Public sector "debt distress" in Argentina's recent stabilization efforts

PUBLIC-SECTOR "DEBT DISTRESS1 IN ARGENTINA'S RECENT STABILZATION EFFORTS. PAUL DECEERMIAN JULY 1990 Summary. Argentine efforts to stabilize inflation during 1988 and 1989 failed, and gave way to episodes of hyperinflation, be- cause the public sector slipped into a debt-distress trap. In their stabilization programs the authorities sought to anchor the price level through a high exchange-rate value, which they defended by running policies that kept domestic interest rates high. Since the public sector had significant domestic debt, however, the high interest rates drove the public sector*s in- terest bill above its non-interest surplus. The public sector .could therefore cover its interest bill only by taking on addi- tional debt, and so its domestic debt grew unsustainably rap- idly. Hyperinflation resulted when the debt became larger than domestic financial markets could be persuaded to hold. The author gratefully acknowledges comments by Werner Baer, Donald Coes, Roque Fernandez, Dale Gray, Ricardo Lago, Millard Long, William Mayville, Richard Newfarmer, Martha Preece, Peter Scherer, Eduardo Somensatto, and William Tyler on earlier drafts. Further comments are invited. Please do not cite this work without first consulting the writer. The views expressed are the writer's, not necessarily the World Bank's, nor of any other institution with which this writer has been associated. 1. Zatroduction. 1.1. On Sunday, July 9, 1989, one day after Carlos Menem suc- ceeded RaUl Alfonsin as Argentinats President, the incoming eco- nomic authorities announced a program intended to stop the hyperinflation that had run through the second quarter of the year. That hyperinflation had come about after the previous stabilization effort, the August.1988 Primavera Plan, coll%psed. Although a Peronist, the new President had drawn his economy team from the executive rapks of the multinational grain-trading firm of Bunge y. Born, and the new program was said to reflect the firm's macroeconomic analysis. Accordingly, the Buenos Airas press christened the program "the BB Plan.*" 1.2. Like several other recent heterodox stabilization plans in Argentina and Brazil, the BB Plan worked at first. The ini- tial measures were a devaluation from 303 to 655 australe per dollar, massive public-enterprise price increases, and an &gree- ment by the largest industrial enterprises to freeze their prices as long as the exchange rate and the public-enterprise prices remained frozen. Inflation eased to manageable single-digit monthly rates. For about three months the parallel and fixed official exchange rates remained virtually equal. The Central Bank*s depleted international reserves recovered rap- idly. Interest rates on seven- and fourteen-day commercial-bank time deposits -- to which the rates on most of the public-sector debt were linked - fell gradually from 15 per cent per month at the outset of the program to 4. 5 per cent in early October. The Government submitted legislation to the Congress for ambitious taxation, privatization, and expenditure reforms. In Septem- ber 1989, domestic macroeconomic conditions were sufficiently favorable to enable the Government An4 the International Mone- tary Fund to negotiate a stand-by program, and the IMF made an initial US$233-million disbursement in November following its Boardts approval. 2.3. In aid-October, however, the parallel exchange-rate premium began to drift upward. Expectations of further devalua- tion induced deposit withdrawals, and interest rates rebounded sharply as banks struggled to retain their deposit base. Finan- cial-market turbulence deepened through November. The Central Bank sold foreign exchange, but faile to reverse the tide against the exchange rate. By early December, with the parallel . premium ranging between 35 and 50 per cent and monthly interest rates back up to 15 per* cent, the program had clearly become un- tenable. On December 10 the authorities admitted defeat: they devalued the official rate from A$655 to A$1010 per dollar, raised public-sector prices an average of 65 per cent, raised xport taxes, and raised public-sector wages. When financial markets. reopened the ,sae week, however, interest rates remained high, and the same percentage parallel exchange-rate gap 're- emerged. The Bunge y Born economic team therefore resigned. -2- 1.4. A new economic team took office on December 18. It in- mediately floated the exchange tate, removed all price controlps and rescinded the export-tax increases. These measures failed again to secure stability: interest rates continued to drift up and the austral continued to sink. Over New Year'?s weekend the anatral depreciated almost fifty per cent against the U.S. dol- lar. The price level roughly doubled over December. 1.5. On New Year's Day, 1990 Argentina*s Economy Minister announced dramatic measures to halt the developing hyperinfla- tion. The principal action was the forced conversion of commer- cial-bank time deposits into ten-year dollar-denominated Nation- al Treasury "External Bonds" ("BONEX"). Approximately US$500 of each account were exempted from conversion and were to be made available in cash. Because the BONEX traded in domestic finan- cial markets at a heavy discount, this "BONEX Plan" constituted a substantial confiscation of private asset holdings. Compul- sory and voluntary holdings of Central Bank and National Trea- sury obligations, yielding interest at rates closely linked to the deposit rates, made up an unusually large proportion of the commercial banks' assets. Under the Plan, the Treasury pur- chased these obligations with BONEX, and the commercial banks then used the BONEX to make the conversion. 1.6. The point of the measure was not merely to erase li- quidity, but, more important, to end the massive public-sector borrowing flow deriving from the interest on the Central Bank and Treasury obligations. The BONEX conversion was the authori- ties* response to their perception that the public sector had slipped into acute "debt distress." Indeed, the failed hetero- dox stabilization attempts leading up to the BONEX conversion had failed, and led to hyperinflation, precisely because they had induced such distress. -The .reasons why the BB Plan failed - and indeed why so many Argentine stabilization efforts have foundered since the mid-1980s - remain, perbaps inevitably, highly controversial. Observers attribute varying degrees of importance to wage pressure,- to persisting non-financial public sector deficits, to persisting Centkal Bank deficits, to devalu- ation and the consequent external trade surpluses, and to sudden crises of confidence in the markets where interest rates and ex- change rates are determined. , These factors all contributed to reviving inflationary pressure. Nevertheless, to understand why the BB Plan collapsed, it is more helpful to consider them as events linked in a dynamic inflationary process, rather than as independent contributing factors,. 'This essay argues that the dynamic process was inherent in the framework of the BB Plan, and .that it was also inherent in the Alfonsin Government's August 1988 "Priavera (Spring) Plan." 1.7,. I the essential argument is as follows. Like the Primavezi Plan - and like other recent heterodox stabilization programs, such as Argentinass Jutie 1985 Austral Plan, Brazil's June 1987 Bresser Plan, and Brazil's January 1989 Summer Plan -3- ("pla.n Ag er.ao") - the BB Plan began with a substantial, pre- sumably corrective, devaluation. The idea was to use the fixed exchange rate as a temporary price-level "anchor", to set a con- text within which to carry out profound public-sector reforms and so set a basis for lasting price stability. In the short term, exchange-rate stability was to ensure price stability by (i) "anchoring" prime cost, not. only by setting the price of in- portable inputs but by setting a referent for wages and finan- cial rates of return; (ii) reducing competitive pressure on tradeables prices generally; and .(iii) reducing devaluation ex- pectetions, thereby strengthening people's willingness to hold assets denominated in austX*a rather than U.S. dollars. 1.8. The problem with this approach was that, in combina- tion, (i) the public sector's continuing credit demand, (ii) persisting exchange-rate uncertainty, and (iii) financial markets* perception of Argentina as risky implied that domestic interest rates would have .to remain exorbitantly high to sustain the fixed exchange rate. High interest rates placed economic policy in a no-win situation, however. Even after the Febru- ary-July 1V89 hyperinflation, .the domestic debt stock of the combined public sector - essentially, the Treasury and the Cen- tral Bank - remained substantial. High interest rates implied a high -public sector interest bill, because the debt was almost entirely at floating interest rates at teras of one to two weeks. 1.9. Moreover, to keep interest rates high, the monetary au- thority itself had to place interest-bearing debt in domestic financial markets - that is, to sterilize the money created through (i) foreign-exchange operations at the devalued exchange rate and (ii) the monetary authority's financing of its own in- terest payments. The interest bill substantially exceeded the public sector's primary (non-interest) surplus. This meant the public sector could pay its interest only by capitalizing it into its domestic obligations - that is, into the anstral-de- nominated assets. The austral-denominated asset stock grew rap- idly through the interest capitalization. As such assets became increasingly abundant relative to dollar-denominated assets, the austral came under pressure to devalue in the parallel market. In sum, the public sector could not afford the interest rates required to maintain the value of the auatral.. It tried, never- theless, to pay them by increasing- its debt; but this failed, one may infer, once private-sector portfolios became- saturated with a-ustra-denominated debt. 1.10. The only way to forestall the growth of public debt would have been for the public sector actually to pay the inter- est on its debt in cash rather than by .borrowing through capi- talization. This would have required the public .sector to run a non-interest cash surplus dufficient to cover the interest bill. The incoming authorities understood the problem if these terms. The initial public-enterprise price increases were intended to meet this objective. Part of the point of the price-freeze agreement, and the resulting decline in the inflation rate, was to produce a favorable Olivera-Tanzi effect, raising tax re- ceipts as a proportion of GDP. Indeed, as the sharp devaluation encouraged exporters to surrender export proceeds, it not only helped rebuild international reserves but also restored the export-tax base. These measures reduced the public sector*s non-interest deficit, and with several further measures, the Treasury secured a non-interest surplus by November. This sur- plus remained insufficient, however, to cover the interest bill, particularly after interest rates rebounded in October. By that point the authorities were in a losing race, struggling to in- crease the non-interest surplus to overtake the surging interest bill. Once the financial markets concluded that they could not win, the exchange rate surged, and the program failed. 1.11. The two following sections describe the historical and institutional background of the 1989 hyperinflation episodes. Section 2 summarizes the reasons why inflationary pressure mounted through the 1980*s despite stabilization efforts, and Section 3 focuses on the unusual intermediation role that the Central Bank of the Argentine Republic (BCRA) had come to play in the macroeconomy. Section 4 reviews the August 1988 Prima- vera Plan and macroeconomic events in late 1988 and early 1989. Section -5 describes the BB Plan in more detail, and macroeconom- ic events between July 9 and December 28. Section 6 briefly discusses some of the parallels of the BB and Primavera Plans. Section 7 briefly discusses the measures taken since DecAm- ber- 1989, including the BONEX conversion. Section 8 offers some concluding observations. 2. Argentine inflation in the 19700s and 19800s. 2.1. The sacroeconomic m rhaniss' that powered the 1989 Lyperinflation episodes developed along with the chronic infla- tion that has plagued Argentina for decades. Some of these mechanisms - notably, dollarization and the various automatic money-creation mechanisms described in the section following - were means that public and private institutions developed to cope with and defend themselves from inflation. The chronic in- flation of the 1970es and 1980ts, along with the concomitant price-level uncertainty and price-system disarray, contributed heavily to the debilitation of the real economy. Over the 1980s, - real annual growth rates averaged only- 1. 2 per cent - annual industrial output fell at an annual avdrage rate of 1.5 per cent (although annual industrial growth rates varied sharply); real per-capita GDP fell at an average annual rate of 0.4 per cent; and per-capita real consumption levels diminished at an average annual rate of 1.9 per cent. -Capital formation followed a clear downward trend as a proportion of real GDP, from 23.7 per cent in 1980 to 12.5 per cent in 1988. Inadequate capital foration has contributed to declinirg productivity, and has. inevitably affected the economy's* international competitive- ness. -5f 2.2. Argentina had a brush with hyperinflation in the mid-1970*s when the Peronist government lost control of the economy: consumer prices quintupled between March 1975 and March 1976 (see Table 3)., largely as a result of the massive . August 1975 devaluation. The armed forces overthrew this gov- ernment in March 1976 . Thereafter, Argentina*s macroeconomic trajectory bad clear parallels with other externally indebted economies with serious inflation, notably Brazil, Chile and Peru. - Credit was then available to both the private and public sectors from international financial markets at what were then low interest rates. Since the new military regime intended to promote economic liberalization and private-sector growth, it had no qualms about permitting the private sector to borrow the resources it needed overseas. It liberalized both entry into the financial system and interest rates to enable the system to capture domestic and foreign resources. 2.3. The public-sector borrowing deficit remained high be- cause the authorities found it easier to borrow than to raise taxes or engage in the political conflicts a thorough pub- lic-sector reform would have entailed. The armed forces, who were engaged in a violent struggle against guerrilla movements, were unable to design nor carry out profound economic reforms. Their economic policy consisted of liberalizing domestic and ex- ternal financial flows and allowing the banking system to ex- pand. Some employment reductions apart, they made few efforts to improve public-sector efficiency. Their laisse&-aire ap- proach did not include ef fective financial-sector supervision. The liberalizing 1977 Financial Entities Law permitted banks full freedom to set interest rates, although it established a system of Central Bank deposit insurance. A deliberate policy of maintaining a high real effective exchange rate in the late 1970*s (see Table 4) - the pro-announced "IabLita (schedule) - provided an incentive for a positive resource transfer (non-factor imports less exports). This was the means by vhich the financial inflow, which took the form of a build-up of pri- vate and public external debt, was transferred as real goods and services into the econamy. 