corporation tax, as a percentage of net sales, fell from 20 percent in 1978 to 15 percent in 1981. The rati.o of net after tax profit to net sales, however, fell more dramatically duringt the period from 9.3 percent-to 0.5 percent. The major factors in the decrease of the percentage of profitability on net sales were substantial increases in the percentage cost of raw materials (11 percent), interest charges (4.5 percent), and other expenses, that wesre partially offset by reductions in the percentage cost of labor, depreciation, extraordinary losses and tax (Table 14 of Statistical Annex). Because the increase in sales in real terms was marginal during the period, these percentage declines in profil: closely represent the fall in profitability in real terms for the group of iirms during the period.  - 54 - 5.6 Debt service coverage as measured by the ratio between profit before interest plus depreciation and the debt service (interest plus principal) declined rapidly, falling from 2.5 in 1979 to 1.3 in 1981. With the continued decline in profitability in 1982, combined with increasing debt service because of higher interest rates and revaluation of foreign exchange debts, average debt service coverage is estimated to have decreased below 1.0 in 1982. In these circumstances, a considerable number of firms were unable to meet their debt service. The substantial growth in arrears of commercial and development banks during 1982 confirms this trend. 5.7 Total assets of the sample of firms increased in nominal lira terms by 255 percent during the period 1978-81. The current ratio (current assets divided by current liabilities) remained above 1:1 during the period, but declined from 1.17:1 to 1.04:1. The debt-equity ratio increased substantially from 3.2:1 at the end of 1978 to 5.7:1 in December 1981. 5.8 The debt-equity ratio is so high because in Turkey the market for equities is inactive and because in the last two years, with low or negative profits, firms could not finance themselves through retained earnings. Hence, to finance their investments--and in recent years to pay the high interest on their outstanding debt--the industrial firms have resorted to borrowing. This has led to the high debt-equity ratios noted above. 5.9 However, the information derived from the firms' historical balance sheets does not adequately reflect the economic situation because it is not adjusted for the effects of inflation on the value of corporate profits and assets. Adjusting for inflation shows that the operating profits of the firms, calculated on the basis of widely used accounting standards 1/, have been over 30 percent lower than the operating profits shown in historical accounting. However, since the middle 1970s the equity position of firms on the basis of revalued assets has not deteriorated but rather improved substantially. If the fixed assets of the 127 companies of the sample were revalued according to the principles of current cost accounting, the average debt-equity ratio would improve from 1:1.33 in 1978 to 1:1.70 in 1981 (Table 17 of the Statistical Annex). During the years from 1977 to 1980, in which interest rates were low, inflation generated for the firms unrealized nonoperating profits in the form of debt depreciation, which are not adequately reflected in the firms' income statements or balance sheets. A simulation on how the Turkish inflation affected the position of the typical firm is presented in Annex B. 5.10 The impact of inflation on the corporate accounts is reflected in the aggregate financial statistics. As noted in Chapter 1, the financial system of Turkey, which was relatively shallow in 1977, declined rapidly in size during the years of high inflation. The corollary to that decline was a reduction in corporate debt in real terms. Between 1977 and 1980, the real value of aggregate advances from the financial system to private enterprises and households declined from TL 147 billion in 1975 prices to TL 94 billion. 1/ Statement of Standard Accounting Practice No. 16, March 1980 and Guidance Notes on SSAP-UK Accounting Standards Committee.  - 55 - Looking at a slightly different aggregate--advances to the industrial sector, including public enterprises--between 1977 and 1980 such advances declineld from 109 percent of industrial, value added to 40 percent (see Table 5.1). 5.11 The average debt-equity ratio of 1:1.70 after revaluation at the end of 1981, which as mentioned above was based on a sample of 127 industrial firms, was not out of line with those in other countries with similar levels of inflation. Despite the deporessed conditions in which the corporate sector had been operating in Turkey during the previous years, the balance sheet position of the firms had not deteriorated on average, and assets were more than sufficient to cover the debts. However, the conditions prevailing in 1982 were very different from those in earlier years. Interest rates on loans had become significantly positive in real terms. Under such conditions leveraging works in reverse, losses are real not nominal, and the firms' debts increase rapidly, as is reflected in the aggregate statistics showing financial deepening in the past two years. With low operating profits and high real interest rates, the debt-equity ratio of the firms was undoubtedly deteriorating substantially in 1982. Table 5.1: PRIVATE ENTERPRISE AND INDUSTRIAL SECTOR DEBTS Item 1975 1976 1977 1978 1979 1980 1981 1982 (T_Lbillions7T Private sector debt to the Banking System Current prices 123 165 210 268 396 689 1231 1757 /a 1975 prices /b 123 142 147 122 110 94 121 136 Industrial debt to the financing system Current prices 90 132 172 212 272 415 638 1975 prices 90 114 120 97 76 56 63 Industrial debt Industrial value added (percent) 97 115 109 78 57 40 40 /a Estimate /b Deflated by the wholesale price index  - 56 - B. The Liquidity Squeeze in 1982 5.12 Although excessive corporate debt is not yet a problem in Turkey, in 1982 a liquidity crisis began to be felt in the corporate sector. The need for firms to refinance both interest payments and maturing debt, combined with restrictions on domestic credit expansion, drove interest rates on general credits to unprecedented levels. It may seem paradoxical that in Turkey the "liquidity squeeze" did not appear in the years 1977-1980, in which the real value of credit to the private sector was declining, but rather in the period 1981-82, in which the real value of credits to the private sector was expanding rapidly. 5.13 It appears that two conditions are responsible for the liquidity squeeze in Turkey. The first is that industrial firms' operating costs and increased working capital in nominal terms plus debt service, dividends, and tax payments exceeded their cash income. 'This condition seems to have existed for many firms for much of the last five years, but it did not create difficulties as long as credit was increasing rapidly in nominal terms. The second condition, which turned the above problem into a liquidity crisis in 1982, was the limited amount of new credit the financial institutions had available to cover the cash shortfall of the firms. 5.14 The first condition, cash payments in excess of cash income, is a condition commonly found in a situation of high inflation and should not be taken to imply that firms are unprofitable in an economic sense. The condition arises because the interest due on outstanding debt may exceed operating income. The apparent losses are created because of inappropriate adjustments for inflation in the firms' current and capital accounts. To correct for the impact of inflation on a firm's profits, one should not only base depreciation on revalued fixed assets and adjust the cost of inventories but also consider the unrealized gains resulting from the revaluation of fixed assets. If the interest payments are set against such unrealized gains, the income of Turkish firms is substantially more favorable at least on a potential basis. When real interest rates are sharply negative, as they were in Turkey during the 1977-80 period, it is clearly profitable for firms to have as much debt as possible. In fact, during the worst recession years, 1979 and 1980, the potential gains from inflation due to the depreciation of nominal debts offset operating losses for many Turkish firms, although those profits were not realized in cash. Interest rates on loans only became positive in 1981 and substantially positive only in 1982; hence, only during this period would they cause the firms to have real losses. 5.15 Even though the firms were profitable in economic terms in the earlier years, cash outlays, including interest payments, exceeded cash incomes. Hence, many firms could meet their interest payments only through additional borrowing. Although credit was declining in real terms, during the 1977-81 period it was expanding very rapidly in nominal terms, as shown in Table 5.1. In 1982 credit to the private sector expanded rapidly in real terms, but less rapidly in nominal terms than in prior years. If the estimated average interest rate paid by the private sector is applied to the average debts outstanding of that sector in each year, it can be concluded, in rough terms, that the interest payments absorbed about 45 percent of credit expansion in 1977 and 1978 and that this percentage then fell to 33 percent in  - 57 - 1979, 1980 and 1981, the years of most rapid expansion in nominal credit. In 1982 the rate of credit expanesion in nominal terms slowed, whereas interest obligations rose sharply. Consequently, in 1982 practically all the expansion of credits to the private sect:or was needed to finance interest payments. This is not meant to imply that firms must borrow to finance all interest payments. This was not true in 1982, nor in prior years. What it does indicate is how little credit was available to finance anything but interest payments in 1982 when comparecl with the earlier years. At the same time, given the increase of the pric:e level, companies would need more nominal credit to finance the same volume of working capital, other things remaining equal. 