Yugoslavia - Financial sector restructuring : policies and priorities (Vol. 2 of 2) : Annexes

Rewort No. 7869-YU Yugoslavia Financial Sector Restructuring: Policies and Priorities (In Two Volumes) Volume II: Annexes November 30, 1989 Country Operations Department IV Europe, Middle East and North Africa Region FOR OFF:CIAL USE ONLY Document of the WYorld 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 without World Bank authorization.  CURRENCY EOUIVALENTS Currency Unit - Dinar (Din.) Currency Unit Calendar Year 1988 ' September 30. 19S9 US$1 Dinar 2,522.6 Dinar 32,521.0 Dinar 1 US$0.0003964 US$0.00003 Dinar 1,000,000 US$396.42 US$30.75 WEIGHTS AND MEASURES Metric System GLOSSARY OF ABBREVIATIONS AND ACROrFMS BOAL Basic Organization of Associated Labor DFEL Deferred Foreigr. Exchange Losses GAAP Generally Accepted Accounting Principles GDP Gross Domestic Product GSP Gross Social Product IEF Inter-Enterprise Financing IMF International Monetary Fund IAS International Accounting Standard JLF Joint Liability Fund JRF J^n"-'t Reserve Fund LBL Ljubljanska Banka Ljubljana-Associated Bank LT Long-Term MLT Medium- and Long-Term NBFI Non-Bank Financial Institution NBY National Bank of Yugoslavia OAL Organization of Associated Labor PEF Preperty and Equipment Funds RAP Republic and Autonomous Province ST Short-term SDK Social Accounting Service (Sluzba Drustvenog Kujigovodstva) SMA Self-Management Agreement YBA Yugoslav Bankers' Association 1/ Period Average. l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~  FOR OFFICIAL USE ONLY Yugoslania Financial Restructuring: Policies and Priorities Table of Contents ANNEXES Page No. I: Modelling the Operations of the Central Bank 1-4 II: The Demand for Money .n Yugoslavia 5-17 III: The Dynamics of Public Sector Debt in Yugoslavia 18-23 IV: The Structure of the Banking System in Yugoslavia 24-28 V: Non-Bank Financial Institutions 29-31 VI: Yugoslav Money and Securities Market 32-33 VII: Structure of Income Statements of Yugoslav Enterprises 34-45 in the Period 1986-88 and the Proposed Structure for 1989 VIII: 1987 Financial Results of the Nine Commercial Bank Groups: 46 A. Investiciona Banka Titograd (IBT) 46-52 B. Jugobanka Udruzena Banka Beograd (JB) 53-59 C. Ljubljanska Banka Udruzena Banka Ljubljana (LBL) 60-66 D. Privredna Banka Sarajevo (PBS) 67-73 E. Stopanska Banka Skopje (SBS) 74-80 F. Udruzena Banka Hrvatske Zagreb (UBH) 81-87 G. Udruzena Beogradska Banka Beograd (UBB) 88-94 H. Udruzena Kosovska Banka Pristina (UKBP) 95-101 I. Vojvodanska Banka Novi Sad (VBN) 102-108 IX. Analysis of the Intermediation Spread 109-lll 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.  ANNEX I MODELLING THE OPERATIOlIS OF THE CENTRAL BANK Since the Central Banking System of Yugoslavia does not extend credits to the public sector, the question arises as to what factors have ;.riven the expansion of base money in the country. The acceleration of base money growth and inflation during the 1980s makes this question even more relevant. There have been sizable transfers of real resources through high and increasing inflation in Yugoslavia. The related question therefore is who have been the main beneficiaries of the inflation tax collected on holders of base money. In order to provide an answer to these questions, this annex provides a decomposition of the National Banks' main asset and liability operations, including the variations in their net worth. The final result allows for the identification of the main uses of the inflation tax in Yugoslavia--the concession of direct and indirect credits to the private sector at very negative real interest rates and the financing of foreign exchange losses linked to the large stock of foreign exchange liabilities of the National Banks. The balance sheet identity of the Central Banking System of Yugoslavia is written in first differences as: (1) NFA - NFA4., + Ct - Ct, + S, - S,, + NOI, - NOI,., - - Ht - 1j + NW, - NWt., Where NFA- Net Foreign Assets (foreign assets minus foreign liabilities to the rest of the world and to domestic banks), C- Credits to the Economy, S- Specific Credits, NOI- Net Other Items, H- Base Money and NW- Net Worth. The specific credits are Dinar credits granted to banks in return for the deposit of foreign exchange by banks with the National Banks (a foreign exchange liability of the National Banks). The credits to the economy are low interest Dinar credits, granted to both banks (rediscounts) and non-banks. Most of these credits are directed to agriculture and to exports. The variations in net foreign assets in terms of Dinars can be further broken down into: (2) NFAt - NFA,.I - E,_1 (NFA4 - NFA.1) + (Et - E,1)NFA11 + + (E, - Et.,)(NFA4 - NFA4.1) The first term above captures the variations in net foreign assets in terms of foreign currency (the star superscript denotes variables expressed in  2 foreign currency), the second term captures the exchange rate revaluation of the previous stock and the third term is the cross product. The variations in the National Banks' net worth are equal to the difference between interest revenues and interest costs, including the exchange rate depreciation: (3) NWt - NWt.,, (i,< + (E, -E,.,)/E, + i,(E, -E,,.)