2*4. Liberalization combined with external borrowing left the economy vulnerable when international credit markets tight- ened and interest rates surged in 1980. As its interest bill soared -and it lost access to "voluntary" external - financing, Ar- gentina devalued heavily .to increase its trade surplus and so . effect the transfer. The devaluation not only increased the domestic-currency equivalont of the public sector's external in- terest bill, but made it more difficult for private borrowers to service their external debt. At the saMe time the devaluation gave a powerful impetus to inflation. Precisely because the higher external interest charges,,..devaluation and inflation worsened the public-sector deficit, they worked against the au- thorities* attempts to compensate for the inflationary effects of the devaluation by controlling public finances. 6 2.5. These problems were intensified by a banking-sector crisis that resulted from high interest rates, excessive over- seas borrowing, rapid lending growth and inadequate bank super- vision. (See BaliWo 1987.) The crisis began in March 1980 with the failure of a large, overextended bank, and continued over the next two years with the failure of 70 other private finan- cial institutions. This had inflationary consequences as well: since the monetary authority guaranteed bank deposits and could not rapidly liquidate the banks (judicial and administrative delays stretched the liquidation processes as long as a decade), it had to create money to pay depositors. - The macroeconomy slipped out of control through 1982 in a spiral of devaluation, inflation, mounting external debt, and corporate debt distress. Capital flight, encouraged by expectations of devaluation, high world interest rates, and diminished domestic prospects, became a significant problem in the early 1980*s. The conflict with the United Kingdom in April 1982, which forced a sharp increase in public expenditure, and the eruption of the international debt crisis in August 1982, compounded Argentina's crisis. 2.6. Neither the banking system nor the public sector could offset the inflationary pressure resulting from the rounds of devaluation. In 1982 the public sector assumed the private see- tor's external debts the devaluation and high world interest rates had made the debt service extremely costly, and foreign creditors would zot voluntarily roll it over unless the public sector assumed it. At the same time the authorities addressed the corporate-distress problem by requiring camercial banks to reschedule outstanding debt over 60 months at a monthly interest rate of 4 per cent, well below the expected -inflation rate. To forestall commercial-bank decapitalization the authorities in- posed controls on deposit,rates for the first time -since 1977. These measures temporarily relieved the domestic distress prob- len, at the cost of renewing financial-system repression. 2.7. The defeat in the war. -discredited the armed forces, and a transitional military regime took power in aid-1982 to prepare elections. In Deceaber 1983 Raul Alfonsin took office at the head of a constitutional government. Inflation worsened over 1983, 1984 and the first half of 1985 as the authorities strug- gled simultaneously to revive real growth and to restore policy control. At the same time they engaged in a series of confron- - tations with Argentina's commercial-bank creditors: on two occa- sions they made interest payments only at the last possible moment before triggering the U.S. regulators' non-accrual rules, and then only on the basis of funds advanced by other Latin American nations or by the U.S. Treasury. Toward the end of 1984, however, Argentina secured a concerted new-money and re- sche4ling agreement with forign commercial banks, after agree- ing to an IMF program. The round- of devaluations and negative per-capita GDP growth in 1982 and 1983 turned the trade balance from positive to negative and even reduced the current-account deficit, in spite of the higher interest bill (see Table 1). -7- The diminished current-account deficit relieved external-debt growth, but the negative resource transfer reduced domestic cap- ital formation and deepened inflationary pressure. 2.8. Over the first half of the 1980's, declining tax rove- nues and rising interest expenditures overwhelmed substantial expenditure cuts and sharply reduced public-sector saving. Gross tax revenues fell because tie economy stagnated and be- cause tax administration deterlotated. Subsidies, principally the costly regional "industrial-promotion" program, rose some- what. Public-sector intereqt payments to domestic creditors began to rise: with external finance reduced and the public sector still having to borrow, domestic debt increased and the public-sector domestic Interest bill came under pressure. Mean- while, the state-owned non-financial enterprises, which operate such basic industries as international air transport, rail transport, petroleum production, and communications, developed heavy operating losses over extended periods of time, through the combined consequences of policies regarding output prices, employment and investable-funds allocation. (See Table 2.) 2.9. Argentinaos state-owned term-financing institutions -- the housing bank and the development bank -- have run heavy losses, partly because of high operating costs, but also because macroeconomic instability made term financial operations inher- . ently unprofitable. Price-level and exchange-rate uncertainty shrank deposit terms to weeks and even days; moreover, financial institutions had to pay high real interest rates to compensate savers for noiible,, not merely eece,, inflation and devalua- tion. At the same time relative-price dispersion and income volatility -affected financial institutionst asset quality. In these conditions -private financial institutions withdrew from term lending, but public financial institutions carried on, partly with BCM financing, and inevitably incurred losses. The commercial banks owned by Argentina's provincial governments have also had high operating losses and troubled portfolios, partly because they are called upon to finance activities - no- -tablyagriculture - that private banks have avoided. Some pro- .vincial banks provided financing to their governments, and some used BCRA rediscounts to finance themselves. Finally, the BCRA itself - owned entirely by the Government -- has run substan- tial losses. As discussed below, the BCM's losses played a pivotal role in the 1989 hyperinflation. 2.10. As a consequence of their lengthy experience of infla- * tion and financial volatility, private Argentines' willingness to hold money and other publid-sector obligations has diminished markedly. The narrow money supply fell from 7.5 per cent of GDP * in 1980 to only 3.2 per cent of GDP -in 1988, while the broad money supply - including interest-bearing =:ammercial-bank sav- ings and -time deposits - fell from 28.4 per cent of GDP in 1980 to onIy 14.7 per cent of GDP .in 1988 (Giorgio 1989). Antici- pated devaluation assumed high importance as residents came rou- tinely to compare rates of return available domestically with those available overseas. 2.11. The economy's diminishing willingness to hold money and other public-sect9r obligations has had a number of practical consequences for stabilization policy. First, any given pub- lic-sector borrowing requirement has become harder to finance and so causes larger macroeconomic problems. If the public sector borrows at competitive interest rates, then, the smaller the domestically available savings stock, the higher the inter- est rates will be .-- hence the higher subsequent public-sector borowing requirements will be and the smaller the quantity of financial resources 'available for the private sector will be. If the public sector finances itself by issuing (non-inter- est-bearing) money, then, the smaller the demand for money, the higher the resulting inflation will be. (For economies like Ar- gentina it may therefore be more helpful to gauge the public-sector borrowing requirement against the broad money supply rather than GDP.) Second, similarly, following a devalu- ation, the money creation resulting from the sale of foreign ex- change to the central bank has a larger percentage impact on the money supply. Third, the monetary control measures required to offset any monetization of public deficits and devaluation pro- ceeds tend to fall heavily on the banking system's supply of private-sector credit. Finally, a narrow willingness to hold money and other public obligations reduces the margin for error in any stabilization program. For example, suppose a stabiliza- tion program is premised on a monthly reserve inflow of US$50 million, -but the inflow turns out instead to be - US$100 million; the monetary consequences could be marginal if the total money supply were (say) US$10 billion but devastating if the money supply were US$1 billion. 2.12. It may help here to summarize the br"ad argument about the origins of Argentina's inflation problem in-the early * 1980's* (i) After 1976 the military regime encouraged domestic liberalization and private-sector growth. It permitted private financial institutions rapidly to expand their activities at free interest rates. (-ii) With credit available at low, but ad- justable, interest rates from international sources, the public and private sectors borrowed freely. In particular, the low in- ternational interest rates made improvements in public-sector efficiency less urgent. (iii) To facilitate the resource trans- fer, the government maintained the exchange rate at a relatively high value. In the early 1980's, however, (iv) :a domestic fi- nancial crisis sharply increased domestic interest rates. (v) Rising international interest rates and tightened credit conditions siaultaneously increased the external interest bill and made it barder to finance. (vi) The authorities devalued to reverse the external transfer; :this was not only inflationary, but. worsened both the public and private sectors* borrowing re- . quirements. Prospects of further devaluation also discouraged holdings of domestic assets, and so contributed to capital flight. (vii) The public sector nevertheless assumed the pri- -9- vate sector's debt, and then (viii) controlled domestic interest rates, in order to limit the banking system's and private firms* debt distress. This diminished people*s disposition to hold banking-system obligations, which contributed to further infla- tion and further complicated policy management. 2.13. When the Alfonsin Government took office in Decem- ber 1983, it inherited an unprecedented combination of external debt, inflation, and economic stagnation. The brusque shifts in month-over-month inflation rates shown in Table 3 trace out the Alfonsin Governments heterodox stabilization efforts. Over the first fourteen months of Alfonsin's term the price level accel- erated, reaching monthly rates between 25 and 30 per cent in the second quarter of 1985. In early 1985, a new economy team took office and prepared the "Austral" Plan, which the President an- nounced in a television address in mid-June. The Austral Plan was a stabilizing "shock" combining (i) a wage and price freeze; (ii) devaluation, followed by establishment of a fixed exchange rate; (iii) creation of a new currency (contracts denominated in the old currency coming due over the rest of 1985 were converted to the new currency according to a schedule that removed their presumed inflationary-expectations component)i and (iv) promises of fiscal reform, including an undertaking that the BCRA would henceforth not issue money to finance Treasury deficits. 2.14. The Austral Plan dramatically lowered the inflation rate for the remainder of 1985. The authorities reduced the non-interest public-sector deficit remarkably and maintained tight monetary policy, as indicated by the sharply higher real interest rates (see Table 1). The balance of payments strength- ened, with the current account actually in surplus over the second half of 1985. Industrial growth resumed toward the end of 1985 and there was a round of wage increases in the private sector, which permitted some recoveiy in sharply diminished real-wage levels. The wage increases attected public expendi- tures, and in retrospect. a renewed fiscal effort was necessary to offset them, but the Government did not attempt one. 2.15. Inflation remained relatively low well into 1986, but sufficiently high that by April 1986 frozen prices had to be liberalized to permit correction of the relative-price array frozen into place in June 1985. Public-sector prices had to in- crease to ensure public enterprises* financial soundness. The authorities also devalued in an attempt to catch up with the in- flation that had accumulated since June 1985. The monetary au- thority made a strenuous effort to restrain monetary growth over 1986 and early 1987, but after the corrective measures of. mid-1986 it had to contend with declining money demand. In Sep- tember 1986 the authorities tightened monetary policy again and renewed price controls. .2.16. The price level continued to drift upward at increasing rates through early 1987. Legislative and gubernatorial elec- tions in September of that year generated heavy spending pres- - 10 - sures in the second and third quarters. Meanwhile, the terms of trade - which had been falling secularly since the beginning of the 1980s -- fell sharply and reduced the trade balance. once the elections were over, the Government announced a sharp deval- uation, liberalization of interest rates, and fiscal-reform mea- sures. Again, the effects of these measures were short-lived: inflationary pressures reemerged early in 1988, and by aid-1988 hyperinflation threatened yet again. 