5.16 The dynamics of inflation and stabilization are complex. During the years of very high inflation in Turkey, nominal credit availability expanded rapidly, but, because prices rose even more rapidly, real credit availability fell. During 1982 the reverse condition held: the growth in nominal credit slowed, but because prices were rising less rapidly, the quantity of real credit rose. During the inflationary years interest rates rose, but they did so less rapidly than prices. Hence, real rates declined. Again it was the reverse during 1982. The two phenomena were directly related: it was the low real interest rates that caused depositors to reduce their real financial asset holdings during inflation; it was the higher rates in 1982 that caused them to increase their financial assets in real terms. 5.17 During the expansionAry phase, firms could increase the real value of their assets and at the same t:ime achieve a reduction of the real level of their liabilities because of t:he benefits they were deriving from negative real interest rates. Those benefits, together with the rapid expansion of nominal credit, made it possible to satisfy the cash flow needs of most firms in spite of the decline in the real value of credit. In 1982 the real interest rates had become extremely high, and they severely affected the profitability of the firms. 'he new loans available were barely sufficient to cover the firms' losses. Thils produced the liquidity squeeze. 5.18 If the recession in Turkey continues, if interest rates remain high, and if the Government continues to restrict nominal credit expansion to control inflation, it is likeLy that the liquidity squeeze in the private sector will deepen in 1983. Only limited funds will be available to finance anything other than the rolling over of existing debts and the interest payments due. Funds to finance new investments will continue to be in short supply. This will be true even though in real terms the financial system may continue to expand rapidly. 5.19 The majority of the iEirms in the private sector had a problem of illiquidity in 1982 but on average the value of their revalued assets still exceeded the value of debts outstanding. However, the persistence of interest rates as high in real terms ais those prevailing in Turkey in 1982 would cause a rapid build-up in debt and a deterioration in the firmst financial structure. 5.20 Even when they suffer from losses caused by high real interest rates, firms often prefer to increase their debts in order to survive rather than to liquidate their activity or to sell their assets and use the proceeds to repay debt. Besides, the economic recession has depressed asset markets; hence,  - 58 - firms would, on liquidation, realize only a fraction of the book value of assets, adjusted for inflation. If a firm survives the recession without selling off assets, and if the markets recover and real interest rates decline, the value of the firm increases. Firms are, therefore, willing to pay high real interest rates, not to finance new investments but to maintain their position in prior investments. C. Assistance to Firms in Difficulties 5.21 The basic contribution to solving the problems created by the liquidity squeeze must come from the reduction of the real rates of interest. A faster expansion of credit supply could lead to that reduction, but, as explained in the final section of Chapter 3, there is the risk that it would be in conflict with the objectives of economic stabilization. To achieve lower real interest rates, several other measures, mentioned in this report, particularly in Chapter 3, must be contemplated. 5.22 In any case, it is probable that the reduction of the real interest rates will be a gradual process that will take time. Meanwhile, the liquidity and solvency problems of many enterprises may become more acute, and the Government may be pressed with growing demands to assist firms in difficulty. Any scheme from the Government to assist industrial firms in difficulty must be carefully controlled and monitored. Already in 1980 a special rediscounting facility of TL 35 billion had been made available for all industrial enterprises having been subject to foreign exchange risk in the period 1978-80. This facility was made available without adequate conditionality regarding the merit and needs of the recipient firms, their economic importance in Turkey's economic recovery, and their viability. 5.23 The needs of enterprises to be rehabilitated are clearly of two kinds: provision of credits to meet working capital needs and equity to strengthen the financial structure. The increases in equity are already being stimulated by the high level of real interest rates. It would, however, be desirable to encourage them further by tax incentives for the reinvestment of profits, the subscription of new equity and the partial conversion of existing liabilities into equity. The resources for credits to meet working capital needs are scarce. If excessive accommodation was tolerated in this area, there would be difficulties in maintaining the credit policy required by the objectives of economic stabilization and the structural adjustment of the economy would be negatively affected. However, if the liquidity crisis in Turkish industry becomes worse, the government will be under great pressure to establish special schemes of assistance to firms in difficulty. The creation of a special Fund for the Rehabilitation of Industry may therefore prove to be useful. 5.24 The creation of such a Fund would introduce a new factor of fragmentation in the Turkish financial market. However, the Government has already intervened in 1982 to assist two comparatively big firms in a desperate financial situation, and has intervened further in 1983. If in the future the authorities feel forced to make similar interventions, it will be preferable to observe a pre-defined set of guidelines rather than to take decisions on an ad hoc basis, according to the special pressures arising from each situation or the particular characteristics of each case. The provision of assistance from the Fund for Rehabilitation of Industry should be subject to strict eligibility criteria and to rigorous performance clauses as explained below.  - 59 - 5.25 In principle, the resources of the Fund could come: (i) from a special rediscounting facility established in the Central Bank; (ii) from budgetary allocations; and (iii) from external credits. The resources provided by the special rediscounting facility in the Central Bank and by budgetary allocations would be severely limited, given the ceilings enforced on the expansion of Central Bank credits and the constraints on the budget. It might therefore be necessary to seek a high proportion of the resources of the Fund from external loans, although this will also be a difficult solution. For these reasons, the size of the Fund should be comparative]Ly small. This would imply that the Fund should be highly selective in its operations of rehabilitation assistance and should restrict the amount of each operation to the minimum which is feasible. 5.26 Since the resources of the Fund would be very scarce compared wiith the probable demand, the types of operations to be financed should be carefully defined. A clear priority should be given to refinancing part of the high nominal interest payments of medium and long-term credits. As explained in paras. 4.82 and 4.83, high nominal interest rates create sevrere liquidity problems for borrowers and should be partially refinanced. The assistance to the servicing of foreign debt of enterprises which have been seriously affected by the depreciation of the TL should also receive a high priority in the operations of the Fund. 5.27 The loans of the Fund should be extended for periods of about 2 to 5 years, depending on the cash flow projections for each firm. The interest rates should be fixed in accordance with the proposals made in Chapter 4 for medium and long-term credits. 5.28 The resources obtained by the Fund from foreign credits should be relent to the assisted enterprises also in foreign currency. A scheme of- exchange risk coverage of the type recommended in paras. 4.77 to 4.81 should, however, be applied to the loans of the Fund in foreign currency. The costs of such loans to the borrowers should be the same as those of domestic currency loans with similar maturities used for similar purposes. 5.29 In order to reduce the risks of inefficiency and wastage, a certain number of principles should be followed: (i) There should be a careful evaluation of the financial situation, of the productive efficiency, and of the economic prospects of each firm to be assisted. Banks and the authorities must accept that firms of dubious economic merit should be liquidated; (ii) The assistance should be made conditional upon commitmentsi from the owners and managers of enterprises to improve their performance. This would imply fixing conditions, for instance, for the provision of additional equity capital, for retaining earnings and for achieving some easily monitorable targetE; regarding costs, sales, profitability and financial structure. The assistance to firms that would not fulfill the agreed conditions should be interrupted, even if the immediate consequence was the bankruptcy of these firms or the transfer of their ownership;  - 60 - (iii) The administration of the assistance programs financed by the Fund should be entrusted to reliable institutions. The use of the expertise and of the operational capacity of the investment and development banks would certainly be very useful. The contribution of some of the leading commercial banks might also be feasible. The problems that might arise from inter-linkage of interests between the bank participating in the assistance program and the firm being assisted should, however, be taken into account; (iv) If a system of assistance to banks with liquidity difficulties was simultaneously in operation, it should be coordinated with the program for the rehabilitation of industrial enterprises. The Government should not double-count by providing funds to the firms as well as to the banks, since funds provided for restructuring industrial firms would be used, to some extent, to repay bank loans. D. Enterprise Accountancy 5.30 The development of accountancy systems in Turkey has not progressed in line with the level of industrialization. This lack of development has been one of the major factors in restricting the expansion of the financial and capital markets to meet the requirements of industry (see Chapter 7). In some of the private Turkish holding companies and larger manufacturing companies, modern financial and management accounting systems have been developed and are operating satisfactorily. These cases, however, are the exception. In private industry in general, accounting is limited to control of assets and liabilities and to the production of annual income statements and balance sheets, mainly to meet the requirements of the tax laws. The information is usually finalized well after the end of the fiscal year. Management accounting is generally limited to ad hoc costing for pricing purposes. 