/E.,,)NFA,,Et, + +i, C,, + is s., The term inside the parenthesis is the depreciation-adjusted interest rate on the stock of net foreign assets of the National Banks. Since net foreign assets are negative (foreign liabilities are much larger than foreign assets), the first term above captures the net interest expenses on the stock of net foreign assets/. The second and third terms capture the National Banks' nominal interest revenues on the stocks of credits to the economy and specific credits. This formulation of interest expenses and revenues is admittedly simplified. Actual interest flows are likely to be affected by var-iations in interest rates and stocks within each year. In addition, interest flows are also affected by the frequency of interest payments and collection and the base on which interest rate changes apply (i.e. whether interest rate changes are applied to all outstanding stock or only to increases in the stock). Nevertheless, this formulation provides a first approximation of actual interest flows in the absence of direct information on these flows. The nominal interest rates in (3) can be further broken down in terms of real interest rates and the inflation premium: A it= r, + p, + r, pt (4) it,1 r, + Pt + r, pt Where r is the realized real interest rate in period t and p is the inflation rate in the same period. As above, the star superscript denotes variables expressed in foreign currency. The changes in the National Banks' net woith can now be expressed as: j~/ The simplifying assumption behind the equation is that the Ceatral Bank .arns on its foreign assets the same interest rate that it pays on its foreign liabilities.  3 (5) NW, - NW,, rt NFA,, Et + Pt NFAt. E, + (E, - E,.,)NFA,, + + rt (l+p,)C,1 + pt C,, + rs (l+pt)St.1 + ASI Combining equations (1) and (5) and dropping out the changes in net other items (NOI) one obtains a final expression for the changes in base moneyJ: * A* * (6) H, - H, E(NFA,I. (l+Pt )NFAt.1) + (Ct -(l+pt )C.1 ) + + (S, -(l+pt )St, ) - rt NFAj E, - rt (l+p, )Ct - r- (14 ) S,, The first three terms on the right hand side of equation (6) represent the real variations in net foreign assets, credits to the economy and specific credits, respectively. The final three terms represent the real interest payments on ret foreign assets and the real interest revenues on credits to the economy and specific credits. Equation (6) may also be expressed in percentages of OSP, as shown in Table 4 (Chapter 2). It should be noted that the capital loss term in equation (2) cancels out the same term in equation (5). A devaluation of the exchange rate increases the Dinar value of the stock of net foreign assets of the National Banks in absolute terms. If net foreign assets are negative, the exchange rate devaluation generates a loss which reduces the net worth of the National Banks but it does not have an immediate impact on base money creation. However, these capital losses may eventually be monetized by either of the following two ways: First, to the extent that the National Banks suffers a withdrawal of foreign exchange deposits (net foreign assets increase) which cannot be offset by an ecual decrease in the stock of specific credits, this withdrawal must be financed with recourse to base money creation. Second. capital losses on the stock of foreign exchange liabilities may also trigger monetization through larger interest payments. If the real rates of interest on the stock of credits to the economy are negative, that introduces an interest subsidy which must be financed with base money creation, as shown in equation (6). The total impact of credits to the economy on base money creation comprises the interest subsidy and the real variations in the stock. In the years when the real variation in the credits to the economy is negative, the total impact on base money is less than the interest subsidy. 2.! Equation (6) depicts the most important and systematic factors of base money expansion. However, changes in net other items may be important in particular years.  4 The creation of specific credits per se does not involve monetary creation, since increases in these credits are backed by increases in foreign liabilities. However, if the real rate of interest on these credits is negative, while the real rate of interest on the depreciation-adjusted stock of foreign exchange liabilities is positive, the difference must eventually be financed through monetization. In addition, if the Central Bank does not earn an adequate return on its assets, the stock of earning domestic assets will diminish relative to the stock of (negative) net foreign assets. The difference between them will consist of foreign exchange losses (negative net worth). When that is the case, withdrawals of foreign exchang3 liabilities (increases in net foreign assets) cannot be matched by decrea es in domestic credits. Under these circumstances, the increase in net foreign assets will also be monetized, as noted above. V S * Finally, in the extreme case where i t i it i 0, that is, all nominal interest rates on net foreign assets and on credits to the economy and specific. credi-s are zero, the equation (6) is reduced to: (6') H, - Ht1 Et (NFA, - NFAt,) + Ct - Ct + St - S,, In this case, the changes in base money are identical to the nominal changes in net foreign assets, credits to the economy and specific credits. Note, also, that when nominal interest rates are zero, the flow of real interest revenues on credits to the economy is equal to -Pt C,.1. In this case, the interest subsidy granted to the recipients of these credits is fully equal to the inflation rate multiplied by the beginning-of-period stock.  