2.17. In early August 1988, the Government announced yet an- other heterodox stabilization program, the Primavera ("Spring") Plan. This program is discussed in Section 4. The section fol- lowing digresses to discuss the complex of relationships cen- tered on the BCRA, among the Treasury, publicly-owned banks, commercial banks, and the wealth-holdir.g public that character- ized the late 1980s, and through which the stabilization and hyperinflation of 1988 and 1989 played themselves out. 3. The role of Argentina' s Central Bank. 3.1. The relationships among Argentina*s Treasury, Central Bank; publicly-oWned banks, commercial banks, and wealth-holding public during 1988 and 1989 may be characterized as follows. (i) The Treasury and the Central Bank taken together owed the domestic financial system amounts varying between US$5 and US$8 billion in interest-bearing debt. (GDP was between US$60 and US$72 billion .a. year, depending on the exchange rate used, or US$5 and US$6 billion a month.) (ii) of this debt, about three quarters - between US$4 and US$6 billion - were interest-bearing debt of the BCRA to commercial banks. This debt comprised (a) rea_nerated bank reserves, (b) "inaccessible" commercial-bank deposits at the BCMA, and (c) marketable SCM bills. in addition, (iii) the ZCM required conventional unremunerated bank reserves of 88.5 per cent against ordinary checking accounts.. Understandably, given the inflationary cir- cumstances, the Argentine public held only the barest minimum in checking accounts for transactions purposes. (iv) Meanwhile, the bulk of the BCRA's own asset position was illiquid or - in reality though not in its accounts - value-impaired. The BCRA's assets included unrecoverable rediscounts to public- ly-owned and. liquidated banks and, after 1988, natured National Treasury obligations, which the BCM had amortized because the Treasury could not. 3.2. (v) In effect, the commercial banks financed their holdings of CM - and Treasury debt by taking -deposits - mainly interest-bearing -time deposits - from the private sector. Since October 1987, comercial banks' liability interest rates have been fully free. In turn, the BCRA paid interest on its obligations to the commercial banks' at rates. equal to the rates on a representative sample of commercial-bank time deposits plus a small spread.B = rdingly, the BCRAs interest was based. on the freely determined commercial-bank time-deposit rates. The - 11 - commercial banks paid whatever interest rates they had to pay to maintain their time deposits, which depositors might otherwise have placed in dollars, passing the cost plus a spread back to. the BCRA. 3.3. (vi) In principle, the SCM could finance the resulting interest bill by (a) creating money; (b) using the proceeds from interest or amortization paid on BCRA assets; or (c) borrowing, either by issuing new debt or by capitalizing the interest due into the debt on which the interest was being paid. The first option would have been inflationary, and would have generated devaluation pressure in the parallel exchange market. The second option was limited by the reality that a large part of the BCRA's assets - obligations of banks undergoing liquida- tion, of publicly-owned banks, or of the Treasury - paid no in- terest, let alone amortization, in cash. The BCRAss only alter- native was therefore to finance its interest bill by borrowing. 3.4. The Argentine system has had two additional features of fundamental importance: (vii) the BCRA has insured the banking system*s deposits and (viii) the superintendency of banks, which has been run by the BCA has functioned ineffectively. Commer- cial banks have paid a fee for .the deposit-insurance service (amounting to A$0.0003 per aaLail of deposits per month). In effect, the BCRA has insured the deposits, and the interest cap- italized into them, through its money-issuing capacity. The . weakness of the superintendency has compounded the problem be- cause the SCM has been unable to detect commercial-bank defi- ciencies before they require -intervention. BCRA intervention has been tantamount to liquidation, because depositors withdraw their deposits once they hear of intervention; since liquidation is: a lengthy legal process, the BCA tends rapidly to create money to pay depositors, acquiring "rediscounts" to the failing . institutions in the process. 3.5. Table 5 shows the 1988 and 1989 month-end stocks of interest-bearing Treasury and BCRA debt to the domestic economy, .including the commercial banks. (This total probably consti-. tuted more than 90 per cent of the public-sector debt to the do- nestic economy at any moment, although compiled data on other categories of public-sector debt - including state-enterprise commercial debt and obligations of provincial and municipal gov- ernments - are unavailable.) The table indicates that the stock of Treasury and BCRA debt -to the private sector has run at about 8-10 per cent of GDP, a relatively low figure by compari- son with other nations. (The comparable Brazilian figure in the 1980*s has been around .35 to 40 percent, for example.) Never- theless, Argentine finaicial markets have been increasingly re- luctant to hold even this low public debt stock, as evidenced by the high real interest rates they carried. 3.6. The narrow m9netary base - currency in circulition and unremunerated bank reserves - fell from 5.2 per cent of GDP in 1986 to 3.4 per cent in 1988. In addition, savings and time de- -12 - posits at commercial banks -- M4 less 1 in the Argentine defi- nitions -- were between 8 and 11 per cent of GDP in 1988 and 1989 (see Table 6). The BCRA has required commercial banks to maintain a high but considerably lower ratio of remunerated re- serves against these deposits: over 1988 and 1989, the ratio was between 22 and 25 per cent. With high reserve requirements, re- muneration was intended to ensure that banks* earnings on their assets were sufficiently high to prevent their charging high rates for ptivate credit. 3.7. The BCRA began to use "inaccessible deposits" (some- times called Oforced investments") in early 1985. Inaccessible deposits were similar to remunerated bank reserves, in that they were remunerated deposits of commercial banks at the BCRA, but with the important difference that commercial banks could not withdraw inaccessible deposits, even if the banks' own deposits had fallen. (There was a partial exception to this rule: after .August 1988 banks could maintain one kind of "inaccessible de- posit" as a daily average balance over each month.) Moreover, interest paid by the BCRA on -the inaccessible deposits could not be withdrawn, but had to be capitalized into the deposit bal- ance. Beginning in 1985, the BCRk required commercial banks to constitute inaccessible deposits from time to time, announcing them in "Comanicadoz* that set a deadline for the banks to do so, typically as proportions of the banks' time deposits as of a certain recent date. 3. 8. Inaccessible deposits were the BCR's solution to two problems that arose with high and remunerated conventional bank reserves. The first was that the yield on remunerated reserves was less than commercial banks could earn on alternative lending operations. An incentive to.disintermediation arose: financial -groups tended to allow their oammercial banks' deposits to di- minish and channeled their resources into Argentina's active interfirm financial market. The second problem was that inter- est payments by the BCA on its reserve liabilities were, in themselves, a source of monetary expansion. To pay interest on conventional bank reserves, the BCRA credited the interest to the commercial banks* reserve accounts, thereby directly in- creasing the monetary base. The inaccessible-deposits system solved this problem partially by capitalizing the interest inac- cessibly into the deposit balance, .rather than making it avail- able to the banks in the form of high-powered money. 3.9. The third form of interest-bearing BCRh obligation was SCM bills, marketable seven-day obligations sold at a discount in daily.auctions.. These instruments were known as "Certifi- cates of Participation" (CEDEPs, because they were presumably participations if the BCM's holdings of Treasury obligations) and also as "Telephone Bonds" (because they were often sold in telephone go-around auctions). 3.10. On its asset side, the BCRA has held *a mix of obliqa- tions of private commercial banks, financial institutions in -13- liquidation, publicly-owned banks, and the Treasury. A common characteristic of these assets has been that the BCRA could not use then to absorb liquidity. They were not marketable, and so could not be used to absorb money in open-market operations. In addition, many were in reality non-performing, so that while the BCRA recorded accrued interest (under accounting standards that prevented the BCRA from placihg its government obligations on non-accrual *status), it received very little in cash. Even the BCRAss holdings of Treasury bonds were mostly unusable in open-marXet operations. Beginning in early 1988, as Treasury bonds issued in 1987 came due, the Economy Ministry instructed the BCRA to amortize them; since the bonds had matured, they were unmarketable. The BCRA received some interest in cash from "rediscount" credit to commercial banks, particularly on opera- tions associated with exports, but it could collect no interest in cash from banks in liquidation nor from the public-sector banks engaged in loger-term operations -- the National Housing Bank (BEN) and the National Development Bank (BANADE). 3.11. Table 7 gives the averages of the BCRA's interest-bear- ing asset and debt stocks during the two years 1988 and 1989, as a proportion of GDP. Argentina's Central Bank has undergone progressive decapitalization, in at least two senses. First, a central bank needs to hold a quantity of assets sufficient to back its monetary obligations. Precisely because Argentina's Central Bank possessed insufficient assets to use in open-market operations, and received only limited flows of cash payments, it could absorb money only by issuing debt. Unlike absorption through receipt of interest in cash, however, absorption through debt issue committed the BCRA to pay interest, and this interest entailed a financing need. Second, the BCRA has been decapitalized in the sense that its accruing interest expenses have generally exceeded its accruing interest receipts. The measure of the flow rate of decapitalization would be the BCRA's losses on an accrual basis. (The appendix discusses the analyt- ical issues involved in this measurement. See also Barbone and Beckerman 1989.) Table 7 gives an indicator of this aspect of the BCRA's decapitalization, the domestic "quasi-fiscal deficit" as defined for use in Argentine IMF programs. This is the dif- ference between interest paid on the BCRA's interest-bearing li- abilities and interest received on the BCR's interest-bearing assets, less the profit received. by the BCRA to the extent in- flation reduces the real value of its not interest-bearing lia- bilities. . 3.12. Since the liberalizing reform of October 1987, interest rates an commercial-bank time deposits have been freely deter- mined - although .the deposits, and the interest accruing on . them, were fully insured by the Central Bank. Argentine resi- dents have easy access:to U.S. dollar assets. Discrepancies be- tween debt-stock and balance-of-payments.jestimates suggest that Argentine residents have acculated a quantity of flight capi- tal on the order of magnitude of the nation's US$60 billion ex- ternal debt. They are estimated to hold between US$5 and - 14 - US$7 billion in dollar currency (more than double the narrow money supply); and they hold substantial quantities in foreign accounts. Because dollars are easily available, anticipated de- valuation has tended rapidly to pressure commercial banks to raise their deposit rates. At the same time, by encouraging people to hold &Ustral-denominated assets, higher interest rates tended to reduce the pressure on .the anatral to devalue in the parallel exchange market. The basic problem, however, was that the bankrupt monetary authority had to pay these higher interest rates. 3.13. To understand how the monetary system evolved into this structure, a brief summary account of its development from the mid-1970s may be helpful. From mid-1973 to early 1976, the Peronist government attempted to centralize banking intermedia- tion. It effectively nationalized the banking systems deposits by means of a 100 per cent reserve requirement: commercial banks were directed to pass deposit proceeds to the BCRA and to pro- vide credit on the basis of BCRA rediscount allocations, all at regulated interest rates. The government intended to use the BCRA to direct financial mobilization and allocation, but this approach proved unworkable because the credit allocation had no rational price basis and because the authorities could not con- trol inflation. 3.14. Since the March 1976 military coup, Argentina's govern- ments have generally intended to liberalize, but have .frequently had to reverse their liberalization efforts. In 1977 the mili- tary regime enacted a liberalizing reform, embodied in a new "Financial Entities Law." This reform returned banking to a fractional reserve system, freed interest rates in the formal financial system, and removed barriers to entry. The authori- ties reduced the required reserve ratio to 45 per cent, fearing that a lower ratio would prove immediately inflationary. This was when the BCR began the practice of. paying interest on re- serves to.enable banks to maintain relatively low spreads be- tween lending and deposit rates. It funded this interest by collecting a monthly fee equal to one per cent of banks' "loan- able capacity," given their* reserve holdings. The BCRA accumu- lated the fee proceeds in a fund called the "Monetary Regulation Account," and paid the interest out of this fund. 3.15. For about three years this system, which was intended to be temporary, worked well enough. The authorities reduced reserve requirements to 10 per cent by 1980, and the Monetary Regulation Account remained in surplus. Banking-system opera- tions expanded in response to the liberalized envirorment. The private sector took on growing volumes of internal and external. debt. Unfortunately, the. macroeconomic policies at this time, centered on maintenance.of. high' exchange rate, required high domestic interest rate levels to. attract and .hold foreign capi- tal. When this policy .proved unsustainable, as described above, devaluation, inflation, and recession cambined to thrust the: * 15 - commercial banks into profound crisis, beginning with a wave of failures in 1980. 3.16. To cope with the crisis, the authorities reluctantly took policy measures that reversed the 1977 liberalizing re- forms. The BCRA increased bank reserve ratios; it relieved the banks from having to pay the fee that funded the Monetary Regu- lation Account, although a substantial part of their reserves continued to pay interest.. The crisis deepened with the exter- nal financing crisis. During 1982 the BCR assumed the private sector's external debt, announced a blanket rescheduling of pri- vate sector loans to domestic commercial banks, and reintroduced of controls on deposit interest rates. These measures permitted the private sector to recover financially, and prevented the complete decapitalization of the banking system. The cost, how- ever, was that the financial system was reconstituted on a re- pressed basis. (See BaliWo 1987.) Unable to attract signifi- cant deposits, the. private banking system had to narrow its lending activities. Disintermediation resulted, in the form of a revitalized interfirm financial market. While the interfirn market relieved the credit scarcity, it was entirely unsuper- vised, and it could provide credit only to large, well-established enterprises. 