5.31 Most SEEs have financial accounting systems designed to produce financial reports required by Goverrment budgeting regulations. SEEs issue their income statements and balance sheets within three months of the end of their fiscal year. The accounting system is geared to enable review of compliance with government fiscal regulations rather than to provide financial information to analyze meaningfully the operations of the enterprise. Costing and management accounting in SEEs, however, is practically nonexistent at present. 5.32 Companies with a wide shareholder base are required under the Commercial Code to appoint auditors who will inspect the books of accounts, report any irregularity to management, and certify the annual accounts. No professional qualification is required for persons appointed as auditors. These audit requirements fall far short of the generally accepted standards for audit in other countries. The accounting profession is not regulated by law nor organized in an association. There are a very limited number of qualified independent accounting firms whose main work is confined to auditing associated companies of international firms.  - 61 - 5.33 The High Control Board, with about 180 auditors attached to the Prime Minister's Office, is responsible for auditing SEEs. In addition, there are three groups of about 400 auditors within the Ministry of Finance: (i) Finance Inspectors mainly responsible for auditing the accounts of Government departments; (ii) Tax Auditors responsible for auditing tax returns; and (iii) Sworn Bank Examiners responsible for auditing banks. Auditors within the public secitor are competent, but their number is limited; the result is that audits are nlot completed within a reasonable time-frame and that the examination is often, by necessity, cursory. Emphasis is largely on the enterprise's compliance with regulations, and there is not much analysis of the financial management and performance of the enterprise as a whole. 5.34 Major improvements in the accountancy standards, with particular emphasis on auditing and management accounting, are required in the private industrial and commercial sector--to enable the financial sector (and the capital market in particular) Ito meet the sectors' expanding requirements and to provide management information essential for the control and development of enterprises. Management accounting is a major factor in achieving increased productivity, an area in which rapid improvement is required in Turkish industry. Improvements are also required to enable the Government to levy and collect tax on an equitable basis and to expand the tax base by the introduction of new forms of tax, such as the value-added tax. 5.35 Improved accountancy systems, particularly in the field of management accounting, are an integral part of the proposed reorganization of SEEs. They would enable the Government to establish objectives for the SEEs and to monitor their performance. Within each SEE management accounting is essential for the delegation of authority and responsibility within the organization--essential for improving productivity--and for monitoring and rewarding managers' performance. 5.36 As part of the program of developing the financial sector, the authorities should undertake, as a matter of high priority, to strengthen the accounting and auditing standards and practices in private enterprises and SEEs. For that purpose, they should consider the establishment of a commission of experts to review the present position of accountancy in the private and public industrial sector and to make recommendations for the development and rapid expansion of the profession to meet the future requirement of Turkey's economic development. The commission's recommendations should cover: (i) the assessment of the manpower, educational and training requirements for accountants and auditors; (ii) the setting at an early stage, on the basis of available resources, of deadlines for the certification of accounts of various corporate categories; (iii) the establishment of generally accepted accounting and auditing standards; (iv) changes in statutory requirements; and (v) an action plan for the phased development and implementation of modern accountancy procedures. In addition legislation for the organization and control of the accounting profession should be passed as soon as possible. Consideration also should be given to the appointment of technical experts from established accounting institutions to assist the commission.  - 62 - 5.37 The description presented in the first section of this chapter gives an idea of the importance of the distortions in the accounts of enterprises that resulted from the inflation of recent years. Given the present stage of development of the accounting profession in Turkey, it would not be feasible to implement a realistic inflation accounting system. A substantial part of the benefits of such a system may however be secured by the revaluation of fixed assets which is allowed under legislation published recently. According to that legislation, fixed assets and the accumulated depreciation can be revalued on the basis of cofficients that take 1972 as the base year. The surplus on revaluation is not subject to the capital gains tax.  - 63 - CHAPTER VI - STRENGTHENING THE BANKING SECTOR A. Introduction 6.1 Apart from the Central Bank, the Turkish Banking system comprises 47 banks altogether. There are 24 all-purpose commercial banks, 7 foreign banks, 11 specialized public deposits banks owned by the Government and 5 investment and development banks. The whole system is dominated by four banks which account for more than 50 percent of total bank assets. 6.2 The authorities have been preparing a new Banking Law whose publication is expected shortly. 1/ It is expected that law will include provisions regarding the capital structure of the banks, the protection olE deposits, limits on credits tc individual customers, qualifications required from the senior managerial personnel of the banks, auditing requirements, penalties for noncompliance and special rules for local banks. 6.3 The main problems of the Turkish banking sector are the following: li) high costs of intermediation of the banks; (ii) undercapitalization of the banks; (iii) high level of maturity transformation; (iv) increasing share of nonperforming loans; (v) absence of an interbank money market; (vi) liquidity risks in weaker banks; (vii) interlocking ciwnership and flow of funds between banks and corporations; 1/ Since the report was written, the Banking Reform Act has been passed. The key features are: raising the minimum capital requirement for new banks from TL 25 million to TL 1 billion; increasing the controL mechanism over the allocation of bank credits by restricting loans to a single person or ccmpany to a maximum of 10 percent of total paid-up capital and by restricting loans to companies in the bani's group to three times the bank's capital; creating a new deposit insurance fund, admirListered by the Central Bank, to guarantee saviugs deposits in banks; and requiring all loans in excess of TL 100 million to be approved by the Board of Directors. These measures do address some of the issues presented here, but not all. In particular, measures regarding the high intermediation costs, the increasing amount of bad debts, the reliability of the auditing and accounting of banking institutions, supervision of the banking system and penalties for non-compliance with regulations appear insufficient.  - 64 - (viii) shortcomings of the accounting and auditing mechanisms; and (ix) underutilization of supervisory skills. These problems are examined in turn in the following sections. B. High Costs of Intermediation 6.4 The intermediation margin of Turkish banks, including operating costs and profits, corresponded on average to more than 11 percent of their total assets in 1980 and 1981, but that percentage declined to 8.1 percent in 1982 (Table 18 in the Statistical Annex). The comparable ratios for industrialized countries are of the order of 3 percent to 4.5 percent. 6.5 The high intermediation costs of the Turkish banks originate, to a large extent, from the regime of controlled interest rates that was maintained until 1980. The controls were far more effective on deposit than on lending interest rates, since the latter could easily be increased by requiring compensatory balances. The banks were thus able to enlarge their intermediation margins. Because they could not offer higher interest rates to depositors, they were basing the competition with each other mainly on the rapid expansion of their network of branches, which in turn implied the rapid increase of the size of their staff. This behavior has been maintained in recent years. Thus, from 1977 to 1982 the number of bank branches increased 21.8 percent, from 5,238 to 6,347; and the number of bank employees increased 15.7 percent, from 113,200 to 130,900. 