The private sector's response to financial liberalization in Turkey : 1980-82

Polly, Planning, and Research WORKING PAPERS Financdal Policy *nd Systems Country Economics Department The World Bank January 1989 WPS 147 The Private Sector's Response to Financial Liberalization in Turkey: 1980-82 Izak Atiyas The private sector's response to the short-lived episode of liberalization in Turkey in 1980-82 did not live up to expecta- tions. lTbPohlicy, Pbaning. and Rese Comple disrib PPR Wa'*gPapu todintcemdefindpgsofwk m pnrs sd to alaiwoge the exchange of ideas amng Bank staff and e3o0 inm eed in developmnt issu. Mum papers cay the nanw of the authoas, ile only their views. ad shoid beused sad cid accordingly. Tbe finding, inetaioms, and cncltsio am the authors own. Tey shod na be ated to i Wodd Bank. its Boad of Dimos its managemea or ny of its member cant.  Polsr Paning, and lRcshm Fln ndal Podby and Sytems Fmancial liberalization was carried out in a An analysis of finn-level data reveals that period when the nonfimancial corporate sector nonfinancial corpor ions were subject to both was in financial distress due to reduced profita- an earnings shock (increases in costs relative to bility. The consequent emergence of substantial sales income) and an interest rate shock. Al- nonperfonming lo-is in the banking sector, though the debt-to-asset ratios cf profitable especially among smaller banks, created fierce firms did not change, those of firms under competition for financial resources. The result distress actually increased, despite higher costs was a rapid expansion of deposits and high real of borrowing. interest rates. Instead of forcing insolvent bor- rowers into bankniptcy, banks refinanced The Turkish experience suggests that nonperforming loans as a way to prolong their financial liberalization may not produce desired own survival, and real credit to the private sector results when it occurs in a period of major increased dramatically. Furthermore, the market macroeconomic realignments that adversely mechanism turned out to be il-equipped to affect the profitability of the corporate sector, induce the exit of insolvent banks and thereby especially when it is implemented without an increase the efiTciency of allocating loanable adequate regulatory framework. funds. This paper is a product of the Fmancial Policy and Systems Division, Country Eco- nomics DeparEment. Copies are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Wilai Pitayatonakam, room N9-005, extension 60353. The PPR Working Paper Series disseminates the findings of work under way in the Bank's Policy, Planming, and Research Complex. An objective of the series is to get these findings out quickly, even if presentations are less than fully polished. The findings, interpretations, and conclusions in these papers do not necessarily represent official policy of the BankL Produced at the PPR Dissemination Center  The Private Sector's Response to Financial Liberalization in Turkey: 1980-82 by Izak Atiyas Table of Contents 1 - Introduction and Summary 1 2 - Developments in the Banking Sector 4 Stage 1 - The Initial Months 6 Stage 2 - Change in the Environment 7 The Response of Policy Makers 8 The Crisis 9 3 - Adjustment in the Corporate Sector 18 Profitability and Its Components 19 The Analysis of Variance Model 20 Empirical Results 24 4 - Conclusion 30 References 32 Appendix 1 34 Appendix 2 35 Tables 36 Figure 1 41 I am grateful to Charles A. Wilson, my dissertation adviser, and Tosun Aricanli, Ishac Diwan, and Dani Rodrik for helpful comments and encouragement. The paper also benefited from comments by Ataman Aksoy, Bela Balassa, Tercan Baysan, Merih Celasun, Alan Gelb, Patrick Honohan, and from discussions in the conference on the Political Economy of Turkey in the 1980s. I thank research teams at the Capital Markets Board and the Central Bank of Turkey for their help and support in obtaining the data.  1 - INTRODUCTION AND SUMMARY In mid 1980, economic policy makers in Turkey initiated attempts to restructure and deregulate the financial sector. The main steps taken were the removal of legal restrictions on interest rates and the allowance and encouragement of financial transactions through new types of financial institutions and instruments. An important aspect of the deregulation attempts was that they were implemented simultaneously with and, indeed, as part of, a comprehensive stabilization/liberalization program. Among the most important objectives of financial liberalization the following were mentioned by policy makers: - Deregulation was expected to increase interest rates -which were hitherto negative in real terms-, and therefore increase financial savings and deposits in the banking sector, as well as introduce competition into the banking system, although, as will be seen below, the nature of the competition that was desired was ambiguous. It seems more plausible to assume that policy makers interpreted competition as elimination of direct government controls rather than non-collusive behavior on the