3.17. The Alfonsin Government intended to resume financial liberalization after the Austral Plan. Through 1986 and 1987 the authorities planned liberalizing reforms, but these plans remained in abeyance irhile the Central Bank maintained tight conditions to keep inflation from reviving. In October 1987, the authorities freed interest rates completely, as part of the package of devaluation and stabilization measures following the mid-term legislative and gubernatorial elections. Unfortu- nately, public-sector credit needs remained substantial, and the stabilization burden continued to fall on* the monetary author- ity. With interest rates free to rise, high bank reserve re- quirements, remuneration of reserves and tight monetary policy proved self-defeating, because the BCRA's own debt to the com- mercial banks accumulated. Since the Monetary Regulation Ac- count had no income, but paid interest on remunerated bank re- serves and inaccessible deposits, it became a standing source of money creation. The BCM could "sterilizeg it only by issuing more interest-bearing debt to commercial banks, in a self-per- petuating debt creation spiral. - Having lost all other means of absorbing liquidity, the BCRA could only bear down heavily on the banks. Meanwhile, although the Treasury's demand for CMA . credit remained under the Austral' Plan prohibition after 1985, the public sector banks - the .MM, the BANADE and some of the provincial banks - came to require heavy financing. In the ab- . sence of other funding sources, the BCM was asked to supply fi- nancing to them, mainly in the form of "rediscount" credit. t 3.18. In early 1988, despite their intentions to reduce bank reserves and inaccessible-deposit requirements, the authorities chose to maintain and even increase these requirements to fight - 16 - inflation. Part of the problem in mid-1988 was that the balance of payments suddenly became a source of inflationary pressure: unusually hot summer weather had damaged U.S. grain and soybean production, and the authorities had to sterilize significant international-reserve inflows resulting from high prices and export volumes. By July 1988, their tightened monetary policy notwithstanding, the authorities found themselvep once again facing the prospect of hyperinflation. 3.19. The essential characteristics of Argentina's peculiar macro-financial system were in place before the Primavera Plan. Demand for both narrow (unremunerated) and for broad (remuner- ated) money, which competed with readily-available dollar-denom- inated assets, was low. Broad money could compete with dollars only by offering high interest rates. Broad money, however, was SCRA debt, intermediated at one remove by the commercial banks. Moreover, the BCRA had no means of paying interest, save through creation of debt that the economy was unwilling to hold except at high interest rates. When the BCRA chose debt creation rather than money creation, it nevertheless committed itself to expand the broad money supply. Tight monetary policy therefore became perversely inflationary. 4. The Primavera Plan. 4.1. In August 1988 Argentinats economic authorities, alarmed by accelerating prices over the first seven months of 1988, attempted to renew their heterodox stabilization strategy through what came to be known as the "Primavera Plan." The core of the Plan was a devaluation, accompanied by introduction of a multiple exchange-rate system through which the BCRA would earn an operating profit by selling foreign exchange at the paral- lel-market rate and purchasing it at an official rate. Monetary policy would be calibrated to set domestic interest rates at whatever level was necessary to hold the parallel rate between 20 and 25 per cent above the official rate. This profit, ex- pected to be on the order of 1.5 per cent of GDP, was meant to reduce overall domestic credit creation. At the same time, pub- lic-enterprise prices were adjusted upward to bolster that sec- tor' s operating revenues. A system of price increases was then instituted under which prices charged by different public enter- prises rise each month to catch up with inflationary erosion. 4.2. The government then buttressed the program by securing agreements with industrialists to limit monthly price increases and with labor unions to modeate wage demands. The industrial- ists went along in the hope of diminishing the Peronists chances in the May 1989 presidential elections. The Peronist labor unions cooperated informally to avoid the charge of having irresponsibly blocked a stabilization effort. Deceleration of prices Was expected to generate a favorable Olivera-Tanzi effect on public revenues. A pre-announced four-per-cent monthly in- crease in the anatral price of the dollar was intended to help - 17 - reduce inflationary pressure, but also to encourage imports and so enable the BCRA to sell dollars. The probable real effective revaluation of the austral seemed justified in view of the terms-of-trade gain resulting from the United States drought conditions. All these measures were intended in turn to provide a few weeks of breathing space for deeper reforms, including a revenue-gathering tax reform. 4.3. The authorities limited their publicly-stated ambitions to maintaining single-digit monthly inflation for the foresee- able future. . They felt that deeper success required reforms the Congress would probably not pass in the months leading up to the May 1989 presidential elections. In addition, it was clear that substantial monetary pressures would persist. The devaluation itself would have an inflationary effect. In addition, the SCRA was amortizing outstanding Treasrry bonds. 4 .4. The main reason the Primavera Plan ran into trouble, however, was that the financial markets continued to insist on high interest rates to hold the parallel exchange rate at its targeted value. This interest bill kept the combined Treasury and BCRA deficit high. The promised structural reforms in the non-financial public sector came slowly: indeed, the tax reform approved in December 1988 incorporated so many compromises that -- although it made a number of advances from the standpoint of efficiency -- it failed adequately to increase tax revenues. Meanwhile, precisely because interest rates remained so high, the BCRA's scheme to 'sell foreign exchange at a profit fell short. The economy was sliding into recession, and, moreover, high interest rates available on time deposits effectively dis- couraged would-be importers from using their funds to purchase foreign exchange. 4.5. Through November, December and January the SCM sought to limit liquidity growth and to support the exchange rate by requiring additional inaccessible deposits. Interest rates rose sharply, but the BCRA*s swelling interest bill itself became a significant source of money creation. As the stock of broad money assets swelled, the parallel exchange rate came under pressure. In January the SCM attempted to relieve this pres- sure by .selling dollars. Once the markets realized that the BCM was losing reserves, however,, the stabilization effort became evidently unsustainable. Commercial banks faced heavy ithdrawal demand as depositors, having earned their interest, sought the safety of the .dollar before the inevitable austral .collapse. The uncertainties created by the mid-may presidential . election provided further motivation for deposit withdrawals. 4.6. Early in February 1989 the authorities bowed to the in- . evitable and devalued. Over subsequent weeks, the economy tum- bled into hyperinflation. As the commercial banks' illiquidity became increasingly acute, the BCM had to (1) release bank re- serves, then (ii) release inaccessible deposits, then (iii) pro- vide rediscounts, and finally (iv) permit widespread reserve de- - 18 - ficiencies and even (v) allow informal overdrafts by commercial banks from their BCRA reserve accounts. in April, to permit a larger transfer of resources to the commercial banks, the SCRA made an upward revision in its method of calculating the inter- est it .owed them., credited the addition to the banks" inaccessible-deposit accounts, and then released a larger pro- portion of the deposits. Meanwhile, the authorities revised the exchange rate policy several times, briefly attempting a float- .ing exchange rate before returning to a crawling peg. By May, however, there was full-blown hyperinflation, fuelled by the BCRA's provision of base money to the commercial banks so that withdrawals would not force them t > close. 5. The 3D Plan. 5.1. The new authorities took office on July 8, 1989, five months earlier than constitutionally scheduled because the Al- fonsin Government clearly could not stop the hyperinflation. They immediately implemented their "inmediate-term" stabiliza- tion measures, an overshooting, devaluation followed by a fixed exchange rate and sharp increases in such public-enterprise prices as fuel charges, electricity rates, and telephone rates. By devaluing at once from A$303 to A$655 per dollar, the author- ities hoped to encourage exporters to surrender foreign ex- change, and -- not incidentally -- pay export taxes. (The pur- pose of this devaluation was therefore largely fiscal.) The point was rapidly to reduce the public sector*s overall non-interest deficit. 5.2. Having taken these measures, the Government's short-term strategy was to secure agreements to freeze the econ- opy's key prices, and to begin working on deeper reforms. Mone- tary policy would be run to prevent emergence of a gap between the parallel and official exchange rates. This would set a kind of artificial stability, which would buy time to establish fi- nancial market confidence. The SCM could then obtain domestic financing at relatively low rates to cover its losses until the deeper reforms eliminated the public-sector borrowing require- ment entirely. Later in July, the Government secured the agree- ment of the nation's 400 largest enterprises to hold their output prices, for as long as the exchange rate and the public-sector prices remained fixed. Meanwhile, the authorities submitted three pieces of legislation to the Congress. (i) The Economic Emergency Law. was an al-encompassing,aeasure intended -to give the authorities permission to suspend or cut subsidies and expenditure programs. (ii) .The Public Sector Reform law was intended to provide the legal basis for a vigorous program to reorgani;e, close, and privatise public enterprises - notably the national telecommuications monopoly (INTel), the railways, and the national airline. Finally, (iii) yet -another tax reform was: intended to close important tax loopholes and extend the base of the value-added tax. - 19 - 5. 3. Until the Government achieved a non-interest surplus sufficient to cover Treasury and BCRA interest, however, success depended on monetary policy. The immediate problem was that the hyperinflation bad so reduced money demand that the money cre- ation resulting from the devaluation and the consequent surren- der of export proceeds amounted to proportionally high monetary growth. Some increase in money demand was likely as a result of more favorable inflationary expectations, but it was excessively optimistic to suppose that remonetization could absorb all, or even, a significant part of, the increase. (in any case, when prices are frozen, the measured real money stock masks an un- knowable degree of repressed diminution in money demand.) The new BCM President made it clear that he would not use compul- sory inaccessible deposits, but rather voluntary open-market sales of CEDEPs -- BCRA bills -- to control liquidity growth. 5.4. After attempting in the initial auction to set a monthly rate of return of about 5 per cent, and facing resumed time-deposit withdrawals and parallel exchange market pressure, the BCRA restored the CEDEP rate to about 15 per cent. Although necessary to secure the parallel exchange rate, the authorities were .fully aware that they could not feasibly continue to pay such high rates. The Treasury and the BCRA had emerged from the hyperinflation with about US$6 billion in domestic debt, and since GDP was running at about US$5 to US$6 billion a month, the interest bill would still be around 10 to 15 per cent of GDP, far beyond any conceivable public-sector non-interest surplus. Accordingly, the authorities sought to reduce interest rates by slowly generating market confidence. By publicizing the struc- tiral reforms, the new Government planned, by securing interna- tional support, and by managing monetary policy gradually and delicately, the authorities hoped to reduce domestic interest rates to levels they stood some chance of paying. 5.5. From Mid-July to aid-October, the Government appeared to make steady progress in this effort. The BCRA gradually re- duced monthly interest rates fra 15 to 4 per cent, while in- creasing the stock of CEDEPs outstanding from next to nothing to almost US$3 billion,, with no significant increase in the gap be- tween the parallel and (fixed) official exchange rates. The CEDEPs were purchased almost entirely by private commercial banks. They bought them voluntarily because, as long as the ex- change rate remained fixed, their yields were higher than any conceivable alternative, and they were able to continue funding their positions from private-sector deposits. The encouraging decline in interest rates was helped along when Congress passed the Economic Eiergency Law. in August and the Public Sector Reform Law in September - although compromises on provisions . regarding regional industrial subsidies and other matters caused some concern in the financial press. Successful negotiation of a stand-by IMF program in September further improved market con- fidence. - 20 - 5. 6. In mid-October, however, this progress reversed sud- denly. Pressure emerged in the parallel exchange market when the monthly CEDEP rate fell to about 4.5 per cent. The BCRM quickly raised the CEDEP rate, believing that it had merely en- countered turbulence, but then it had to raise the rate again as anxious depositors sought withdrawals. Commercial bankers found that the' withdrawal pressure were. hardly less dangerous when their deposits were backed by marketable assets rather than in- accessible deposits. If they tried to sell their CEDEPs to re- cover liquidity and pay depositors, they would have generated a sharp decline in the assets' value, hence further pressure on interest rates. 