6.6 The high intermediation costs of the banking system contribute in large measure to the wide difference between the average interest rates received by depositors and those paid by borrowers. The role of the banking system in the economic development of Turkey will be strongly affected if such a large wedge between the returns to savers and the costs to borrowers persists. This is a structural problem, however, and it would be unrealistic to expect that it may be solved soon. 6.7 In order to reduce their operating costs, the banks must increase their productivity, reduce their staff, close unprofitable branches and cut their overhead expenditures. Those improvements cannot be easily implemented. But the banks would be put under pressure to become progressively more efficient if adequate taxation and regulatory measures were adopted. Such measures might include: (i) introducing a limit to the growth of operating expenditures which can be deducted from total revenues for the purpose of determining the taxable profit; that limit should be set significantly below the rate of expansion of the value of total assets of each bank, (say, 5 percentage points), and should be maintained during a certain number of years; with this system the banks would be penalized by heavier taxation if the proportion of their operating costs in relation to total assets was not reduced gradually;  - 65 - (ii) establishing limits to the number of branches and to the ireal estate assets of each bank, which should be related to its total equity; the Turkish authorities have been considering the introducti.on of such limits which could force the banks to place a larger proportion of their assets in credits producing operating revenues; the limit to the real assets would only be meaningful if such assets were periodically revalued in order to reflect the effects of inflation; (iii) stimulating the competition to the banking sector by revitalizing the bond market on the basis of the recommendations of Chapter 7; since that market can operate with low intermediation costs, it can become a highly effective instrument of competition with the banking system; and (iv) reducing the specialization among the different types of banks; the specialization among banks exists mainly in thie areas of government-owned specialized banks and in investment and development banks; specialization has some advantages, but when it is strictly enforced by legal barriers it may also have important costs; less specialization and more competition among the different types of banks would force the banks that benefit from monopoly positions to become more efficient, would reduce the artificial fragmentation of the market that negatively affects efficiency in the allocation of financial resources, and would create more favorable conditions for the exploitation of economies of scale in banks' operations. C. Undercapitalization 6.8 Banks have already been compelled to increase their capital. It was expected that such increases and the revaluation of fixed assets might place the Turkish banking system in the upper group of capitalized systems in the future. However, by the end of 1982 only 40 percent of Turkish banks had complied with the law requiring equity to be raised. Banks which cannot comply with the minimum capital requirements can elect to remain in business, provided that their scale of operations is curtailed to a level proportionate to their equity. Otherwise, they might be forced to liquidate or to seek mergers with stronger banks. 6.9 The experience with these regulatory measures shows that minimum capital requirements applied across the board are not the best available instrument to ensure the adequate capitalization of banks. Minimum capital requirements should be conceived as a first line of defense against banks' notorious undercapitalizatioin. The requirements should be set at a relatively low level, and adequate capitalization for each bank should be ensured by means of a customized solvency ratio. For instance, risk exposure of individual banks could be better taken into account if banks were to abide by a solvency ratio that related the amount of unimpaired own funds to total assets, with the ratio weighted by category according to the assumed degree of risk.  - 66 - 6.10 There is presently a worldwide tendency to favor such a ratio, and in Turkey it would offer some side benefits. It would give bank supervisors a weapon with which to attack the worrisome problem of interlocking ownership of banks and corporations. If implemented, as in Belgium, for instance, banks would be obliged to increase marginally their capital for loans to, or participations in, a single entity in excess of the general limit set forth in the banking law. This would make it costly for banks to favor borrowers from their own group. Besides, if bank examiners were to be allowed, as mentioned below, to classify bad loans and to deduct them from the banks' unimpaired own funds, the banks would be given a disuasive instrument against the building up of bad loans. 6.11 The customized minimum equity ratio could be set at different levels for commercial banks and for investment and development banks. The assets of investment and development banks tend to be much less liquid than those of commercial banks and their equity participation may involve a considerable degree of risk which it is difficult to measure. For these reasons, the minimum customized ratio of equity to total assets imposed on investment and development banks should be higher than that of commercial banks. In addition, it would be desirable to prevent excessively dominant positions of other banks in the management of private investment and development banks. It might therefore be desirable to introduce a maximum limit to the participation of a single shareholder in the equity of investment and development banks. D. Maturity Transformation 6.12 It is common, in particular in developing countries, for the financial sector to assume a fair amount of maturity transformation by borrowing short and lending on a longer-term basis. There is, however, an element of risk involved in the process. Banks may encounter funding difficulties, especially when the interbank market is nonexistent and when changing interest rates may lead to unforeseen losses. 6.13 Some Turkish banks may have reached the sustainable limit. No precise data are available on this subject but an overall picture may be drawn. Apart from a very limited amount of bonds, most of the resources obtained by Turkish banks until recently had maturities of not more than six months. Graduation of interest rates offered to depositors along the maturity scale was too moderate to combat the strong preference for short-term savings instruments. In addition, holders of deposits of an initial six months' maturity could, in practice, withdraw them without, in most cases, being charged the interest rate penalty provided for in the gentlemen's agreement. The recent changes in interest rate policy reintroduced one-year time deposits. The interest rates on those deposits exceed those for six-month deposits by 5 percentage points. However, on a compounded annual basis, the difference is very small, and it remains to be seen if the share of one-year time deposits becomes substantial. 6.14 On the asset side of the banks' balance sheets, repayment of loans is increasingly assured by the disbursement of new loans, which, in practice, amounts to a lengthening of the maturity structure of bank loan portfolios. Equally, Turkish banks traditionally carry a sizable portfolio of unquoted  - 67 - capital participations, and they have in the recent past increased significantly their fixed assets, mainly because of the aggresive extension of their branch network. Consequently, the amount of long-term investment, including equity participations and fixed assets, exceeds that of the bankcs' owned funds by about 40 percent. 6.15 At present, very little is done about monitoring the degree of transformation effected by banks. The only constraint that binds banks is the liquidity ratio, whereby liquid assets must now cover 10 percent of banks' deposits. Compliance with this ratio was checked until recently by the Central Bank on a monthly basis, and banks were only obliged to respect it at the date of the drawing up of the financial statement. It was well known that some banks ignored the liquidity ratio the rest of the time. An amendment was subsequently introduced to make the ratio compulsory at all times. This is a welcome measure, but to make it effective the sworn Bank Auditors will have to ensure its enforcement. 6.16 In a broader context, some use should be made of the breakdown by maturities of the data on assets and liabilities provided by banks to the Central Bank on a monthly basis, and information gaps should be filled by the Board of Sworn Bank Auditors (SBA) in the course of their examinations. 6.17 In addition, the bankzs' use of funds for fixed assets and capital participations should be limited to the amount of their net worth (as implemented in France for insl:ance). Besides improving the liquidity of banks, such a regulation would also contribute to reduce their operating costs, as suggested in paragraph 6.7. E. Nonperforming Loans 6.18 One of the most troublesome problems of the Turkish banks to date is the increasing amount of bad debts. As explained in preceding chapters, the prevailing high level of real lending rates is not sustainable in the long run for borrowers who have no access to preferential credit. Confronted by the choice of whether to declare the borrower in default or to roll over the loan, banks are opting for the latter in an increasing number of cases. 