5 THE DEMAND FOR MONEY IN YUGO0LAVIA 1. Spacification The estimation of the demand for money in Yugoslavia follows standard practice. The demand for each mo.aetary asset/ is formulated as a function of real income, the nominal rate of return on alternative financial assets and the expected rate of inflation. The inclusion of the expected rate of inflation is justified because real assets may be relevant alternatives to fixed-income financial assets. In addition, the relationship between interest rates and inflation in Yugoslavia has never been a very close one. As a result of the determination of the nominal interest rate on deposits by cartel-like agreements among banks, the real rate of interest was negative during the entire estimation period. That has increased the importance of expected inflation for the estimation of the demand for financial assets. The dasired real stock of money at period t is, therefore, written as: A (1) log m* bo + b1 log y, + b2 iTDt + b3 pt + u, Where m* - desired stock of money, y real income, iTD - nominal interest rate on time deposits, p - expected inflation and u - random error term. The subscripts indicate the period of time. Actual stocks are assumed to adjust to desired levels according to a standard stock adjustment function: (2) log ;4 - log;.1 - d ( log mn* - log m.1) Combining equations (1) and (2) gives: (3) log m = a. 0fi a, log y, + a2 iTD, + a3 p, + a4 log M. + ut Where e1 - b, d, i - l..4 and a4 - 1 - d. The a, 's are the short- run coefficients. The long-run coefficients are obtained by dividing the short-run coefficients by one minus th.e coefficient of the lagged dependent variable: b1 - a, /( 1 - a4 ) ; i - 0..3. The semi-logarithmic form implies variable inflation and interest rate elasticities along the demand curve. The demand for/ currency, sight deposits and time deposits in Yugoslavia were estimated according to equation (3). The next section discusses the data, while the last section discusses the estimation methods and the results. 1/ Currency, sight deposits and time deposits.  6 ? Data The demands for monetary assets were estimated using quarterly data for the 1980-87 period. The data on monetary aggregates, interest rates and prices were centered in the middle of the quarter (February, May, etc) while the income variable was represented by the accumulated flow within the quarter. The data on monetary aggregates are from the quarterly bulletins of the National Bank of Yugoslavia. The domestic interest rate used in the estimations is the interest rate on one year time deposits (households' or enterprises' deposits, according to the case). This was the only short-term interest series available during the estimation period--Deposits with shorter maturities did not exist prior to 1983. The consumer price index was used to deflate the nominal variables and to measure the rate of inflation. Inflation was measured by the quarterly variations of the CPI. The CPI series are from the International Financial Statistics database. In the case of the households' demand for money, the income variable was proxied by the real personal income of households. The series on nominal personal income of the household sector are from the Index?. The real income of enterprises was proxied by two alternative variables: the industrial production index (from the IFS database) and the real revenues of the enterprise sector (from the monthly statistical bulletin). The series on the Central Bank's discount rate are from the quarterly bulletin of the National Bank of Yugoslavia. Data on exchange rates (Dr/US $ and DR/DM) and foreign prices (US and German CPIs) are from the International F'nancial Statistics database. Foreign interest rates (on one year euro-dollar and euro-mark deposits) a-e from Morgan Guaranty's World Financial Markets. 3. Estimation and Results 3.1 The Demand for Money by Households A. General Results The OLS estimates of equation (3) with household data are shown in the upper half of Table 1. The first and second rows show the estimates of the demand for currency and sight deposits, while the third row shows the estimates of the household's demand for Ml. This is a narrow definition of Ml, since it excludes the sight deposits of enterprises. Each demand equation was estimated with two alternative proxies for domestic inflation. The equations (A) were estimated with current inflation as proxy for expected future 2/ This is actually a series on wage income. Therefore, it is just a proxy for personal disposable income, since it does not include, for instance, transfers from abroad.  -7- Table I OWJTIU EUTINM OF TN M lWS' DENIS FM NM 1 YUUUV14A 13 - t7 3 Estlaisi EguatU: Im t= + I TDt*ai P t