5.7. It is unclear why the situation deteriorated at pre- cisely this moment, nor precisely why it then deteriorated so rapidly. It seems probable, however, that the financial markets fully understood the policy-makers' trap. The BCRA had driven interest rates too low to compete with dollar assets, but with the domestic debt now swollen to US$7 billion, the public sec- tor's debt service would soar uncontrollably if it tried to pay competitive interest rates. As in the Primavera Plan, once the private sector bid interest rates up, the stabilization effort became unsustainable. The authorities could not reverse the drift, particularly when they found they had saturated the market and received few or no bids in the CEDEP auctions. Even the announcement of emergency taxation measures in mid-November -- including heavy taxes on the profits banks had earned in the previous hyperinflation - failed to mitigate the crisis. (The effectiveness of the taxation measures may, in fact, have been offset to the extent people reduced private saving rather than consumption or investment expenditure to pay taxes; diminished saving might have added to the pressure on interest rates, leav- ing the higher tax receipts offset by a larger interest bill.) 5.8. Through the first week of December, both the parallel exchange rate and deposit rates rose sharply. On Sunday, Decem- ber 10 the authorities announced' .a package of corrective policy measures, including devaluation of the official exchange rate from A$655 to A$1,000, and higher export-tax rates. The mea- sures also included a two-year rescheduling of amortization due on most outstanding Treasury bdnd issues. Many observers be- lieve this rescheduling was a serious mistake, because it dam- aged financial-market confidence at a mament when the Government badly needed it. Nevertheless, the authorities hoped that these measures, combined with the emergency taxes announced in Novem- ber, would permit a new equilibrium.. A bank holiday was de- clared for Monday; when the markets reopened the next day, how- ever, they rapidly reestablished the same percentage parallel exchange rate premium, and interest rates came under renewed pressure. Accordingly, by the end of the week the Bunge y Born ministerial team resigned. 5.9. These processes can be traced in the figures provided in Tables 5 and 6. Table 6 shows that the broad liquidity ag- - 21 - gregate, M4, which includes time deposits, grew both in dollar terms and as a proportion of GDP from the third quarter of 1988 into the first quarter of 1989. This was because the real in- terest rates on such time deposits (see Table 1) grew steadily over the period, not only encouraging deposit holding but also increasing the M4 stock through capitalization -- at least until February 1989 when the exchange crisis burst, leading the stock of M4 to decline precipitously into the second quarter of 1989. With the BB Plan the M4 stock resumed its growth, although from a lower level, into the fourth quarter of 1989. Table 6 also shows the sources of monbtary-base growth. Here the significant points to observe include (i) the significant growth resulting from BCRA interest payments throughout 1988 and 1989; (ii) the contribution to monetary contraction of inaccessible deposits in the fourth quarter of 1988 and of primary open-market operations in the third quarter of 1989; and (iii) the expansionary contri- butions of the external sector through the period, notably in the periods inmediately following announcement of the programs. 5.10. Table 7 shows .the main components of the BCRA's losses and borrowing requirements over this period. (The Appendix dis- cusses some relevant technical issues.) The significant points to observe here include (i) the sharp increase in the stock of inaccessible deposits and in the interest paid on them between the third quarter of 1988 and the first quarter of 1989, (ii). the sharp increase in the stock of BCRA bills and in the interest paid on them between the second and fourth quarters of 1989, and (iii) the high nominal interest payments accrued by the BCRA. Broadly speaking, the only categories of interest re- ceipts that the BCRA received in cash were the dollar-linked rediscounts -- mainly export-related credit - and illiquidity rediscounts, which were relatively small in magnitude. 5.11. The BCRA's borrowing requirement was larger than its total losses, and indeed scarcely less than its total interest payments. This point was especially important in the third quarter of 1989, when the nominal interest bill remained sub- stantial even though the BCMA ran a surplus on an accrual basis - a surplus so large, as -a consequence of the lagged indexation of some of its assets,. that the BCMA's accrued asset stock actu- ally came to exceed the liability stock in the third quarter of 1989. In real teras, the BCM's interest payments amounted to their nominal value plus the inflation gain ork the corresponding liabilities. Particularly in the fourth quarter of 1988 and in the fourth quarter of 1989, however, the BCRA's nominal interest payments due substantially exceeded the inflation gain through erosion of liabilities. 6. Common characteristics of the Primavera and U Plas. 6.1. Many observers noted the parallels with the Primavera Plan even at the outset of the BB Plan, and there are evident retrospective parallels. Both failed because they sought to - 22 - stabilize a macroeconomy burdened with excessive, and exces- sively expensive, public-sector debt by temporarily increasing public-sector debt. The architects of both plans were fully aware of the contradiction: they hoped, however, to borrow time at high interest rates, win sufficient confidence to reduce the interest rates, and then carry out structural reforms quickly enough to enable the public sector to pay the interest. 6.2. The orders of magnitude of the aggregates involved made this approach too difficult to execute. With monthly interest rates on the order of 5 per cent, domestic debt totalling US$7 billion, and monthly GDP of US$5 to $6 billion, the monthly interest bill of US$350 million was 6 to 7 per cent of GDP. Even after the massive devaluation and public-sector price in- creases of July, the non-interest public-sector surplus was barely positive at best. It may have reached 2 or 3 per cent of GDP in November. An additional increase of 3 or 4 percentage points in the non-interest surplus, whether it had been obtained through expenditure reductions or tax increases, would have strained the economy severely at a moment when it was emerging from ,the economic disarray induced by the hyperinflation. Even if successfully carried out, such fiscal improvements might have had short-term recessionary consequences so sharp that they might have diminished private saving, which might in turn have increased domestic interest rates. (The peculiar hydraulics of Argentina's macroeconomy made this possible, among other things, because demand for broad money depended 2sitively on interest rates and the public sector*s own interest bill was so large.) 6.3. Interest rate deteraination in the Primavera and BB Plan contexts clearly operated rather differently from usual. In conventional settings, borrowers and lenders presumably come to market with real freservationn . interest rates based on their propensities to borrow and lend, then negotiate nominal reserva- tion interest rates by taking account of exogenous anticipated inflation. In a financial system as open as Argentina's, lend- ers* opportunity cost is govexned by anticipated devaluation and external interest rates. High interest rates induce private lenders to ration, because they realize borrowers are less likely to be able to pay. Where the public sector owes the debt, however, the nature of the market changes. A public sector can promise any nominal interest rate as long as it has unquestioned capacity to inflate. Knowing this, private lenders can demand any interest rate from the public sector they collec- tively wish. The public sector can place debt at any interest rate the market demands, effectively financing the interest by placing additional debt. 6.4. A perverse distress cycle may result, in which the public sector borrows from the private sector; the public sector rolls over this loan and borrows- ore from the .private sector to pay interest to the private'sector; and the public sector bor- rows still more to service still more debt. A regressive trans- fer results, from inflation- and conventional-tax payers to - 23 - lenders. The process breaks down because the capitalization of interest rapidly increases lenders* portfolio holdings of astnadnminated assets: once they decide to balance their portfolios by acquiring more foreign exchange, they bid down the value of the austral and bid up interest rates. At this point, with further devaluation anticipated, increases in the interest rate serve only to increase the anticipated devaluation. In such circumstances, the markets cease to perceive the inflation rate as exogenous in forming expectations. Indeed, the average Argentine financial-market participant, observing higher inter- est rates, is apt to conclude that everyone else's inflation ex- pectations have increased, that hers or his therefore should as well -- and that thiis is appropriate because the authorities are likely to inflate to pay whatever interest rates the markets demand. 6.S. It may help to consider the process from a game-theory perspective. Imagine the macroeconomy as comprising two play- ers: (i) a government and (ii) a financial market. Suppose, first, that the market is fully confident that the government will honor its commitments. Government debt should then carry a relatively low interest rate, incorporating a low risk premium; the government is therefore more likely to meet its obligations. Suppose, on the other hand, that the financial market believes the government is likely to default. Government debt will then carry- a relatively high -interest rate, incorporating a high risk premium; the government is therefore more likely to default. This is why governments in such circumstances concentrate heavily on providing *signals" to financial markets, to indicate that they intend to honor commitments. A government incapable of persuading the markets may prove incapable of avoiding de- fault. The problem Is complicated, moreover, by the reality that the "financial market" comprises many decision makers, each of whom are guided not only by their perception of the govern- ment's intentions, but -also by their perception of other market participants ' perceptions. 'A single market participant might, for example, -be persuaded thit the authorities intend to honor government obligations, but maight also believe that other market participants remain unpersuaded. This person, and others simi- larly placed, might therefore reason that the government is still likely to default. 7. stabilisation easures of December 1989 and January 1990. 7.1. Following the Dnge y Born team's resignation, a new economic team took office, and immediately announced a new set of measures. On December 18 they unified and floated the ex- change rate, announced the virtial elimination of al: price con- trols, and reversed the December 10 export-tax increases. They provided another wage increase to' public-sector workers. The new Central Bank President aiinounced that he would permit no money creation, although the financial press clearly understood that a significant part of the monetary base creation - in par- * 24 - ticular, the interest paid on BCRA obligations -- was automatic. The exchange rate fluctuated between A$1,400 and A$1,900 over the remainder of the week, while real interest rates remained . high, partly on account of seasonal pressures. Prices rose sharply; the price level roughly doubled from November 30 to De- ceaber 31. The new authorities hoped the floating exchange rate would reach equilibrium and that interest rates would then de- cline. Unfortunately, the instability deepened. After Christ- mas, the exchange rate devalued sharply, reaching A$2,,500 on De- cember 28. Interest rates soared.as anxious depositors re- quested withdrawals. Over the last weekend of the year, the floating exchange rate surged suddenly to A$4, 000 and monthly interest rates rose to between 500 and 600 per cent. 7.2. There was widespread speculation that the authorities would dollarize the economy, possibly by devaluing heavily, de- monetizing the austral, and setting a short deadline for conver- sion to dollars. Instead, however, they formulated the "BONEX Plan," the core measure of which was the conversion of the bulk of banks' time deposits into BONEX yielding LIBOR plus a small spread maturing in ten years with two years* grace. The author- ities thereby eliminated virtually the entire flow of BCRA losses through interest due to commercial banks. Since the BONEX traded at a discount, and since the discount was bound to deepen as the outstanding BONEX stock increased, the conversion constituted a substantial unilateral write-down of the public sector*s obligations to the private sector. 7.3. The BONEX conversion had four consequences, two stabi- lizing and two destabilizing. The stabilizing consequences were (i) the direct elimination of the doaestic quasi-fiscal deficit and (ii) the extinction of roughly one half the economy's broad liquidity stock. The destabilizing consequences were (iii) a profound blow to financial. market confidence; and (iv) a sharp reduction in commercial banks' solvency and profitability. Moreover, the measure was taken hastily, and it took two months for the authorities to decide essential details of the conver- sion, which generated additional uncertainty. Over the first two months of the year the exchange rate and the price level re- mained highly volatile, rising at average monthly rates of 70 to 80 per cent. In early March, however, following a month of ex- tremely tight monetary policy and announcement of substantial fiscal measures, the exchange rate settled down and the infla- tion rate moderated somewhat. Over March and April the exchange rate revalued sharply in real effective terms even as the mone- tary authority purchased more than US$1 billion in foreign ex- change. Meanwhile, the Treasury ran a surplus sufficient to purchase about one fifth of this. It indicated that it intended to apply to external debt service under the.teras of the renewal of the IMf - stand-by program negotiated in April. Nevertheless, with money demand still very low and the real economy severely depressed, the economy0s condition .remains highly unstable.. - 25 - 7.4. Argentina's hard-pressed authorities now face a formi- dable task in restructuring their financial system. They set out from a situation in which demand for money, or indeed for any financial. obligation, is sharply reduced. At the same time there are too many undercapitalized private commercial banks without effective supervision. Somehow, a way must be found to increase the willingness to .hold Argentine money and private assets, and then to restore the commercial banks* credit and payments functions -- all while restoring macroeconomic stabil- ity in a context of unsteady public finances and unsettled ex- ternal debt. 8. Concluding observations. 8.1. When a government slips into a debt-distress trap, the available policy options are disagreeable -- essentially, to continue the cycle until they can somehow get together a suffi- cient non-interest surplus to stop it, or in some sense to de- fault on commitments. As long as the cycle continues, pol- icy-makers lose their capacity to carry on normal policy activ- ity. On the other hand, it they default, given the government's role as the economy*s principal debtor, they set a troubling precedent and make future public-sector borrowing difficult, if not impossible. 