6.19 This problem and its possible solutions go far beyond the scope of bank supervision. Nevertheless, adequate regulatory constraints and appropriate reporting mechanisms could prevent the process of rolling over bad debts from accelerating rapidly and at the very least could provide the supervisory authorities with the opportunity to monitor it. In fact, no aggregated information is presently available to the authorities on the pervasiveness and evolution of this phenomenon, but the liquidity problems of some banks is a precursory sign. 6.20 There is to date no obligation for banks to report loans extended to customers whose repayment capacity is doubtful. Banks are only obliged to report those loans for which decisive steps have been taken for collection or liquidation. Besides, there is no incentive for banks to report these doubtful loans since the loans cannot be covered by a provision deductible from the banks' taxable income. The loans which banks have to report in their financial statement under a separate item are those which are being  - 68 - renegotiated with customers, with a view to alleviating customers' payment schedules, those which have already been rescheduled, and those which are being settled in court. At the end of 1981, the share of these bad loans in the banks' loan portfolios represented 4.27 percent for the banking system as a whole. To be considered as tax-deductible expenses, loan losses must meet very strict conditions. They may cover only the unsecured portion of loans and must relate to loans fitting one of three situations: the repayment is overdue and a formal order to repay has been given at least twice; the loan is being settled and collected in court; or the borrower has been given by the court a certificate of inability to repay. Another consequence of the nondiscrimination between sound and doubtful loans is that banks consider accrued and uncollected interest to be part of their income. If the loan is eventually declared in default, such a practice will have generated a cash-flow deficit because corporate taxes of 40 percent and dividends would have been paid but no interest payments would have been received. The current practice of rolling over credits corresponding to uncollected payments on interest and principal contributes to a potentially dangerous situation, although the practice may be unavoidable in the face of high interest rates and the liquidity squeeze. 6.21 The following loans should be reported separately by the banks to enable the supervisory authorities to detect any worsening of a bank's position: loans that are overdue by, say 90 days; loans that display some well-defined weaknesses that may inadequately protect the banks' credit position at some future date if not corrected. In the classification of loans it would be necessary to ascertain whether the difficulties of repayment of interest and principal by the borrower are mainly due to problems of liquidity caused by very high nominal interest rates or whether they exhibit a serious weakness, (such as a borrower's unbalanced financial structure or negative profitability), that make the collection and liquidation in full highly questionable. 6.22 In addition, the conditions under which provisions for banks' loan losses may be set as deductions from their taxable income should be relaxed. If properly administered, a relaxation of the rules would not affect tax receipts significantly in the long run but would contribute to smoothing out short-term fluctuations and would force bank management to plan better. Two methods are possible. The first is to allow banks to charge to operating expenses an amount equal to the average ratio of net charge-offs to average loans for, say, the most recent five years., The second method is to give banks' management total freedom in determining the required amount and to check ex-post, by means of on-the-spot examinations, the adequacy of the policy. Although the second method provides for a more flexible system, it requires some coordination with the government tax inspectors, whose objectives of necessity oppose those of bank supervisors. 6.23 As regards the problem of accrued and uncollected interest payments, a policy must be adopted to deal with nonaccrual of interest on delinquent loans. Banks should not be allowed to accrue interest on any loan when principal or interest are in default for 90 days or more. Obviously, even loans that are rolled over would be classified as delinquent loans. In those conditions the rolling over of loans would not in effect relieve the banks for more than three months; beyond this period, the accrual of uncollected interest would be prohibited.  - 69 - F. The Interbank Market 6.24 An interbank market is practically nonexistent in Turkey, apparently for the following reasons: (i) the keen competition for deposits and the high demand for- credit facing practically all the banks in the country discourage those financial flows from provincial banks to illiquid urban ones which are a major source of funds for the interbank market in many countries; (ii) there has been insufficient penalization of shortfalls in reserve requirements, compared with the marginal cost of bank resources; consequently, interbank borrowing has been more expensive thain the cost of not meeting reserve requirements; and (iii) the financial. transactions tax considerably increases thie cost of interbank borrowings. 6.25 Although the first problem is largely structural, the last two might be solved by increasing the 0ost of not meeting reserve requirements to a level higher than the cost oiE, say, time deposits and by abolishing immediately the taxation of :interbank transactions. The creation of an interbank market would be highly desirable to provide a "safety valve" for the liquidity of the banking sysitem. G. Stabilization of the BankLing Sector 6.26 At the end of 1981, and again in 1982, there were runs on the brokers ; the run in 1982 was accompanied by substantial deposit withdrawals from the smaller commercial banks. Even without such withdiawals, the liquidity problems of the banks have been quite serious, mainly because banks have been forced to refinance a large proportion of the amortization and of the interest on their loans outstanding. The Government must consider both how to prevent instability in the banking system and how to deal with it should it occur. The following recommendations, if implemented, should help achieve stability in the financial system. 6.27 A measure widely employed in other countries to keep the confidence of depositors is deposit insurance. In Turkey the face value of deposits is guaranteed by Government, buit not accrued interest. The authorities are considering the improvement of the system of deposit insurance in the context of the new banking law. In an environment of such high interest rates, a guarantee of the interest dtLe, as well as the face value of the deposit, may be needed to prevent deposit: withdrawals. Of equal importance would be to assure depositors that their funds would not be frozen in case of liquildity problems, as can now be done for up to three years. Only deposits up tto a l/ Brokers are intermediari.es dealing primarily with corporate bonds, but until recently were also handling certificates of deposits.  - 70 - certain size (perhaps TL 1 million) should be covered by deposit insurance. Of course, deposit insurance should not cover the commercial deposits that the banks require borrowers to hold in the form of compensating balances. Furthermore, deposit insurance should assure the payment of reasonable but not "excessive" interest rates on deposits; it might guarantee rates up to the level of the gentlemen's agreement but none higher. 6.28 A somewhat more drastic measure to strengthen the system would be to merge the weaker banks with banks whose liquidity and capital position are stronger. Mergers would improve the situation in the short run only if banks in difficulty were merged with sound banks. The merger of two or more weak banks might not be effective. But, because few banks are presently in a strong position, this would increase concentration in the banking system. 6.29 The measures mentioned above are intended to improve the position of the banks and hence to prevent excessive deposit withdrawal. In the second half of 1982 and in the beginning of 1983, the Government has successfully dealt with the difficulties created by deposit withdrawals from individual banks by enlarging the rediscounts of the Central Bank in favor of banks with particularly acute liquidity problems. Some people have commented that withdrawals of deposits from smaller banks would be of little consequence to the financial system. Although it is true that the assets involved are small, the danger of contagion cannot be dismissed; indeed, that the problems in the broker market in 1982 so quickly spread to the smaller banks is indicative of the danger. 6.30 In all assistance measures, the Government should distinguish between the banks' depositors and equity holders. The objective of any intervention should be to protect the deposit holders; no concern needs to be paid to the equity holders. H. Interlocking Ownership and Flow of Funds Between Banks and Corporations 6.31 Today the banks are certainly competing for deposits, and this has been of benefit to the deposit holders. There is far less competition on the lending side. One aspect of the problem has to do with the interlocking ownership and management of banks and industrial corporations. Firms not part of a banking group have less opportunity to borrow and, hence, can be prevented from competing in particular product markets. Moreover, profits can be shifted among banks and firms in the same group by varying lending charges. This has an adverse impact both on tax collections and on the position of minority shareholders. Finally, the lack of "arm's length" appraisal of credit applications may lead to excessive lending to particular corporations, adding to the instability of both the corporations and the banks. 6.32 In the absence of an adequate equity market the Government cannot require divestiture of shares. There would be some gain, however, in trying to separate both ownership and direct management of industries from banks in order to reduce conflicts of interest. 