part of banks. - It was hoped that an increase in interest rates would push corporations in the private sector, which were typically highly leveraged, to reduce their indebtedness and increase their equity base. Owners of firms were advised to liquidate personal wealth and transfer it to their firms as equity capital. "Sell your villas to finance your corporations" was the call oE the day. The increase in the interest rates was expected to allocate loanable funds to most profitable users. - Bankruptcy was believed to be a major regulatory force in the corporate sector. Inefficient firms, especially those in previously protected import competing sectors, were expected either to adjust to the requirements of the  -2- new economic policies, for example, by vegrina production towards exports, or to leave the market. On the macroeconomic side, the most important developments in 1980-82 were export orientation and disinflation.' The comprehensive stabilization/liberalization program included policies aimed at trade liberalization, real depreciation of the exchange rate and reduction of domestic absorption. The economy responded quickly. The Gross National Product (GNP) growth rate, which was negative in 1979 and 1980 picked up and reached 4.1% in 1981 and 4.6% in 1982; although below the 1973-77 average of 6.5%, the increase in growth rates did reflect a recovery. In the meantime, the composition of demand changed drastically: The rate of growth of domestic absorption, averaging 8.2% in 1973-77, was only 1.6 and 2.8% in 1981 and 1982. While the contribution of foreign balance to GNP growth was negative between 1973-77, it was positive in the 1980's--with the exception of 1983.2 Exports grew by 47 and 25% in constant Turkish lira prices in 1981 and 1982 respectively. The decline in the rate of inflation was even more dramatic. The annual rate of change of the wholesale price index declined from around 107% in 1980 to 37% in 1981 and 25% in 1982 (see Figure 1). In short, 1980-82 were years of drastic realignments in major macroeconomic variables, which were bound to affect corporate performance. The deregulation episode in the financial markerc lasted two and a half years. The objective of increasing deposits and financial savings in general 1 For a detailed overview of macroeconomic policies and performance in the 1970s and 1980s, see Celasun and Rodrik (1987). 2 Data from OECD, Economic Survey: Turkey, various issues. Let Y, A and B stand for GNP, domestic absorption and the current account surplus respectively, all in constant prices. Then, Y - A + B and Y - aA + bB where hats denote percentage changes, a - A/Y and bA- B/Y. The contribution of foreign balance to GNP growth is defined as bB.  -3- was met with considerable success. However, a financial crisis broke out in 1982 and several brokerage houses went bankrupt. In 1983, the Central Bank took over the administration of some "problem" banks, put all the others uneer close supervision, and started to reregulate deposit interest rates. Overall, it can safely be said that the response of the financial sector to deregulation was worse than expected. The objective of this paper is to present an overview of the events that culminated in the crisis. This will be done in two steps. First (Section 2) the events in the banking sector will be summarized. In the banking sector, shocks to the corporate sector made portions of banks' assets nonperforming. This was true especially for some smaller banks. Rather than forcing liquidation of their clients, these banks engaged in a fierce competition to collect funds, to raise resources both to meet their liabilities and to refinance non-performing loans. Interest rates soared. Some banks were thus able to survive even though they were insolvent, until the government finally intervened. To get some clues about why bank loans might have become non-performing, Section 3 looks at the private corporate sector. Analysis of a panel data set of corporations reveals that in this period firms were subject not only to an interest rate shock, but also to a gross earnings shock (increase in costs relative to sales income). While one would expect that higher interest rates on loans would make firms reduce their indebtedness, debt to assets ratios of profitable firms did not change much during 1980-82, and those of firms under distress actually increased. These observations suggest that financial liberalization may not generate desired responses if it is carried out when there are major changes in the  -4- macroeconomic environment that adversely affect the profitability of and cause financial distress in the corporate, and consequently, banking sectors. Furthermore, the market mechanism did not seem to be well equipped to carry out its most essential regulatory function in an efficient manner, that is, inducing the exit of insolvent economic units and thereby decreasing inefficiency in the allocation of loanable funds. Section 4 will conclude the paper. 