8.2. If the incoming authorities had taken sufficiently drastic emergency measures in July 1989, they might have secured a rion-interest public-sector surplus sufficient to pay the com- bined Treasury and Central Bank interest bill, thus achieving overall public-sector balance. - They might therefore have avoided further debt accumulation. They clearly attempted as much as they believed politically and economically feasible. It cannot be assumed A griori, however, that rapid achievement of an operating surplus so large was in fact within the authori- ties* reach. Over any time period, there is assuredly some max- inum feasible improvement in any public sector's non-interest surplus, particularly for an economy setting out from depressed circumstances. It is therefore at least gossible that the BB Plan strategy was simply impossible -- i.e., that the inter- est bill exceeded the public sector*s capacity to pay it. It is at least possible that Argentina's public sector was, in this sense, truly bankrupt. 8.3. One of the striking lessons from Argentina's experience is that contractionary monetary policy can be peculiarly coun- terproductive in an economy where structural problems persist, where money pays interest (and where money demand therefore de- pends on interest rates and on expectations), and where the public sector itself has a high interest bill. Over the latter half of the 1980s, Argentine pollcy-makers applied vigorous contractionary monetary policy, forcing interest rates to levels no private sector activity could pay in any context, let alone in the depressed conditions are the Argentine economy. one con- * 26 - sequence was that the formal banking system gradually stopped lending to the private sector. The scarcity of working-capital credit -- let alone of investment credit for new enterprises - has become a standing obstacle to economic progress. A second consequence Vas that the public sector itself became heavily in- debted, arguably to the point of distress. 8.4. There is a widespread view that governments, unlike private entities, cannot go bankrupt. It appears to have been Argentina's unhappy lot to have provided a counter-example to this view. The January 1990 BONEX conversion amounted to the Argentine Government's declaration of bankruptcy. If indeed the nation had become bankrupt -- and there is, at a minimum, a strong case that it had -- then the authorities would appear to have had little choice. Having taken this difficult and regret- table step, however, there are grounds for hope that this funda- mentally resource-rich nation can return to economic progress, if it can maintain fiscal austerity, if it can gradually restore private-sector confidence and credit flows, and if it can work out a realistic arrangement with its overseas creditors. ARGENTINA: INFLATION RATE w|C 1987 - APR 1990 210- ' 200- 190 - 180- tro 160- 150- 140- 0 130- 120 - 110- 100- 90 go - .70- 60 - 40 E 30- 20 - 10 - 50, Da-87 Jun-88 Dee-88 Jun-89 Deo-89 0 fCotblned'* Index ARGENTINA: FREE BANK INTEREST RATES: mIC 1987 - PmB 1990 .210 200 190 180- 170 160 150- 140 130- AM0 - 110 - 90 70 60 '50 40 30 20 to- 10 Doo-87 Jun-88 Deo-88 Jun-89 Dee-89 1 Dank flability rate + Iltilon rate ARGENTINA: MONETARY AGGREGATES J~N 14ffi - m1 1990 17- 16 15 13 6- 10 9 8I X4~ I 7 o 6 5 e . 4. a Jan-88 Jul-88 Jan-89 Jul -89 Jan-90 IZZ M1 M4 lees ML ARGENTINA: REAL EFFECTIVE EXCHANGE RATE az0 (Combined consumer, wholesale priceg) 300- 260- 240- 140- 220 - 160 140 100 80- Jan-87 Jul-87 Jan-eg Jul-88 Jan-89 Jul-89 Jan-90 Jan 87 - Apr 90 0 Official + Parallel ARGENTINA: SAVING AND INVESTMENT, 1970-1989 lao- 25 20 t0- - *10 • 1970 1973 1976 1979 1982 1985 1988 etetet+ Dome~tt oaving 10 Import% exports ARGENTINA:PUBLIC SECTOR DOMESTIC DEBT 啊叩1955一力唱C 19e9 祖； 觔睡。一肥細。一舫·觔。一群刀匕。一冊刃匕。一日9 __q)N間目闈煙勵A鸞鸞日居pA訕U勘亂屆叩BANG蔥RA彎蔥 匕乙】）憑細dB&kl以NN此細nd hea•”叮 Об-.рд-00 TAaLS 1Ак 11ПбВfГ11Ч1г рМрц, eV�ROBCOl10ltIC 1fOКATO:R. 19�-10а0. Опга0 !М 10в0 1os1 10вп lоеа ivaв lовв lvев lоаа 2ове 1vв9 AMB1N� OR 1ФЧlК7 IlаКАТ1а11ЧТЕ9г /1 Nq1Np1в �гiСр '/.1f O.qt 1а.01 19.8R 11.1t 6.1я 6.lа 11.1t 88.6t �qiwae•t �вlsвв . �.В� 1п.8S 1в.бп 16.Ot 18.вf 8.Оя 0.04 14.Ot 30.6t А119Иа! ОР 10П1�Т рЕМ. ]NfBtl�! 1U1?О1к /4 а. ce....et•г lааг е1м �мвди /а -1.та -1.и -4.4tiг -п.1х -1.пх -1.ая -о.о: а,► m.иsst•г-ыnгг гем. �а s.8s -о.вх о.1я -о.1я п.1s а� tв0а�м оо.в•Ча» /а п.еs о.ак o.ss п.1s o.es 1.ав •o.1s 110l�tNtц АОО<�1{flбе /4 �►м ам ы то) а.ви s.1s е.ав tьnо..емгг wда{r (гп) erow.e,�,.,mb р� пе.а па.па пд.оs 18.аа tп.0а 12.4s 17.?а 1в.аs 1s.яs yp�n у� ар ���д� о.в s.s в.о в.в �а Ыгttмг .еваrмl м1w• sa:wгe.e •е иго. ы амг.. в.«. ,овем• v*•гг•г мем) l�elqql Тг•вlYqr 0.7 0.7 0.� 0.7 а�{►{р 0.4 0.4 О.в 0.8 реггр� р„г� 0.8 0.8 0.4 О.в �г � а.о е.в 4.8 4.7 1 � �г N Медмf fRf гNв (ialО. i•ег.) /1 10.ai 1lР.8М 48У.А! 818.4ц в2f.0а 7аt.1я 86.7а 127.7а 80в.ОS 47J8.2t � hrotltl ввеА.-гоN ог•вiw /в О.па па.аf 40.7t ав.?f 8а.0а f7.8t 11.вЕ ?8.1t ?в.0Е 20.ES Rи1 •Nвeeioo вьд+. г•Ч (•�t1e1•1) 46.0 а1.4 00.Н аб.п ta.a а6.Е 01.1 100.0 06.6 1п8.? 1 9A41tlN[Е ER ОцrlQlflS � Ыг11м)г /1 ew.«иs•екмо wrotw -4.е -а.7 -п.в п.4 -п.в -1.о л.о -4.п -к.в -1.в иегемыг...е.ыо.wвlw •1.s о.7 п.а в.7 4.о 4.о пл 1.о ss в.в 1lмdqn�iвe q1егМ 9.0 4.1 Т.б !.! 6.1 6.4 в.4 в.4 4.f О.в iйвs<,м1iи lаоогlв (t11� -0.4 -8.4 -4.0 -4.1 d.L •8.6 -4.4 -а.4 -�.9 -в.п 8вг+lsа N/ако •а.4 •а.1 -а.п -а.п -в.а -а.б -а.8 -б.п -а.е -б.в РI.иМ 4м.-идь. гоигw• •eesk а.1 8.8 п.Е 1.п 1.п 8.8 п.7 1.6 8.в 1.8 0�6в ООСЕ6IК Nt00t1C4s fд 1RY бtftp► 1а4.8 18t.4 вВ.0 70.0 7а.4 вд.0 70.1 а0.1 01.f ое 196а ргiи► t108a • 10р) /1 ia.0 07.б 00.п 102.1 104.0 i00.0 iбS.a 107.а i01.? teиsswв•lwerв...) 1.ss к.вs 1.м a.os п.вt -в.вв в.ая 1.v: -а.1п т► аыгsеог tв.гвw►ае. мггии) а.ав Oo.as 1ва.оа a40.1s as•.аа ваа.4а 7в.оа 1?т.пя т.ех /!к 2е0вгмtiаиl Nsмesryr 1rмf. /Rs 1NтtA W►e Тв1lр (1гsгИ !ы�). /at iseмsle Ргевг•гь ы ОАв Аврме{м Ооvв►моnе. /4t Qe•aewis Cawtиion tet Wiп Awrtи. fai СоМ►o1IN ввМ• �гов ам 10ап • QCD 1Уа7. /а: A�bt4s ftпмдого. г�а� s а� � п�� 3�»=z� ��: з 3�� �д ��'�� � Уп� z �$�� �: � � � �� s _п_- '�i - � �_� ��п�� g �ё � п.. ��- � � -� i� i �- �п� _д а4 � . �'. �_ 3 � Ч � � О �� '� � � �� � � '���+$� �� � � а О п � - w �� . д а� � �_ ��� - � � � п_ �� � _�� �'� g�_!� $8 g�; � � �� �, ру. S 3 а ёг,� �.. v . S 3 � � � �� - � � � � � п � ��- +� +g �� � $ $ $. � ..� � в � �g• �$ $S'л � � »' �� • ��� �° S ;��� � � � � � S�_ ��� � #�з � : i8 ' ��:' ' � • � �� _ �sг �� s . а� � g� ��� ��., � � �i � w лi вв в р iв��гЯ� рУ00в�вq ппвlfY FЯ i� - р� ЯМаг1lдхпi�д ië1R � i�в q e • q � 1.�рΡ�pN$ ОвiдОп0� Уp�р�рΡ V V• гf���в Фi►�a� 5лгыrрпiв�уΡ�Qь0 OаEi � У iqRMY ОООЧ Ч V М М п i� Мп АМ �Оьв RA7fw�i�R�вf N� ьв ��tр� ApCвfSO Rвгп iг 1�i 1�R �$�$ ����ы'�уМММ 1�М � « bp,�рΡ'д•�� lлo..�SjSi oлaesл,"'ыа o►Gp дуΡ��$ ёаp&влWвsёлwуΡо!�плаSл g � i W О �ОМ! а0 О М 1А ii �О i�ONp �ЯЯц�R RАвГR Я1� ai Sы�'аb� wŭ.•O',!Т$ oapaS'sа«апeп wрΡ�w.~". е.л'••'алтлwрΡоьq8p1p3аL8 � й�М PV ьв01 •�10 �М� �М�в�0�! �R V Ур �МдЯц�RММ й� А t V � • ! • ~ ~ � � У / М У `� • va�.� иwилbRе eaai3w.•av w�88 s..aavoa....o$ � ri`si�sR �''vo�avм °��йуг�веj°��' угооо п$ii�E�°�ëO1йS�iEii•й Q й i�f �lьiьi� j�"о'�wav �'iйа`iййEi�i� °�isaw ���+�$i�ER�FiiiiR � •• � EiiйЪi �'va�wo йR.°iZi°�.4а°�� $°g° .R,°:йй�йа'4'.��пG � й Ф.. пЯ М� д V V О�Р ��� i!N ��А 1RФ Ф$ Я�р Мй �М71�М М i! � F � - Ч9Z � ов-лм-во тАеt.е ise A�пna►f аимтааv и�еаимк� пакглтоак. �oвs.�oes. ва�ао м fоев i ioeo ii iosa цi iоы iv zoe6 � foss и �ов3 ккт �oes кv �oss к :oss кк �oss кsг �vвв пг A1�BtA0B OF 110MBtY 11/'lATtON 1т11в6к �/1 . - Мюlеwlв иiеи , fe.ss i�.er рп.вt ts.ox пв.ss ss.вs а.ж s.as a.is <..: �.м s os tа+sмг иfАи iв.пв iв.вв ю.rn u.TS n.л ав.оs , о.вs о.вк о.п з.и 7.ss .,в At�ASB OR lФ1i1t1s,Y 11в/16 Z11T@Ч�Т (цTpie /в м �е�..еи� ии� et« +аиtи /б -е.оя -в.ss -s.ss -i.вк •б.м •в.аs о.а: о.вs -о.� -�.as -� os -о.эs es ееоввг�еЫ-►s�г lевм• /б -6.if -а.4в -в.Ов о.lц -s.q . -l.sR i.oa �.st �. п -o.2s -i.вs �.� аи �ви.г�..аониtм• /а -а.аа� о.тв -п.т в.ов -o.st sr.os в.оа п.вs �.м .o.ss д os z.os етаtтАяr �аоатилаs: � /в ' �ягг awa ег аоу w►га. Аоиr «ав�r с�п� в.кв а.vп а.юс аsв а.о: а.юs в.о�г в.п a.ss б.вн 6.as s.a егОвб веМ�r wP►ff Мвi 1в.вц 14.of tп.в! i1.тt 1LOY . 10sS 12.9t 11.Ot 1в.5х 17.вlt lт.ilt 17.оlt Овl�TiC О�Г р' 11Е Я16L2С sBTatt . 6.6 7.0 в.1 8.8 (tlis 61t1/ew� woMh-мi гоlwв tanwвЧS N в�о. ы м.... ..... .�ип,• ....�i.� ..и.� ибiм.г т.аw� a.z а.в а.в а.� sвsu.teiи i.7 к.Т �.: i.a � ��_��' ��мв о.в о.в о.т о.7 • Смм°1 вм1' в.а в.7 6.7 б.Т N О� Е7УО1а10Е tЫTEat � Е'' ие.�мi ивв г.и b.ea. tм..) /i sе.ю� п.оs es.вs оо.яс es.тs ias.�s вs.вх о.о: о.о: в.is iв.м �в.ы t Овsв1Ы екА.•таlв вгмtw /а ввsе вs.es лв.аэс �e.at ав..в вв.�s кs.вв ц.в: и.юс в.ж о.� �s.es кее►.г.о.емго.ил. гsM�et41a1s1) e:.i та.а 7о.в ».т so.a ва.в оz.в оа.в оп.о оо.о oo.i оо.Т е�цааее а► мnегВ �ust ицtа�)� р � а*►мв-вееамs wna. • -o.i •os -о.в -�.л -os о.п o.i -о.а -о.е -os -о.а -i.к мог<ммn..-a.t. arotw , iл i.в о.о о.в о.о t.�т i.в о.о о.в о.о о.в о.а и..�л�«..оегев в.п п.в п.� i.a �.в п.в п.а s.T s.в s.o к.о i.a и..лмпн« t.►ега ceкsf -o.s •s.e .i.п •i.п .о.о -o.v -о.о -о.о -о.о -s.� -в.а -i.: s.w�и. и�ми .i.s •к.е -к.а -i.e -is -is -is -�.а -к.а .к.а _�.s _�.. ►а•-мб б...-им. .'.•и.�,...еееб s.s �s is в.п о.в i.т s.o а.а ss в.а а.в а.Т ааоаа вaмesttc гмоi►ог: . � tn 1RS бtliм �� вг�� h� ��) /t :oi.s sas коа.о iю.е iод.� оо.а оа.е :ов.е ioo.s юаs ias цi.о tвжмаМ ��агвав� • -s.oi as6 -i.� 7.8к -o.1t -oss •б.б.з la.es •а.п в.ев -о.тs в.м aD' Чfts� (еаsм�ув 1ns.вess) . ...�........ . .. /iг tмвлиsiмоl �а.вь.гу �ме. /�г Уовl� Ов10 Твl1и Q1Аег1А 8адг). /8r Sгasestt lrевгвв е! рв Агвмihа оо.егмомt. /И teenмte tавlрlт 1ев Ьв0iа Аж1ев. · 26) 눌 ---.;[,!i,.&..:.:.?&.,.．·（ · ов-,м-оо tАв� 1ег иговп:fв►г цшягеа.r �ис 1юlитонs. 1ов�-1оео. впгsо ои 1оеУ к:ом. t1 19ет 11t lоет 1v lоев к 1�е n 1� 11к 1о� 1v lово 1 lоао n lоао 1n lово 1v . . AY9tA0! ОР 1011КУ 116qAT1p1 RАТв3г /1 У1ю1ив1в Огiср Т.в! в.48 1t.ot 10.oi 11.4в 1Т.Ох 41.вf 7.2i 11.ef 7St.2R вв.вs 1б.в= Оеавwов О►iеов e.71t ♦.р 1s.ee 11.7в l8.У1[ 41.8а IIO.e* в.7в 11.3f 06.1f b0.9f 1'.вв в� OF lAN11tT 1t�ц. DrfвЛЕвТ R1lT!'!г /t ем оееов►s{в1 1►и4 еiвв Aqos1M Jb •1.lд -1.l1t -в.р 0.4f О.оlt •1.вв -1.ef 2.7f -O.St -1B.os -16.es еА �е....оаl.иам 1ем. /а -o.rn -о.1в •п.п8 п.�в а.аг 1.п -1.ь: а.вs э.п •4.1к -o.es м trи.rn.. еигеоlаа /a o.es 1.ot о.» п.1в 1.rn -о.в: •a.as sss 1.n в.зs -1..ss Момtглаr яоаееоiiп�г /л (1е► аме е! ФРj па+гое.еыq. «а►Н fiп) ' 8.тs в.ви o.as в.ае< в.ов ass s.n а.а а.тя х.а 1.п о.ма .м.r'•�roo1r Р� • 1о.вs 1о.в1г 1s.a1t 1в.>1с 1в.а 1е.ы1 1а.вs 1в.вк 1о.в: 1г.в: a.es oa�etle огагг ем пе welac веегоsг о.а � а.в ю.в о.У т.• в.о в.а а.1 4.п в.z в.в У.п Qei! 6i11tм1 веnМиМ аlиев омwгЧА N овв. ев swr.. ►го.. вагМо' ssro11o1 nово) и.ыемl тга«мr а.а а.п а.а а.п а.о п.о 1.о 1.1 1.1 1.в 1.в 1.в 8owt1NN 1.в 1.Е 1.в 1.t 1.t 1.5 1.в О.о 1.р 1.8 1.4 1.5 0►tle.osгf ssвtap � О.� 0.1 о.0 О.в О.в О.в о.в 0.4 O.t O.f 0.1 О.! tмомl в�мь s.1 s.a в.е s.s в.4 ..о в.в а.о а.о s.s в.о в.в . 1 � �г и aмotasl t�t rои beeo. 1вег.) ./1 a1.os 1б.lt вя.7s во.оs 2r.Yr вв.мс ыs.as 1а.п 1в.вs вv.cs аво.еs ве.ов � м.вии .г�.и.и о•.вt.+ .' /е пв.вв пв.пв sa.a гтsв ва.в+1 пл.1к ю.вв 21.ат п.оs тв.ая s.� Yo.os пвsl .tв.estw маь. г.и (en1o1s1) ов.� • ов.2 ов.е 1oeJ ив.в ioвs a1.i eas sa.o lввs 1ао.о 1z..� 1 6Nt/0kt ОР 1ArlВ11! р1Вв Ч 11/еа) г /i Coгtмt-fCtetlвO вw's1q • •1.0 •0.8 ' -1.1 •1.8 •о.о • -о.1 О.О .o.S -о.8 -0.1 0.4 10ггвМйлiвиовеtо вwptq О.в о.в 0.1 0.1 О.в 1.0 1.8 1.Z 1.0 Z.6 1.о 1Мгдiм/1р оs�ег�о 1.� 1s 1.е 1.в t.т . s.a п.� s.4 п.1 п.в п.о 1ьг�1и t.оем. � ; •1.1 -1.в •1.в -1.в -1.1 •1.в •1.в •1,п -1.1 -1.о -1.о so..1.0. иtо«о. -1.4 -1.в -1.а -1.л -1.а -1.4 -1.в -1.в •1.о -1.� •1s w.-ме вм.-им: 'гаегор алов4 в.е 1.а t.o 1.в 1.в ' 1.о в.о в.в 1.в о.о п.а 1.в аroьа ва8lвтк оаоииег� iв t!!в бiltiм • as lовб eгisq р9lв � 1о� ./1 1ав.в ков.в 1�.в 11п.У юв.в ют.а ов.У 1а.в tи•вм+.ео 1м+.м.) .в.вя в.т -sss e.os -в.�s 1.oz -s.мг s.os aDF 1М1еwг (�вг�рвоqв lаsгров) . ®®� /1к IдМгмЧеивt 10еММр• l1мА. /?е Уог1А Ов1i Теltи @ba�1A bм4). /ss Ееааеаlа Рвевгвв вt tla Аrвмtlм оswмьмt. ре B�eneвia Севвlрiаn tег to�ia Ми1ев. - � . ・二’-1&.---r.&&&&--(!--------1--------&---!t- -TgZ柳 аа•.wмоо *м�к 1е� ивов4гтки: моиги.r кr�акпккlс 1гоlитопа. аеев. о�:ао ги .aw lаае ►eb 3oes ws 1о8в дог lоев wsr 1о� ,ми iом .лгк lоав див ioee вео lоsв осе lовв н,.. lоев г,.е 1ок� маиnа.У 1г�lдпаи ап8е: п Миlавlо fгlеа 1п.1s 1s.ser 1e.su 17.� па.аs sas: пв.аа s1.or в.s 4.ва a.or а.7а Сеиоwов вгisоо о.18 10.4а 14.7М 1f.1f 1в.7а 18.ta Ув.бf 2Т.ва 11.7f о.о! 6.7а в.0т кокгпаr канаа� Уме�т wпедк �4 а► ce:oв:�is1 lonk t1ь► Овреliо /t i1.ri i4.4s 14.ia 1a.is sв.1s 1s.тв ю.7а io.sf: в.1ц е.ьs v.a8 u.va an seв:lesslsl-6в:�1г lевм /б 1а.ац 1б.18 17.аа 17.вf 1о.вf t1.oa 2b.1Я 18.в1t 1Q.11 1;t.p�a 18.71 14.Уа en 1sЧвliгв евпN�аи :/а 1а.8а 11.Ot 1б.1R 11.1а 1а.оа 18.8i 28.Оа 14.4Е 7.tt 8.8х 9.8Х I1.1f lФ�1`АП1Г 1i0oл60+tTl6е /1 lо.оьЧ� аиг в.о.lеw вмlер . ио.гев аему :::Mny рр) -а.бЯ i.oa 7.as 8.7а 1.к: ss.oв a.8s n.as n.1s 1o.:s о.>х п.8s �м0 вмквD ooe►�f lMh B.aS 9.Ot 18.7! 11.аЕ 14.М 40.оа 21.вЕ 21.1а 16.8t О.а! 18.?i х.ж 11евоМt8 MN Obi 8.аУ •8.af 10.8Е IO.BS 18.7П 16.оli 8A.4R i8.61: 72.25 а.ЕЕ 17.bf 48.ta ооига►tе о®r он tиt wвttc аасеаек в.е в.о е.1 в.о sл в.о а.о е.п a.s в.т а.о 7.1 (�а0 Ф111iакк к�еnУим4 гоlrи «:wslоl sв ora. е1 еии.. fвw. rмЧи� ysailol м�а) ��1 т�ог1 S.a tt.8 а.1 Y.II II.1 2.0 ?.0 1.8 1.а 1.в 1.а 1.в �101p � S.1 4.! i.8 1.7 1.Е 1.Е 3.а 1.8 2.! 1.1 1.о 1.о о1��амгr вwtмв • 0.4 0.4 0.8 О.а О.8 о.а о.а о.в о.а о.а о.в о.е �м�� �ь �' а.а а.4 а.о а.е s.o а.о а.о 4.в в.о а.1 в.8 а.а . � О� RдтЕак ro 14iым1 1bt е�вlв fвefq. taer.� /1 11.'2а 10.вt 18.ва 16.а1 ia.oli Q1.4a 10.6ft II1.86 0.0lt 1.oi 8.ef е.вв � wт.�и1 ..а�._ии в.м�w �в sв.в8 ю.аа ю.1s ю.7а я4.вs гг.и ав.ак к4.ак 1v.тs n.� п1.п ю.1ti � а�.е� игsеlw..ви. гNS tе101дв1) 11а.п ю7.о 1о7.1 m7s 1ае.о 1а.о ов.7 о1.8 аа.а sa.o ы.? е?.а еалгке ог вxr�ra пиа е�гi�е~эк n к..емк�еt«-мвЧ «мв�w о.1 os о.п о,п о.а о.4 о.4 о.в о.п о.я о.4 о.в Мидt.41Ч Чвегбв 0.6 О.а О.в О.б 0.8 О.о 0.0 i.0 0.7 О.а о.8 1.0 MO/fUM4�p �овОвlо (С1Р� •8.4 -0.8 -о.4 -0.4 •о.а •О.а •О.Е -О.а -о.в -о.4 -о.а -о.4 Р4•'+�0 Ом•-oвsfi• �iввгws a0eek 1.4 1.8 1.а 1.8 1.6 1.0 1.о 4.7 8.о 4.0 8.0 е.в � /1к fлlвr:a►iaat а�ккквМq l6М. , �_�•~�_� /ак 11xtt Овl9 Твltва (O�ettO 8м4). , /3к ts+eвeвis Ок�а0гокl М 01и Агамtiм Омиммl. /Ок Ееакавiе Саокliввlеn 4ег 4t1n A:�niq.� ое.лм-Оо tAel.s кг ма9+гlw►: жнt�б.r wиап0ЕСаиоикс 1юlсатопs. 108+. атга0 а+ • Jen 10l0 Fo► 10и Мрг 108V Аог 1080 Му l080 .hм 10В0 Jo1 1080 Аяо lО8У 8вр 1080 Oet 1060 !br 1оВо Dве lо80 • 1ФИ11L1г 113iЛT101i RAt63s /1 • Мюlеоеlв в►iеи в.9f l.вf 16.41i . 66.08 1М.бЕ 188.и 100.!! б.бх П.St l.E1t 1.81i 49.68 Се�wвев егiиs б.0У о.бlt ».oi 86.4R а8.ая ки.а: lvа.бА ат.Оt o.4s ь.вs в.sя 4t+дя MONПS.r tWMLNЧ. ЯП91Bt RAtE6s /4 w еевереiе! Оыi О�вв Ирееiр /б l1.b4 11.0! l6.7R 6i.11t 70.М о4.0У 1l.4х 14.Oi 6.Ot б.st 0 п м иевепlоl•►aR 1им ro ц.б► 4o.8t 44.Mt ао.ОС 106.бя • 1ю.оs 48.8х !е.lt 11.2s о.lt 1?.6t1 os tew.гl.. e....вten. /6 1о.ео loss пl.вя ю.� 1ао.ов laa.бs s8.81t 1p.aR в.а1с a.aR 1101�ГNdr AO�ltfdг /4 (0►евв1� евЧ огегlеw вегlе� lьгro..емr ирb ор) •i.ap i.бs iе.б# 1в.вв оs.бц aoss 1оа.