6.33 To reduce the problems created by interlocking ownership of banks and corporations, the new Banking Law will reduce the existing limits on credits  - 71 - to companies related to a ban'k, as well as to large customers. The solvency ratio recommended above in para. 6.12 could also be used to attack the problem of interlocking interests, as explained in that paragraph. I. Shortcomings of Accountin,g and Auditing Systems 6.34 The reliability of the accounting and auditing systems differs widely among banks. The first three or four leading banks, and a few of the snualler ones, have endowed themselves with efficient systems; most banks, however, still have poor accounting systems and virtually no internal auditing policies. There are no uniform accounting standards in Turkey, not even guidelines for the recording of bank operations laid down by the supervisory authorities. Banks are free to choose any internal system, provided they comply with generally accepted accounting principles, and they are only obliged to follow a format for the drawing up of the financial statements required by the regulatory authorities. Only a few banks have adopted an internal accounting system that fits the format. 6.35 As regards the auditing systems, banks' shareholders are oblige,d under the commercial code to elect at least one internal auditor who is answerable to them. These auLditors are also obliged under the Banking Law to report any infringement of the provisions of the law to the regulatory authorities. Functions assumed by the auditors are purely nominal, and they rarely report to the authorities. Besides, there is no legal obligation for external auditing, and banks do not resort to these services. 6.36 The recent limited openness of the banking sector to international transactions and the example set by the bigger banks in the strengthening of auditing systems may induce other banks to upgrade their accounting and auditing standards. Nevertheless, this will take a long time if the trend is not encouraged by the authori'ties. Some thought is being given by the regulatory authorities to compel banks to adopt particular standards. Experience in other countries indicates that such a solution is very costly to banks, and it requires from the authorities a body of knowledge that Turkey may not presently possess. AL more viable solution in the short run would be to refine the contents of some of the sensitive items that banks show in their compulsory financial statements. Some basic accounting guidelines should be laid down, in particular for the recording of doubtful and bad debts, for the writing-off of loans, for accrued and uncollected interest payments, and for provisions for loan losses. 6.37 To improve the auditing system, banks should be encouraged to plan a medium-term strategy for its development, and shortcomings duly established in the course of examinations by the Board of SBA should be severely penalized. The Board of SBA should closely monitor the remedial actions taken by the bank. J. Supervision of the Banking System 6.38 The supervisory system has been established since 1958 and involves about 100 people. Apart from the collapse of half a dozen small banks in 1966-67, which was handled efficiently by the Government, the banking system as such has not experienced any serious alert until the crisis of 1982 and the system of bank supervision has on the whole been considered satisfactory. But  - 72 - the present economic and financial conditions may prove to be a major test of the resiliency of the banking sector and of the adequacy of the supervisory system. Some rearrangement of the respective responsibilities of the various units in charge of the supervisory process is needed to eliminate current duplication of effort and to put the present facilities to their best use. 6.39 The Turkish Banking Law entrusts "the control over the application of the Banking Law and of the provision of other laws pertaining to banks and all sorts of banking operations" to "sworn bank auditors associated with the Ministry of Finance". Other provisions of the Banking Law pertaining to licensing, opening of branches, setting up of prudential regulations, merger, or liquidation of banks designate the Ministry of Finance as the relevant authority. The Central Bank is legally committed to bank supervision only in an advisory capacity; the Central Bank Act provides that "whenever requested by the Government, the Bank shall give its consultative views on matters connected with ... banking and credit questions in general." For the conduct of montetary and credit policies, however, the Central Bank is the authority in charge of monitoring and of regulating the banks' activities. It is also the legal depository for the Liquidation Fund. In practice, three different units are involved in the supervisory process: the Banking Department within the Central Bank (40 people), the Bank Department of the Treasury (40 people), and the Board of SBA, (40 bank examiners and 12 support staff). The last two are within the Ministry of Finance. 6.40 On-the-spot examinations are carried out solely by the Board of SBA in accordance with the Banking Law, but the banks' financial position is monitored routinely by all three units. Banks must submit quarterly financial statements to the Ministry of Finance and to the Central Bank, and in 1981 they were also requested to submit in a different format monthly financial statements to the Central Bank. Banks provide the Central Bank with all sorts of additional information--their foreign exchange position, a detailed breakdown of their loan portfolios, their biggest loans, reports on bad checks, credit files on customers eligible for Central Bank rediscounting, and so forth. No use is made of this information to ascertain the individual bank's financial position. 6.41 On-site examinations, intended to appraise the performance of banks' management and their overall financial position, should be given the highest priority in the supervisory process. The growth of banks in the past two years, their opening to international financing, and above all the recent difficulties of some of the banks call for frequent and in-depth monitoring. The Board of the SBA has been fulfilling its duties now for 24 years and is staffed with well-trained bank examiners who are also familiar with advanced techniques. The workload of the Board, however, has grown. Besides the examinations of an appraisal nature, bank examiners have increasingly been called upon to assume various other tasks. In 1981, for example, in addition to 23 general bank examinations, bank examiners, (who then numbered 34), had to carry out inspections of 16 brokerage firms and of 149 bank branches, to follow up on 24 notifications of complaints, and to do 6 research studies. On top of all this, the Board was also supposed to monitor the banks on the basis of their periodic financial returns.  - 73 - 6.42 Bank examiners should be assigned mainly the role of checking into the solvency, liquidity, and profitability of the banks and of assessing the banks' managerial skills by means of on-site examinations. In particular, greater emphasis should be put on the evaluation of the actual equity base of banks. Appraisal of bank assets, with respect to equity participations, fixed assets and loans should occupy at least 80 percent of the examiners' time, as is the case in other countries. On the matter of equity participations, bank examiners should exercise their legal right to investigate affiliates when the situation so warrants. As far as the liquidity of banks is concerned, bank examiners should strive to assess the degree of transformation effected by banks, assisted in this respect by the preliminary analysis that the Central Bank makes of the banks' monthly returns. Profitability should also be looked at carefully; in particular, foreign exchange profits, interest charged on doubtful loans, and assessmeni: of adequate provisions for loan losses should be given greater emphasis. 6.43 To enable bank examiners to deepen their analysis of these matters, they have to be assisted or relieved of other tasks. Since it is proposed that the Capital Market Board (CMB) will be responsible for the inspection of brokerage firms, bank examiners will be relieved of this task in the long term. In the interim, while bank examiners are still involved in the work of the CMB, a temporary increase in the SBA's staff ceiling should be authorized. As mentioned earlier, inspection of bank branches should be carried out only for the sake of test checking, and this would be a major time saver. Bank examiners should also be assisted by outsiders in the course of examination of banks, and the Banking Law should be amended accordingly. As noted above, once operational,, the group of auditors of the CMB could help bank examiners to scrutinize bank operations in securities and could provide bank examiners with their own assessment of banks' affiliates, if supervised by the CMB. Central Bank sta:Ef should also be drawn upon for controlling the banks' fulfillment of their obligations under regulations pertaining to monetary and selective credit policy or to foreign exchange control. Finally, the preliminary processing by the Central Bank of various data pertaining to banks' operations would facilitate the examiners' work. 6.44 As regards the routine monitoring of banks based on documentary evidence and factual information, it is suggested that the Central Bank play the major role, if only to ease the workload of the SBA. 6.45 It is believed that greater involvement by the Central Bank would contribute to increased efficiency and to enhanced utilization of all available sources of information. Even in the limited number of countries where the central bank is not the primary authority responsible for bank supervision, its participation is much greater than in Turkey. This is especially true ir countries where bank supervision comes under the responsibility of an autonomouls commission; in practice, the staff of such a commission is usually made up of central bank staff, and most of the preliminary treatment of banks' returns is taken care of by the central bank itself. The rationale for a central bank's larger responsibility in the supervisory process is multifold. Since the central bank handles the daily cash transactions of the banks, it is in a position to detect promptly any sudden change in a bank's cash position, in interest rate fluctuations, or in foreign exchange transactions, and it comes to learn about any impending  - 74 - difficulties. A central bank has at its disposal a cadre of officers experienced in banking practices and in monetary regulation. It collects valuable statistical data on the working of the financial system and is well equipped to process information. Finally, a central bank is the lender of last resort. 