2 - DEVELOPMENTS IN THE BANKING SECTOR There were two principal sets of players in the financial crisis of 1982. On the one hand were the banks. The Turkish financial system has been dominated by commercial banks. At the end of 1979, the commercial banking system was composed of 12 state owned banks, 24 private banks and 4 foreign banks. The market was highly concentrated. The share of the largest 4 banks (one state owned and three private) in total assets was 58% and 56% in total deposits. Each of 11 smallest private banks held less than 1% of the total assets of the banking system. Most private banks were owned or controlled by industrial conglomerates.3 This pattern of ownership was the result of two trends: Some of the private banks were actually established by industrial groups controlled by individual families, while smaller and provin-cial banks were established by local businessmen and later acquired by industrial groups and transformed into nation-wide banks. New entry into the banking system was subject to the permission of the government. Prior to deregulation, the governments had been very conservative in granting the necessary permission. As a result, the 3One important exception is Turkiye Is Bankasi, the largest private bank in Turkey.  -5- number of banks had been stable between 38-42 in the 1970's. Prior to deregulation, interest rates on both deposits and credits were fixed by the government at low levels. vlith accelerating inflation in the 1970's, real interest rates on depcsits were largely negative (see Figure 1). Another set of players that proved especially important in the events after the deregulation were the brokerage houses. Most brokerage houses were establishsd around the year 1979, when industrial corporations started to issue bonds.4 Some "bankless" industrial groups, unable to enter the banking business, formed their own brokerage houses.5 The rapid development of the-- admittedly small--bond market also encouraged the establishment of independent brokerage houses.6 Deregulation in the banking sector was started by two important steps undertaken in July, 1980, whereby: a) Legal restrictions on deposit and loan interest rates were removed, and, b) banks were allowed to issue negotiable certificates of deposit (CDs).7 Following Artun (1983), the events that followed these deregulatory steps and culminated in the financial crisis of 1982 can be summarized in two stages:8 4Artun (1983, pp. 70, 77) argues that bond issues resulted from new financing requirements due to the impact of devaluations of 1978-79 on corporations. 5 Examples are Meban of Transturk Holding, Eczacibasi Yatirim of Eczacibasi Holding and Oyak Yatirim of Oyak Holding. 6 One has to make a distinction between the brokerage houses discussed in the text, which traded, at least in the initial stages of the process, in securities of industrial firms and banks, and, other unorganized money market institutions that collected funds solely against Rersonal cheques and IOUs. The latter type of institutions (dubbed "market bankers" in Turkey) mushroomed following the deregulation and were the actors of another crisis that unfolded at the end of 1981. These institutions and their evolution will not be addressed in this study. 7 For a more comprehensive overview of financial reform, see Akyuz (1988). 8 The following summary is primarily based on accounts given in Colasan (1983, 1984a, 1984b), Ulagay (1987), and Artun (1983, 1985).  -6- Stale 1. the Initial Months' Soon after the reform program was announced, larger banks encouraged members of the banking system to form a cartel and set deposit interest rates collusively, at a rate higher than the pre-liberalization level (30% on annual deposits). The monetary authorities did not seem to object to collusion, although they did think that the interest rate was low, given that the inflation rate exceeded 100% in 1980 (See figure U.9'10 In any case, the so-called gentlemen's agreement that was drawn between the banks was not adhered to; some (mostly smaller) banks offered higher deposit rates. Initially, it seems that the breakdown of the cartel agreement was due to an attempt by smaller banks to explc.lt their competitive edges, which arose from their lower intermediation costs. By the end of 1980, these banks offered deposit interest rates that were 2-5 percentage points higher than the 30% envisaged in the agreement. In February 1981, a new gentlemen's agreement was signed, whereby the rate of interest on one year deposits was raised to 50%. Soon, this agreement was also broken by some banks. Besides the deposit interest rate, CDs also proved to be an important tool of competition and were widely used. A substantial proportion of CDs were marketed through brokerage houses, both through those that were independent and those that were subsidiaries of banks or holding companies. A mechanism was developed whereby banks in effect used CDs and brokers to circumvent the gentlemen's agreements and t:Aed to increase their share in the market for deposits: CDs were issued to brokerage houses in large volumes at