� бт.оес es.aY lО.вп п4.,к 8�етЕ веiаа ирls (ОМ) 11.ю ю.1к 16.8L 8а.т as.4f 110.8f Оl.оЕ 4Т.9lt 81.б: 1б.0Е 1в.оЕ Menelевlr bee (1� �1.48 8.46 71.б1t g►.II8 1o9.4f 64.У'! lЧ.бs 40.Ыt 48.4f 24.6f 2s.os ео�вsТке Оsвг ер пе wльlе еееtоп� s.a е.о sл s.4 т.! е.! б.т в.о е.л в.в а.в в.4 (U66 O/tltм! вм�-!wa valwe имrе►МЕ ев в.о. е1 оwг.. в►вв. ваг.Ма� еегеltеl гоМе) 14Иаа1 Тгеiирr 1.а 4.� R.e �.7 1.0 1.б 1.6 !.1 !.4 !.б 1.6 !.3 8еwвlЧй 1.1 1.а п.о �.п !.б !л !.а 1.? !.а !.а 1.4 !л ! оеl�о.ео.у .е.lеоо as о.в о.в о.в о.а о.п о.1 о.1 о.1 о.! о.1 о.1 сМ1го� 8вп! в.о 6.0 в.8 1s ss 4.а а.а 4.6 4.о 6.2 4.в п.о � zy eкarrio6 �► ! lbвtи� tp8 ай (,е00. 1еег.I /! 4.!! e.61t б.бц 8Q0.8! 08.8lt Ce.Bt l10.4R l6.4fc O.of O.O1t 0.08 aA.Bf о...iи! .,.в.-..е. о+иlгw /в se.o8 so.ls l�.1s -8.86 s.es w.a! l4sя п.ох -оsк a.6s 86.4s 1e.es Rи1 eNиtlw иоА. гоЧ (ot11o1e1) а0.0 79.0 70.0 208.0 406.6 366.! l86.8 teo.б l28.а 110.0 116.Е l40.0 811tJ1tlR 08 1AtA�118 (1bt Ч t(1еи) е /! ииеhааitее.ивtе ewr�w о.8 os о.8 о.4 о.в о.в os о.в о.в о.в w.sмetlи ивога о.а о.а о.а о.т о.о o.v !.о о.о o.s о.е 11вгаАМ1lи tqoгM (С11'j �0.4 �0.8 •о.8 �0.1 -0.8 �0.4 �0.8 -0.8 -0.4 -0.4 РО.-и! tи,�виА, Nвегвее в0ееlг !М 1N 1.0 !.8 1.6 0.0 t.i !.0 4.8 4.8 1.0 1.6 /1г Iw0егnоЧвМ1 11емМвr �n0. /IIr Mer1d ОАв 1еЪ1и (Уог1� Qndr). ps 8eeeeeie rге9гав ai Уи Авембlае Орwмовnв. р! Ее0гаевiе GОвtи{и tОг Ii,I11 АиГIи. тшие гк ааУтпlеа: влrlис ас wпм. Роагn►тта+. lvao-1oee. 1т•Jo1•УО .." • 1о9о 1061 lобп 19еп lоее loas lоав lоат lоьв .....�.. t/oмRtпяnelsl ►еllts aeslег: •»..»»..« ..... 1bn-fsMмe► 6win0 • a.af -О.ах -а.Ох •8.2t -4.а� 8.8t 8.5� 2.7Х Овпем) Оs.вгав�Л0 а.оа •У.О'Х •a.Z>< -a.1f -2.OR У.аа 1.ац О.М� 0.8i Геlat Сwмяо lbeetPM 2т.б1t гб.8i 73.OR 2a.ai 22.7f ?7.7t гв.аf г8.�t ге.гlt Tst Авев)о!в �.аа >10.лК fB.TR ts.� l8.2t 22.tt 2?.8t Z1.sa ?О.Ов 1bn.?и Rвиbb в.аs в.оs е.пк в.м1 в.аs а.яс а.гs s.o: г.а: w�-lееиие w..ме tгомеtам. •гв.аs •тт.7в ла.олr -�.а: -па.вs -4а.е: -га.оr -га.а •r1.оя а..1�о et мf11s ывогоии. , о.ав к.в: o.aR •o.1s о.вв 1.os 1.os п.гк rегеtи амиие г.�г..м. е/ -о.еп •я.пп -в.» •а.ох -в.гк -е.тs -а.вs -а.оц �.г.! ор...�wнм •о.ах -1.1я -п.ая •а.с1с .я.ыs -У.тк .г.1и -1.п -1.тs Puбlie 8аtвво.tаво ' -о.бб -1.16 -в.г! -п.ох -f.1ft -г.о7t •1.е![ -1.СА OewN�1e Iаlемвf РвувмМ 6/ -�.6t -6.4К •6.71 -о.оХ -O.af •o.OS -О.е>< -o.L1t Ооаевв) Qевогмм0 •1.8f -4.вц -a.OR -о.lGx -О.ая -о.гR -О.аа -0.бх -O.as РмЪ11е �lвгогlрв -1.1lt -4.вц •1.о!{ -о.а$ •o.aR -0.8t -O.1f -о.1я Tofвt аввiд� •о.311 •Т.7ц -1a.8f •0.4f -т.8� -4.1t •о.вt -1.л>< � кеввslвиN е1 -o.4t •О.в>< •8.1s -У.ац -l.бf -6.6t -7.et -7,as -А.бв омемt оs.егмме •6.8t -в.48 -в.е8 •б.as •е.2в -е.оя -1.ts -в.ак -а.б+� PuD1)в 6►и.Ри.» •б.е% -а.пб •а.а: -в.вб -а.аlс л.еt •г.бl1 -а.4: -4.т+е иоs е.tеме .о.лц -tт.111 -n.п -яs.Т+с -1e.os -s.o>< -г.тх .в.ов 1 Pгtwb aeslе► � •-•-••" о 8sв)ав osalodip Улiеиав. 2еvевМмf fnceвe 11t.0!< 1Т.т!< ft8.af ао.4! 4a.a1t ZB.�C га.аа Ja.7R ) иее Регеlее tеи.иt еев tр...мме t�ком ess а.ыс -s.1s -asx -г.ья -л.гs -в.� -в.сz Ри>+11а 8к�ев lдtвгвИ (eas1, аGЧ) 6J p.8t 6.4� b.tf 0.9t 0.6Х О.бх 0.4У o.бt Тоав1 авvl�р 19.4lТ 4a.qt а2.1* 46.ОХ 4б.8! 2а.0li 10.о![ 1a.7R кМ.о.аме е/ .1s.ов -1п.еs -1л.1s .1п.вs -1л.вs -1а.ек -1в.ов -1л.� -1a.ru Nвб аfми s.4t /a.йR lf.Q!< ia.S!!t 1г.2!< Т.ва л.0li л.0l1t Еsbмвl sвe0sr •......•...«» • 8srlns (RввsУ►ев Веlsдее) a.o1i 4.611 -a.as •е.вэс -e.oi ..l.вх • -е.ах -8.7t -6.ot No0 Рвеfет авwiаи е/ -8.8! -1.at б.бб 6.4!< в.тli B.Ot 8.;)! 8.6R В. � и.а а�.1.ае. �аг.«.е яeeev,�s агiеlе> s.1s а.а: a.az а.е: г.ж 1.ю) а.ж е.ок 1.es мие Iw..e " оо...в1s s..reolаог 1а.ак 1r.a па.ва гт.аа< па.т: гч.вs пв.»с гв.а: го.1я и.ымеl aw►f�/аоР 1o.1s 1в.ж :a.n и.еа 1v.as ао.оп ке.ая u.as го.л4 Огер Imnal1м01ФР s! Wг.ме Р.1еее 42.'1t г?.47t 42.2а Уц.40 4У.2lt 72.2i 42.гR г2.г1t гг.гt аа.. tе.ваии�►е/mv и>ато w1и. га.п 1о.вц sв.п 1в.пt 1п.ат io.mc и.в>< 1а.ха 1x.as аоР tUat лlгliм) t/ 1a1.s 1ае.а ss.o То.о vг.7 тS.а ».п м.т ю.1 ввввiввввввеввввввв�в�вввв�вввв ■ваь�ввв s/ Inlвгвt� вдА м0 Аiгее! iмrqlьм� inca�в.k 1nReroвt иrемм�в an ввбвнаl Авlt. :/ HвetvQи tiи es-pыr fsfls0iew saoer,мt е а/ llof ов eaPipt вевомiе• вwгвбln0 0.6 Рев � д/ 2ettuAlвs е1�вА0и fw lwvмоегlи. fnиNиo 1е !Ав BGRA мЧом1 estaмb. вAief. wв о feliвteteд fАвб А1�tЧв lrеь tAвt е1е веео�пlв. в/ Uкtеевв мв vпгвprlssd ммвtвм. •/ Са�ьегtа'вА Чв оеевогпlвtе (efO1e1a1) • 5еипиt tмtaat 6оМ ot Чи АгумОiм Rwubti 0�-�ег-00 1ABt! II. ApOB7t1Ml: tцTIaN11L :REASIAlY MD CENIRnL В><►ic OfBT, ru�-nEC iose. 00к49 РМ Jм-вв PeL-sB М►г•вВ br-ss Мвr•вв Jиn-8a .hr1-88 Аvв-вР 8ел-вВ Qп-ВВ ►bv-ss рп-вв В1t11мв ef wггмb ЧпtгslNt Овt{ед влАз telвtк �О.в BS.s �1.о 1о.) ba.s М.О о1.7 11�.: 1?о.1 140.Е 1в?.в 1в9.в ивЧаеsl tгиwг:.: ia.1 1а.8 1t.s 17.7 ю.в 4в.о ао.а в3.е s1.в аа.а эв.i ае.1 8ееwlЧр 11.1 18.4 14.4 1�.0 • tб.a 1в.6 22.7 2�.6 21.в 2У.в 2t.6 ?8.Т и{:ввеегr вв.{.о. 4.о п,4 4.в а.т а.1 в.в е.1 о.э 1о.о ко.г ил 1..л СмОгв1 Вмtг 1Т.в 2D.4 4f.8 81.о 8в.2 �.0 в0.о в0.1 07.7 10'1.1 11s.8 18Q.6 t1Sa ►i/11a► (uetзq вwгввв et tvгввМ вnд feltas{пв вenW�a еsи/1о1 вьвА. иа>: Теа1: s.s s.o в.в в.в в.а в.а Т.1 в.1 е.в о.а о.в 1о.в йвЧааl t►ввwqrt 4.8 4.а 4.4 4.в ?.4 ?.2 2.t ?.4 4.? •1.? ?.? ?.а eвcwiNos 4.о 2.� i.a i.o � 1.7 1.7 1.t 1.7 1.в 1.в 1.а 1.в 001{рlзн{► 8tквlзlQв 0.1 0.4 0.0 0.8 O.s O.s О.в O.i 0.7 0.7 O.t О.о eмs..i е..а: а.1 а.в в.1 в.п ' в.4 в.а +.7 в.а в.7 7.i т.в в.о Ротем! е1 еммгоl{квв ФМк Теleh 1.ff lsft 1.вЕ 1.� 1.ая i.бt 1.6t t.бя 1.вя 1.Ыt 1.6s 1.7я юзиаа{ tг.гwгу: о.вя o.�t о.ы► о.вп о.ья о.Ея о.вя о.Ея o.as о.вt о.вя о.вя а.иn{а1м • л.вЕ о.в1г о.м о.ис , о.вя о.а о.п o.as о.ая л.м� о.:rя o.ss алnв.мгr sw{ив. o.1s о.1я o.1s о.1Е о.1к о.1я o.1t o.is о.1Е о.1я о.�я о.1к Смlмt 6>адг О.вя О.вя 1.оя 1.ОЕ • 1.оя 1.1я 1.1! 1.1я 1.2R 1.2f 1.?3 i.J1t Саеltкюд P►isв iuQn О.а7М О.о4? 1.Овв 1.41в 1.4о8 1.В11 2.?70 2.оЬ2 8.20в 8.417 8.677 а.700 ( Ровsмlввв еАвз►вв оwг еиевl{кю мгlеi 10.60Е 1l.80t 1в.в0! 17.81Е 1o.sOR 40.в8я 28.а1я 4о.77я о.Овя 6.7o1t в.7вя в.4tя �Ч Ф Рвго11s1 ивАмвв гвlв (AtJЧS4? s.i101 E.7ssi 6.840а в.ве10 а.4.�0 1о.4019 i4.f�8 1а.87Е1 i1.2qe U.oot.1 fa.RLt.a f6s?w { (Риtе0 оwг�о�) ваи►гsок Свдогоt Ева1г е! Чи Arpaslм RгооЫ 1с. . Ов-Мт•о0 TABLC в. AЯ�M1Wl: NATIOMЧ. t1a�RV A►D Сeтмлi еАм Ожi. .д4и-vМ 19в0. OssII! РИ ,hл-80 РвL-80 Мвиво Ауг-80 /br'80 Suu•80 .hrt-BO 1а�р-в0 Еер•вУ OCt-80 tb+-8V Ore-p0 9it11eм et ампм► Awl.ntи{ ов*tad мАе Teq1r 1ов.в ?8�.4 вов.4 в80.7 0в1.1 4кзо.8 58s1.0 1а10., вв5в.в S2s7.в ввов.7 веь7.п 1tвЧааl tвpwyrr Ч.О в8.т 1р►.4 1l8.0 4�?.8 в71.в овТ.в 10ов.! 1001А 11во.6 lвво.0 237У.4 О.мг/вiи 4У.в sS.6 70.Е 1в8.т t0e.0 ' в00.о оп.о о1п.Е oa0.t 1о3з.8 tавв.в s218.o вDtta.мn swt.o. ка.в te.4 sа.т по.в во.е то.в �o.s оа.в 1о1.а 1ов.в us.s 1в1.в Cмfnt @мR квп.е к�о.в 4оо.о аи.в жо.v 1ase.e 2+4s.s sa1a.4 ыа�.в <оsт.в авв.г ssтr.e ивв ыиг.а cd.tws...no. ее и».ма мв tellaiгq в+оа�b'о аиltвl вдеU. го!е)s tо�еlг о.т т.4 в.к п.в �.к в.г а.4 а.в s.z в.� в.1 в.в маЧмиt т...мг„ 4.4 4.о п.о 1.о к.1 к.1 1.в 1.а 1.в 1.в 1.в 1.в sкr.tвt.. к.в к.в к.в 1.в о.о к.о к.а к.• 1.ь 1.s 1.п 1.а олllв.м.r к�ь{г�. о.е о.е os о.в о.4 о.1 о.1 о.к о.п о.1 о.1 о.1 смв►ег ем« 7s п.в �.к а.е в.о в.о а.т а.о в.в в.п в.в в.о fБиrмв et вмимttме тРг 1оаг: 4.о8 4.пи 4.п 4.ж п.тr 4.ж к.в: к.вк 1.вs 1.ж 1.п 1.вк а.вtаио тг..wг» o.rn о.вs о.в1с 1.ое о.ж о.ж о.п о.п o.ss аа о..: о.вs tsw►101и о.вв о.1! , о.бf о.е! О.вf о.вs о.вs e.as o.sft o.ss о.а o.as oвtte.ae.r s../r�. о.пs o.4s о.пв о.пs o.1s о.1в o.os аа: о.оя о.сп o.os о.о1г Смвгвt в�м! 1.бв 1.вf 1.вf 1.ж 4.о'i 4.вй 1.Ох 1.trk 1.S1f 1.3t 1.31 1.МУ Сад{м1 вгlев 1hМв 4.0о� 4.ввlt 6.9ее т.Т2! 1l.оЕв в8.'►кв к00.00о 118.о36 izs.4l5 120.8г1 к8l.аве 1vв.60S г voгs.�e.o• а�в. м.. ож.еlео о.гtы f.va�c O.ook 1т.оаt • в&.ве+с о1.вв1{ iss.oex юz.етк zs.lтs в.оsх a.ssu л.1мг и.аат �' оо v� nагв11е1 •вгtапе гоЧ tлslusИ 16.rno 4i.вTSO и+.вRм г�а.отв� 1а2.вюп tов.амв вЧ.выо sw.lcn вго.t� coa.vwn nм.маs >~члs.,с�в t �Роиiеl sив�ов) вои►tОг Смtгвt 6мАе е1 !М АгQмв{м Rваllie. -28с- ' � �i�ë",�ixi ..G LQG �L"L�,Ë�L��L�''»+���: �t"г�'Ъ4� � н w.�в й«.д lV w V h« L .w :в v О ь. L« Ч w<. N Св о � О м i. �[ь О о s s О Ф« wнi 1� О О 8 8 8� Ф А i4 �д � �1 � « Rn.�i, и�' нi �..� . «�.;r,�оп••о� ci.+r;eic: • н w w н� � v 1 � l��L;lоlг ос�г�ла i�E3:�"ilS�`iSb'�$3� �ili�ЪS� + « ой£iww й� г: чс; а �д::одоо�о�,'� сi�сvйй ь� S � •, ! • w i� А О Ф�� ^ Ф w л w О W i в вн. Ф� О� Ф(О Ф f. О О � � • ��&2�1: •,аай Моwнео�лое;о» йrii:. � О � М 1 S CJ в«�в�0�� а0�:в�wн �p�w♦ y1wO0AQp т�о0и�о у� .s» 3«��� ооа�о.:.: аааломо�йл�.�й we«wн Яi $ � :°�,�'�,::�О", а�.оо ��о,�lрΡë���лi,��te`,� ��г�b',!S s, � �t��ij рwвJдьl 4i$р i�wi�Y^�w�w vp0�^ b� dSdsD� jj ja � ��Чр М о :+���� ro..wso �L���i:����ë�`� �ri���:�: g " ��ri.г3.�3 «.:�ci �$�i"�$4i$�$° :и�7•.�« � S.. � � .. �` lе3�:t�oрΡlо!�`орΡ, awoco ��г"�`:L",:��b`,$'�О S°o�q`iC�i� � н - Mi �F в♦ �°ww�i •�•в••w0 bО.вдвQ Фeww« .r � � � � Ё�i�w�itv bhлФФ �����i�r��� Wc� �wвОла � Ж�rSB$ '°'°�:::; ��:о«Фооо ;о o.o�S:3 � � � � g ::�:ZS�L`, лvаал !о"3.",�ffi�й��i,��}� 10�`,�СЬ`!� � . н ерΡ t p t� !ь N �ОΡ p а� в~в �itl f: hI. pawww ��aV paqOwi°i,p i рр йвi�вi .� � • М Zo',`:i�S��'и, ...mesa �i��lSb�:°.lг�,?�о�3"о �"��SЬ� � � М � " �irS�ФL° a�sos ���:•id�eO:w« iq.�•i М е 1 1 А �в � � � ~ pOf4О N 4i0 W�iP1. V AФIV f i wwp w .. .ааедд �°�е°".°,.�'бо�о w:::::.°. � � 1 1 w � • . . У А � � � � � • • S ^ � .,f � ' ;�._. • у �'g � 3�5� : �� � �г+s 8 ��ае о $s g ��ОΡ-$ Lq'� �� �� sgд�п �! .. � � �-��i � • 1�1 ■ j' О О О • р q� ��. • i V � �О�� �1= �_� О� ��i � � � � � s � в�; �;У: �j g� �� � ��в�1f�s а� � в � ���� § ' � а� �к s a�s в���$ ������ �а _ .� tlf g� � _����� �_�� . ��s�=���._g�рΡ- t�� "� � ��q`��_,g� .�� в а t• д$�п � ьш�и�дйа`й.йао $ gV � uaiY ��k Ь� �mxii�� � mri�i3� �� ��a�i� g�! � • л v й л' �j 17-Jo1-00 tABLE 7. AAfп/R1NA: COЛRAL ВАЧ: 51�Рк.U8. 148B:f-1060:1Y. 1пгD6:0? ивв 1968 1 1ов8 11 1489 1II 1981► 1У я98V 1 I48o II 1069 It1 10Во IY 1wt1в6}емАд3vвlед 8uиi-fisaвt Ewo1w в/ -R.804 -о.ЕК 3.b8x •1.� •8.i7lt .Ra.1'� 17.lа% -5.8St к .мв►iе Сореnм0е 'II.� •O.btt 1.а8а •1.оlК -T.11f -пп.8Вt 19.68t -4.bOt 1bьlмt Rвwtiк -о.76! o.1Sf 1.051< •0.98Х -б.б41i -п7.Ь2л 14.7Ьх 0.77f lгsМвув Rвb 01flвгмЪiвt 0.о08 o.o07i o.82i О.беl; O.OD�s O.ODi O.ODS О.ОО'i Inамвl lьbгиl P1ons •о.7вх 0.45Е 0.73ц •1.40!< -б.б4t •27.b2!< 10.75А о.77ц tn4stes4 Rивi04sк 11.491 1b.T1i 14.�% 9.978 14.15>< E8.98t 80.11t 18.04L Rйlиамtв в♦. RвОвlоbд Rвlи 8.64R 8.77!< 8.п0'Х п.�% а.90!< 1б.16lt R.Abt п.вбlt � 1пдв.-llо►.д lадiками п.4а1с 4.аоц о.овs a.eafc а.иа a.as 1в.ог8 а.о� оо1а.-tикя:.е,аеl.вовввв 1.71я п.воаs 1.амс о.ид 1.аох 4.гтк 1.г1х а.гоs вм~аи ts ttvвгд.Чеиг о.оах 1.1вя о.ваs о.ых о.аsк 1.ев�: a.sss о.1зП 1в«.. и оге.. аее е.вж 1s.1a пп.оls a.e1s 7.оьп �.в7и и.о�[ п.евs м�.е..вs tм аг«sslos. -e.ass -u.oss -пl.аеs -4.4вк -в.ык -п8.оох -4о.71я -п.П,>< t�гl,ptllу, о.ооп o.oos о.омс о.оац a.aas 1.авz о.оас о.пок о,.вгме„b F1мм. о.вtвs 1.вак 1.as 1.9юs а.�1с п?.аsя s.�s а.аоп о18., s.,9.г.и а.в.tв+д 4.оьи я.оаs o.eo�t о.ыП о.вrn в.вsк o.sn o.a1t я.ии.в г.,r.ввак -11.о1я -14.7ьк -18.1оя -1о.авк -пд.г» -e4.a1s -1а.еок -18.1» емег.l е.кrк eslis -о.аок -о.еsк -о.а1х •o.RVx -о.87к -11..os -a..os -s.нти я..м...ме ки..... -о.еок -7.sos -п.па+к -о.пюс -1.oas -7.в.х -1.мк -o.ess lмsииlы• о.оо.tи -1.a1s -�.cas •ю.паs -v.амг -1а.п7s -es.a7� -в.ачх -вsss 1 ацд,. ц�е...вв Р,r.н,ев о.омг о.оо: o.ooz о.оо�1 o.ofls о.ооlг о.оох -о.воs � 1ntlsiiaa Adjwtвen4 -f.б8ц -1.о0!< 0.35!< -о.10а •o.77ft L.14a 4.94l. •b.a5t lвtанок: ie.. аи А.ии -1о.tбп •1а.ооа -1в.п21< -в.tos -1п.в7s -вl.гсх -к8.1п .1а.оа>: 1 1s,r�.et« свlо еи ивы 11лtи о.1т 1п.9ва ае.а:хс а.191� 11.71s ев.49к 1т.оsк е.ае�1 t.t.r„d Севве„«s o.o2s o.oDS а.ооя o.oos -1.opt -о.7а -п.1оt -1.пвs t,евв Ргмlоiеи tм pN 6� -О.в71 -o.79t '1.06t -О.б�t 'О.вSf -1.60� •П.7ад -o.15t quвei-Fiseвt ЕvгРlиа вдуwЧl !ег �И lевои -п.�f •1.8мt O.iкD% -1.б7>< -8.91П -II4.781 11.б5х -5.08Х In�вrost-BвsHuo АвиЧ MaI 1.1s6it1d{и: Ави{в: � 81.LS 48.Бх 28.8! 88.9t 80.7� 10.4f 19.2t t4.td а.еьиегми: п4.ts пz.ss пп.оs 27.1s 81.7s кп.r<к ц.ак sв.ох двввlвlвд Rвйв 7.L>< 6.Ьlt 6 0>; 8.бх 7.Qlt 8.8t 8.оц 4.8f tг:еи-1tка.е о.в: o.4s а.яп 14.в: 1s.7s в.вs в.вs кв.еs ОstНипмьы a.es 4.as а.о�[ s.as a.as 1.se 1.ss 1.os 1111�,�е1� о.ох о.ох o.os о.ок о.ои о.о: o.oz о.1к св...к�.,и мккмао о.юс п.sк и.вя а.лк 7.о�с 4.т T.os Т.о: и.лиьыи: па.г1� пs.ss п7.�с ео.ок ю.1z а1.7п па.а: �.а�1 aw► еи1. o.os 1.п 1.мс о.тц oss 1.тs to.m o.as п..ви.вlве енаг авв..« пт.оус 1a.4s a.7s 1.ыс а.юс 1.ов 1.ех 1.ок lмеввиiбlв Оввевitв 2.Ot 10.08 п1.оlt �.4R 82.Ах 17.4t 14.Е8 18.� �sвинееее аа.имиеаиев ваиаавsпв Мгмгвмtавг р�Р tAt wi � 1 iм: av.wll:вd) 89Q077 Еб�?.б )о1дls/х. 1ЯВ.t' Fa9 lйМ4tА 6«"1о!!п8 а9�Ио0f bt?71aor r� 1•� satiмiбieд. 1./ .tf-Mrecмb preьisi�aa sp 1Ммt 1е вп! iисоМ !гм Чм �i. 27 Table 3. Ugentina: Percentage monthly increase in consumer pricest 1971-1988. 1972 1972 1973 2974 197S 1976 1977 1978 1979 Jan -0.24 5.23 4.59 -6.31 3.28 8.83 8.02 13-36 12.77 Feb 3.34 3.62 7.38 2.44 4.76 19.00 8.25 6.21 7.44 Mar 0.96 4.23 8.6S 1.20 7.58 37.60 7.55 9.49 7.75 Apr 0.95 4.83 4.07 2.3S 9.86 33.86 6.01 11.07 7*00 May 2.47 2.64 4.59 3.44 3.85 12.14 6.49 8.69 6.92 Jun 3.09 5.49 -2.60 3.34 20.99 2.74 7.63 6.50 9.69 Jul 4.33 C93 0.08 2.15 3S.20 4.24 7.37 6.60 7.15 Aug 2.S6 -0.07 1.71 2.10 22.27 5.Sl 11.32 7.81 11AS Sep 0.93 2.45 0.75 3.10 10.80 10.52 8.30 6.39 6.84 Oct 1.03 4.78 0.60 4.00 13.93 8.50 12.49 9.75 4.34 Nov 2.65 4.93 -0.40 4.81 8.80 7.93 9.04 8.79 5.14 Dec 11-90 8.81 8.12 11.93 19.S5 14.37 7.31 9.07 4.53 Mean: 2.83 4.24 3.13 2.88 13.40 13.77 8.32 8.64 7.59 Standard dew,iation: 3.01 2.12 3.50 3.83 9.09 10.72 1.79 2.06 2.48 1980 1981 1982 1983 1984 1985 1986 1987 1988 Jan 7.21 4.90 11.93 15.99 12.51 25.13 3.03 7.56 9.09 Feb 5.35 4.28 5.28 13.02 16.99 20.67 1.69 6.50 10.43 Nar 5.79 6.00 4.72 11,27 20.26 26.50 4.65 8.20 14.74 Apr 6.17 7.88 4.19 10.27 18.50 29.47 4.73 3.36 17.23 May S.78 7.54 3.06 9.06 17.08 25.12 4.03 4.17 15.72 Jun 5.74 9.37 7.90 15.82 17.91 30.54 4.55 8.00 3.7.96 Jul 4.57 10.2S 16.26 12.45 18.28 6.19 6.76 10.12 25-64 Aug 3.42 7.92 14.67 17.24 22.85 3.06 8.78 13.72 27.63 Sep 4.54 7.24 17.09 21.37 27.55 2.00 7.23 11.69 11-69 Oct 7.61 5.82 12.69 16.97 19.32 2.95 6.05 19.55 8.99 Nov 4.69 7.21 11.35 19.23 14.97 2.37 5.30 10.27 5.70 Dec 3.82 8.80 10.62 17.70 19.68 3.17 4.74 3.40 7.00 Mean: 5.39 7.25 9.98 15.03 18.83 14.68 5.13 8.88 14.32 Standard deviation: 1.21* 1.72 4.66 3.63- 3.63 li.43 1.82 4.47 6.68 overa I I mean: 9.13 overall standard deviation: 7.13 - 28 - Table 4. Argentinas Quarterly real exchange rates, 1976,-1989. (Real exchange rates are geometric averages of Argentina's bi- lateral real exchange rates with 14 trading partners constitut- ing 73 per cent of total trade, weighted according to trade for the sample period, based at 1987 = 100. Percentage figures below each real exchange rate indicate the average parallel-market premium.) Year average I II III IV 1976 74.4 95.9 73.2 63.9 68.3 83.1% 157.8% 77.5% 78.1% 37.9% 1977 88.7 90.1 93.0 87.4 84.6 1.0% 3.1% 1.1% -0.6% 0.3% 1978 76.1 85.3 79.4 73.3 67.6 -0.3% -0.3% -0.7% -0.7% 0.4% 1979 55.8 62.1 57.6 52.6 51.5 0.1% -1.0% -0.6% -0.7% 2.5% 1980 45.9 49.8 47.0 45.4 41.6 0.2% 0.2% 0.2% 0.2% 0.2% 1981 51.4 41.4 54.4 54.3 57.1 23.5% 0.2% 2.1% 45.4% 56.1% 1982 90.2 68.6 76.2 129.9 97.6 29.7% -1.4% 56.0% 30.2% 41.3% 1983 86.2 92.3 91.0 81.2 81.1 38.7% 34.9% 24.2% 59.6% 38.4% 1984 75.8 82.1 73.5 70.4 77.7 38.0% 45.3% 52.8% 38.0% 18.5% 1985 86.5 79.3 83.4 91.4 92.6 17.3% 25.4% 14.1% 18.5% 11.6% 1986 91.1 92.9 90.9 90.1 90.7 11.6% 11.5% 6.7% 9.8% 18.8% 1987 100.0 96.7 98.2 96.8 108.7 27.6% :5.6% 28.9% 35.4% 22.3% 1988 96.4 108.4 105.2 91.1 83.2 24.3% 33.8% 24.1% 20.8% 21.5% 1989 121.2 75.9 186.4 130.9 124.7 39.4% 77.8% 64.7% 7.0% 20.6% - 29 - 90 APPENDIX. 