6.46 In Turkey, the Central Bank could process the monthly and quarterly financial statements on a computerized basis and could provide the Board of SBA with selected indicators of the individual banks' financial positions. An initial computerized control of the banks' compliance with legal requirements could also be provided. In the same way, the risk centralization scheme would be utilized with a view to giving the Board of SBA selected information on the composition of each bank's portfolio and, in particular, a first assessment of credit concentration, prior to a bank examination. A summary of individual credit files in the name of borrowers who benefit from rediscounting facilities could also be made available to the Board of SBA, together with a rating from the Central Bank so as to facilitate the examiner's work when he assesses the quality of a bank's portfolio of loans and participations. In the longer run, if the Central Bank engages in a systematic collection of financial information on firms in the industrial and commercial sectors and sums up its assessment of the firm's creditworthiness by giving it a credit rating, the scheme could be interconnected with those of the Central Risks Service and of the unpaid checks. As already implemented in other countries, such a system would provide bank supervisors with current information on the quality of a representative sample of loans for each bank. 6.47 In this context, the staff available in the Banking Department of the Central Bank could continue to analyze the banks' financial statements on a current basis and to draw the attention of the Board of the SBA to any foreseeable difficulties of individual banks. The Banking Department of the Central Bank should also be the focal point of contact between other departments of the Central Bank and the Board of the SBA. The relevant departments of the Central Bank should notify the Board of the SBA of any suspected infringement of monetary rules via the Banking Department. Under these circumstances, there should not be any further need for the Treasury to duplicate the efforts of the Central Bank. The Banking Department of the Treasury should be mainly concerned with the application of the specific provisions of the Banking Law that provide for its involvement--namely, licensing, opening of branches, revocation of license, liquidation, and setting up of prudential regulations.  - 75 - CHAPTER VII - THE CAPITAL MARKET A. The Case for StrengtbeninDg the Capital Market 7.1 Given the high internediation costs of the banking system, there is a clear need to introduce elemeints of competition from instruments and institutions outside that system and to create a more direct link between savers and the corporate sector. This should permit a reduction of intermediation margins, diversify the supply of financial instruments, and effect a better allocation of resources and mobilization of financial savings. In addition, more direct financing from industry would to some extent bypass the interlocking of banks and large industrial groups, with its distortive effects on the allocation of credit and the concentration of economic power. 7.2 The experiences of many countries show that credit ceilings and high intermediation margins in bankring often lead to the creation of new instruments--such as money-market paper, debentures, and leasing--if new entrepreneurs are permitted access to the market on fair terms. To some extent this has also happened in Turkey, but in an unregulated environment. This competition led to a rapid expansion in the market for bonds and other debt instruments, but it was Followed by the brokers' crises of December 1981 and June 1982, whicb were the result of inadequate credit information and analysis combined with insufficient government supervision. 7.3 There clearly is a need for an adequate regulatory framework that will promote rather than prevent the formation of capital and money markets, for which there is an equally great need in Turkey. The development of a more efficient market for corporate bonds and other debt instruments may initially prove easier, if adequate measeures to restore investors' confidence are taken, but the greatest long-term beniefits will be derived from the development of an equity market. 7.4 It is important to realize that the capital market is only a segment of the financial sector, and ithat its modernization will yield mostly long-term benefits; consequently, one should not expect from it an immediate solution to the problems affecting the country's economy. The Case for Equity Financing 7.5 As is typical in many inflationary economies witb subsidized credit schemes, Turkish corporations have relied largely on debt financing. In addition, the availability of credit from affiliated banks has removed the need for large industrial groups to raise new equity to maintain prudent debt equity ratios. As long as real interest rates were negative, this lack of new equity did not put serious strains on the financial position of these groups, since inflation "wasbed debt away". However, as discussed in Cbapters 5 and 6, with the present bigb real interest rates, the scarcity of credit and the high costs of foreign finaDcing, there is the risk of worsening undercapitalization in both industry and banking, which is likely to continue in the future. The need for new capital is thus much higber now than it has  - 76 - been in the past. However, there is evidence that, given the high level of real interest rates, many entrepreneurs are now much more interested than before in increasing the equity of their firms. 7.6 On November 22, 1981, the Central Bank, worried by the serious undercapitalization of the banking system, increased the minimum capital requirement for banks, thus forcing them to raise new equity. Altbough new equity was raised in the course of 1982 by a number of institutions, particularly banks, 1/ many otbers have been unable to raise new capital from their controlling groups or from the public. The undercapitalization of many banks is now such that, if they bad to write off their bad accounts, they would probably show negative net worth. 7.7 Besides providing a remedy to the undercapitalization of the corporate sector, an increased supply of equity through a well organized securities exchange would contribute to the creation of an environment more conducive to innovative entrepreneurial activities and would help to dilute the concentration of economic ownership and control and to spread participation in the country's wealth among a wider group. For these reasons, it should be the focus of a program for the development of the capital market in Turkey. B. Constraints on the Development of the Capital Market Need for a Long-term Program 7.8 The Capital Market 2/ in Turkey is small in comparison with the banking system, and its development is hampered by some of the institutional weaknesses and constraints that also affect otber sectors of the financial system. 7.9 The following bottlenecks should be addressed as a matter of priority; (i) the lack of adequate accounting and auditing standards, whicb limits the availability of accurate data and has contributed to a regulatory environment that does not require the disclosure of reliable financial information; (ii) the interlocking of banking and industry, a connection that supplies large industrial groups with a privileged source of credit and greatly reduces the need to seek sufficient equity capital; because the same groups also control most of the bond underwriters, they beve similarly privileged access to the bond market; 1/ From data supplied by the stock exchange, it appears that 13 banks were responsible for about 80 percent of the amount of all new capital increases by 71 companies listed on the Stock Exchange in the course of the first ten months of 1982. 2/ Defined as the primary (new issue) and secondary trading markets for long-term debt and equity instruments.  - 77 - (iii) the absence of a strong network of non-bank financial institutions to support capital market activities; and (iv) the inadequacy of the present secondary markets for securities, which is reflected by the fragmentation of the market into a number of over-the-counter dealers, without an effective securities' exchange; without an adequate secondary market, issues of securities become extremely difficult, because they lack the liquidity that would make them appealing to investors. 7.10 The recent financial crises have been painful but provided healthy lessons that have prompted act:ion by the authorities to regulate a market that had grown in a legislative vacuum. The resulting system is, however, characterized by artificialities and distortions, as well as by a certain lack of reciprocal understanding by private operators and public regulators. 7.11 The enactment, in February 1982, of a Capital Market Law, with the creation of a Capital Market Board (CMB) responsible for the supervision of the market, is a significant step in the right direction, although the Lalw may need improvement in a number cf areas and should be interpreted as giving the CMB responsibility for the development, and not just the regulation, of both the primary and the secondary markets. It provides a framework for futur,e reform of the system and for t.he formulation by the CMB, in cooperation with the private sector, of a long-term program for the orderly development of a modern capital market. As a matter of high priority, the supporting institutional and supervisory framework should be strengthened. This should include: (i) strengthening the capability of the CMB to develop, regulaite, and supervise the market; (ii) improvement of accounting, auditing, and disclosure standards and practices; and (iii) amendment of the Capital Market Law to include the regulation of the secondary market and intermediaries in this market. 7.12 Before these measures are taken, it would be unwise and even dangerous to make a large-scale effort to foster the growth of capital market instruments. Once considerable progress has been achieved in the fields mentioned above, further steps should be taken to: (i) develop nonbank financial institutions and institutional investors; (ii) review and eliminate constraints and distortions affecting the demand and supply of securities, including tax and other reforms to remove disincentives for the equity market and artificialities in the bond market; and (iii) reform the secondary market, including the establishment of a new Securities Exchange.  - 78 - C. Strengthening the Regulatory, Supervisory and Institutional Framework The Capital Market Board 7.13 The recently established CMB is expected to play a key role in the regulation, supervision and development of the capital market. The CMB bas already made important efforts to set up a regulatory framework; bowever, given its relative lack of experience and of adequate staffing, there is a risk that regulation be given precedence over developmental and supervisory activities because regulation requires less skill and manpower. This priority can lead to the formulation of rules that bamper the development of the market or that are unenforceable. The CMB should strengthen its staff by; (i) hiring qualified financial analysts to screen new issues and supervise market operations; and (ii) recruiting an experienced advisor who should help design training programs for the staff and provide advice on accounting and auditing standards, disclosure requirements, screening criteria for the issuing of securities, and supervision of market operations and measures to develop the primary and secondary markets, including the organization of a securities excbange. 7.14 Initially, the CMB should focus on strengtbening its abilities to set and to enforce high standards for the development of accounting, auditing, and disclosure of information and to carry out effective supervision on the operations of the securities market. Once this is achieved, the CMB could concentrate on studying and implementing measures to increase the supply of and demand for securities and to establish a new securities exchange. Disclosure of Information and the Screening of New Issues 7.15 In line with its responsibilities under Article 16 of the Capital Market Law, the CMB, besides preparing and monitoring standards and principles for the disclosure of financial information by publicly owned corporations (tbat is those with at least 100 shareholders or those which have made public offerings of securities), is responsible for approving public issues of securities on the basis of a judgment on their quality. This role of the CMB, althougb perhaps justified by the relative lack of sophistication of investors and intermediaries, puts it in the dangerous position potentially of suffering discredit if the issues authorized prove financially unsound. In addition, even if the CMB bad the staff to do a financial appraisal of the issues, it would have to work on unreliable data because of the lack of adequate accounting principles and auditing standards. 7.16 Recommendations concerning the improvement of the systems of accounting and auditing of corporations are presented in Chapter 6. Because, however, the implementation of the recommendations will take time, before adopting otber measures to foster the growth of the primary and secondary markets, the authorities sbould initially require the external auditing by  - 79 - licensed firms of at least all financial institutions and of the companies offering securities to the pubLic or having a sufficiently large number of shareholders. Because this might discourage public offerings, at a later stage large, closely held compamies should also be required to submit audited accounts up to the same standards for tax purposes. Meanwhile, they could be subjected to frequent, extensive inspections by the tax auditors, whose staff should be strengthened. 7.17 Once satisfactory accounting, auditing, and disclosure standards have been established and the sophistication of investors and intermediaries has increased, the role of the CMB should be limited to monitoring compliance with disclosure requirements, which should be the same for equities, bonds, and other instruments. The appraisal of the quality of issues should be left to the market. Strengthening the Supervision of Securities Market Activities 7.18 Until September 1981, brokers were in principle under the supervision of the Ministry of Commerce, whlich along with provincial authorities was also in charge of their licensing. None of these bodies was equipped to regulate and supervise an active market properly, and even less so to track down unlicensed operators. In September 1981 a Lending Act subjected brokers to licensing by the Ministry of Finance and to supervision by its Board of Sworn Bank Auditors (SBA). Before any corrective measures could be implemented, however, a number of small brokers began to collapse. 7.19 Following the crises, the Capital Market Law of February 1982 and rules issued by the CMB and other agencies have greatly improved the regulatory framework by: (i) setting criteria for floatation of securities and disclosure of information; (ii) prohibiting brokers and other securities firms from carrying out some activities (including lending without a specific license, dealing in certificates of deposit and certain othier money-market insltruments, issuing paper representing their own liability, and giving guarantees of the repurchase of bonds at predetermined prices), which were perceived as being among the causes of the crises; (iii) strengthening thie liquidity of brokers and securities firms by establishing a Securities Regulation Fund, whereby securities can be swapped for cash, with a repurchase agreement; and (iv) requiring brokers and securities firms with underwriting activities, in order to continue operations after February 1, 1983 (a limit which was afterwards postponed by one year), to have aL minimum capital of TL 200 million and to obtain a license from the CMB, which is responsible for their supervision; for brokers and securities firms without underwriting activities the minimum capital required is TL 50 million.  - 80 - 7.20 Given the small number of firms currently active in the securities business and of those likely to be licensed in the near future, it should not take long for the CMB to achieve, perhaps with the initial secondment of SBA from the Ministry of Finance, the capability to properly supervise these firms. This supervision can be carried out by: (i) requiring securities firms to submit periodic reports with frequency, format and underlying accounting principles to be set by the CMB. The reports should be submitted at least on a quarterly basis and should include a breakdown of loans obtained, of positions in securities, obligations in respect of security underwritings and net positions resulting therefrom, of customer's debit and credit balances, and of repurchase agreements and other relevant information; and (ii) undertaking on-site examinations. In principle, the CMB should be responsible for the examination of securities firms, but, according to Article 45 of the Capital Market Law, the Ministry of Finance may also act on its own. To avoid confusion and ensure unity of supervision, it would be logical to make the CMB fully responsible for these examinations. 7.21 A more serious problem is caused by the difficulty of detecting unlicensed operators, whose number could become quite large if licensing standards are set too high, if measures to favor the growth of the capital market are taken, and if trading is eventually concentrated by law on the securities exchange. To detect the potentially destabilizing activities of unlicensed operators, the CMB, besides strengthening its staff, should closely cooperate with an association of licensed securities firms. This association should adopt and enforce a code of fair practice for its members. Its role with nonmembers should be confined to the reporting of unlicensed operators to the CMB which would be responsible for the follow-up and enforcement of the law. Non-Bank Financial Institutions 7.22 A variety of non-bank financial institutions are necessary to diversify and to make more efficient a financial system presently dominated by banks and to support capital market activities. In many countries, these roles are typically played by investment banks and securities firms (as intermediaries) and by such institutional investors as pension funds, mutual funds, and insurance companies (as investors in capital market instruments). The latter class of institutions often referred to as "contractual savings institutions" can play a key role in mobilizing long-term savings from middle income households. 7.23 In particular, there is a clear need for firms that can act both as underwriters in the primary markets and as brokers-dealers in the secondary markets for money, bonds, and equities to fill, in an orderly and regulated environment, the gap left by the disappearance of most brokers. Some of the existing intermediaries, particularly the few remaining bond underwriters, do presently to some extent play such a role. Their activities, however, are  - 81 - usually limited to the handling of securities issued by their controlling groups and this limitation prevents the development of their own ability to appraise securities. The evolution of such intermediaries should be encouraged through training programs to acquaint them with investment banking practices abroad; in addition, the creation of new institutions not linked to specific groups would be desirable. 7.24 Given the desirability of financially strong securities firms, the TL 200 million in capital required for licensing by the CMB of intermediaries with underwriting activities is perhaps modest. There is, however, justification for the present requirements according to which the minimum capital for dealers not engaged in underwriting is only