80312 UMCICt. 288=88 R8 GRING CZNTRAL-SM LOSSES A D 8ORROWING REQE : : g. 9.1. Over the 1980s Argentina's Central Bank undervent pro- gressive decapitalization, in the form of deepening indebtedness and progressive losses on assets. As a consequence of its decapitalization - that is, as a consequence of its loss of an independent net asset position -- the Central Bank lost the ca- pacity to carry out effective monetary policy. Having no mar- ketable assets, it could not absorb money except by borrowing at high interest. Its interest-bearing domestic indebtedness became so severe that it slipped into distress, with its inter- est bill running ahead of its capacity to place new debt except at higher and higher interest rates. This became a core cause -- arguably the core cause -- of the hyperinflationary pressure that emerged twice over the course of 1989. 9.2. This appendix discusses certain conceptual issues re- garding central bank losses and deficits. The expression "quasi-fiscal deficit" has been used interchangeably to refer both to the Central Bank's losses and to its borrowing require- ment. Nevertheless, it is helpful - indeed, important -- to maintain a clear distinction. 9.3. Like any other financial enterprise, a central bank may earn interest on its assets and pay interest on liabilities. (A central bank that has access to seignorage - the capacity to borrow at zero interest by issuing an obligation that people hold for its liquidity characteristics - should not need to borrow at interest.) In addition, central bank assets may take capital gains and losses, ar may revalue through currency fluc- tuations, or losses on account of non-payment. Such events give rise to profits or losses. In any time interval over which the central bank distributes no dividends nor receives additional capitalization, the flow of losses (profits) equals the flow de- crease (increase) in the central bank's net capital position. 9.4. To discuss central-bank losses and borrowing require- ments more precisely, it helps to work w1th several definitions and distinctions: (i) operating and non-operating losses; (ii)- external and domestic components of central-bank losses; (iii) accrual and cash bases for measuring central-bank losses; (iv) nominal and real bases for measuring central-bank losses; and (v) the distinction between central-bank losses and borrow- ing requirements. 9.5. *- O2eratina and non-operatina central bank losses. In Argentina, the principal source of central-bank operating losses has been the excess of interest paid on- obligations over inter- est received on assets.7 (A *classical" central bank whose lia- bilities were restricted to issues of non-interest-bearing base money wuld have no significant interest expenses, and should therefore generate no o arating loss through interest flows.) -30- Additional, but less significant, operating losses might arise through the central bank's own administrative and payroll expen- ditures. In some nations (such as Brazil in recent years) cen- tral banks carry on subsidy operations that may also give rise to losses. The important point about operating expenditures and net operating losses -- is that they require the central bank to issue debt or money as they are incurred; that is, the central bank must capitalize interest due into a debt stock or .pay -- i.e., issue -- money to cover it. (Vice versa for oper- ating receipts; the central bank either capitalizes them into assets or receives -- i.e., absorbs -- money in payment.) 9. 6. Central bank non-operating losses include such things as increases in the domestic currency values of net dollar de- nominated liabilities resulting from currency devaluation; re- ductions in recorded values of assets through recognition of value impairment; and reductions in market valuations of (capi- tal losses on) longer-term marketable (but sound) assets. In themselves, non-operating expenditures and losses reduce the central bank*s capital position, but do not directly require the central bank to issue new debt or money. 9.7. Operating and non-operating losses both reduce a cen- tral bank*s capital position. In principle, the distinction be- tween the two is that, unlike non-operating losses, operating losses must be financed directly, as incurred, through increased central-bank itet indebtedness. The flow increase in the central bank'-O-met indebtedness equals -the -flow increase -in the- central bank's total debt - currency in circulation, debt to commercial. banks, debt to the non-financial private sector, debt to the private sector (deposits of public entities, if any) , and exter- nal debt -- less the flow increase in central bank assets - in- cluding international reserves and commercial bank, private- and public-sector obligations to the central bank. 9.8. While non-operating losses do not force a central bank to contract additional net debt, over time the reductions they entail in the central bank-s capital position may affect its ca- pacity to act. Value impairment of, capital losses on, or ad- verse currency shifts involving central bank assets reduce a central bank's subsequent capacity to absorb money. 9.9. ii- External and domestic sources of central bank losses. A central bank may take operating and non-operating losses on both external and domestic activities. Operating ex- ternal losses consist principally of .the difference between in- terest flows on external liabilities and assets. Non-operating external losses consist mainly of the flow increase in the do- mestic currency value of external- liabilities resulting from de- valuation, if external liabilities exceed external assets; if external assets .exceed external liabilities,. devaluation gives rise to non-operating profits. Vale impairment of, or capital losses on, extetnal assets may also give rise to non-operating external losses. Profits or losses may also arise from mtiple - 31 - exchange rates. A central bank that purchasen foreign exchange at a high (local currency per dollar) rate and sells at a low rate - presumably, to set differential incentives favoring im- porters over exporters -- would tend to run an operating loss. 9.10. 111- Accrual and cash.bases for .c..entral.b.ank. oDeratinS losses. For simplicity,- suppose a central bank runs losses through interest due on a single obligation. Over any time in- . terval, this interest might te paid in cash, capitalization into the obligation stock, or a combination of the two. The losses . measured on a cash basis would be the interest actually paid out in cash, or, equivalently, the total interest accrued less the interest capitalized into the obligation stock. Income on assets can be treated similarly. Thus, Central bank operating losses, accrual basis = Central bank operating losses, cash basis + Interest due to be paid by the central bank accrued into liability stocks - Interest due to be paid to the central bank accrued into asset stocks, or, Central bank operating losses, cash basis = Central bank operating losses, accrual basis W Interest due to be paid by the central bank accrued into liability stocks + Interest due to be paid = the central bank accrued into asset stocks. The base money issued in the process of "financing" the central bank *s operating losses constitutes the flow "monetary impactm of the central bank0s losses. (Non-operating central bank losses have no direct monetary impact.) 9.11. 1v- Nominal and.real basesl for measurin gadomestic cen- tral bank losses. Where inflation is significant, the central bank generates a profit to the extent its liabilities denomi- nated in local currency exceed its assets denominated in local currency (or takes a loss to the extent its assets exceed its liabilities), because the inflationary erosion of the liabili- ties exceeds that of assets. Over any tine interval in which prices rise x per cent, each domestic currency unit loses [x/ (1o0+x)] per cent of its purchasing power. (If one gna is held over a period in which the price level rises by x per cent, its purchasing power value at the end of the period will be (100/(100+x) 3 times its initial purchasing power value, a loss of- [x/(l00+x) 3- per cent.) Accordingly, a given "nominal" cen- tral bank loss may be placed on a "real basis" by adding a flow equal to [x/ (100+x) ] per cent of the difference between the cen- tral bank0s domestic currency-denominated liabilities and assets. For this calcolation, AU.1 of a central bank's liabili- ties - in particular, non-interest-bearing monetary base - and assets should be taken into account. (The definition of the - 32 - quasi-fiscal deficit used by the Argentine authorities did not take account of the narrow monetary base, and to this extent overstated the BCRA's own real decapitalization. On this point see Barbone and Beckerman 1989.) 9.12. v- Central bank losses and the central bank*s borrowing requirement. Por some purposes, the borrowing requirement re- sulting from the central.bankts own activities nay be a more useful indicator of the macroeconomic pressure exerted by these activities than the losses arising from them. The point here is that, where a central bank has an asset position on which it ac- crues interest income but does not receive it in cash, the cen- tral bank's own operational flow may nevertheless imply a sub- stantial borrowing requirement, even if the operational flow shows a significant profit on an accrual basis. In a macroeconomic context, the borrowing requirements, not the un- derlying losses p= AS, measure the burden the central bank's activities place on domestic financial markets. 9.13. The central bank's own borrowing requirement may be measured as the sum of (a) its losses on an accrual basis and (b) interest receipts capitalized into the asset stocks rather than received in cash. (Note that the central bank borrowing requirement as defined here includes only what the central bank borrows to finance its own losses.) The central bankts losses will be less than the central bank's borrowing requirement to the extent the central bank is only accruing, not actually re- ceiving, cash payments of interest on its assets. 9.14. To discuss these points in analytical terms, define the following variables, all as period-end stock values (the nota- tion convention is that a prime indicates domestic-currency stocks and an asterisk indicates U.S.-dollar values): ' net obligations of the non-financial public-sector to the central bank; H' m obligations of the banking system to the central bank, including rediscount credit* e = exchange rate (domestic currency per U.S. dollar); A* =net external assets of the central bank; RO comercial-bank reserves; B# = the monetary base (i.e., B*-RO represents currency in circulation); U' = non-monetary financial obligations of -the central bank; and Vo = the central bank's capitalization. 9.15.. The central bank's balance sheet, accordingly, has the following strWcture: * 33 - F' Rf H' B'- R A0 = eA* U' V 0 9.16. Let the respective nominal rates of return on the assets and liabilities be given by f', hI, a*, r', and u' (there is presumably no return on B' - R). (The notation convention is that an apostrophe indicates a nominal domestic-currency rate and an asterisk a nominal U.S.-dollar rate.) Over a given short time period -- taken here to be one accounting accrual period -- the central bank's o2eratina profit in nominal terms is given by J (f' 7' + hO H' + e a* A*t) -W W-2It1 PlTV, l (The central bank's own operating costs - wages, supplies, etc. - are ignored.) 9.17. (The asset and liability stocks and their corresponding interest rates may be thought of as vectors. 'Thus, for example, f0 and F' would be vectors. incorporating rates of return on and stocks of different kinds of government obligation held by the central bank.) 9.18. The central bank's agn-oneratin profit, again in nomi- nal terms, is given by NO - (e - e.)A - A* d , [2] - 34 - where "d" represents the proportion written off the stock out- standing at the end of the preceding period of commercial-bank obligations to the central bank. The first term in this expres- sion represents the appreciation in local-currency terms of the net international-reserve position as a result of currency de- valuation. 9.19. For simplicity, assume that the central bank receives no additional capitalization nor pays out any dividends, so the central bank's capital position simply increases by the overall profit: V' -V = V' + N'. [3] 9.20. The "real" part of this profit flow is calculated by - subtracting the "inflation adjustment" of assets and liabili- ties. Let "x* represent the inflation rate (now in decimal terms) over any given period. Then the inflation profit over the period on inflation on assets and liabilities is given by -[x/(3.+x)J ][ + N1 + e A* * B# - Ut1], [43 F-j13'.1 *1 -1 '1 -11 4 vhich may be added.to the nominal profit flow to give the corre- sponding real-basis profit. Also, note that the real-basis profit is not the same as the deflated value of the profit flow. Note that this last expression takes account of all central-bank assets and liabilities, including the monetary base. - 35 * 9.21. The central bank's loss flow is the negative of For- mula (3). The inflation-adjusted loss flow, which measures the change in the central bank's real capital position, is given by Z = ZO -[x/(l+x)] [F'-1 + H'1 + eA*_ - B - U-], [53 where ZO,= -(Vt - V' ). 9.22. The monetary impact of the central bank's operating loss may be defined by the expression = (mr r* R'1 + M u' W ) + (M f'1 F* + ah h' ' + ma e a* A*.), [63 vhere ar and an are the respective proportions of the central bank's interest payments on liabilities actually paid in money, while a , mh and a are the proportions of the central bank's interest receipts on its net assets actually paid in money; JAB*sJ is then the change in the monetary base directly resulting from i', the central bank's operating loss. Since the propor- tions (1-ar) and (1- eu) of interest paid by the central bank are not paid in cash, and the proportions (-mf), (1-sh) and (1-ah) are not absorbed in cash but instead are capitalized into the respective assets or liabilities: m36 M lraR'r = r Rr1 ' [7] /aUO = c u U s 18] AN = ch h'1 R' H, £103 and A'A = ca e a*_ A*1f £11) where o r (1 r) UMU) C = (1-rn) cf mf c (1-n), b h ch =(1 -ah), and ca =(1 a The coefficients an may be called the "monetization" rates, and the coefficients a the corresponding "capitalization" rates. 9.23. In addition to the growth in each asset and liability that takes place through the capitalization described by the formulas above, each asset and liability grows or diminishes through monetary events and policy: bank reserves change as com- nercial banks receive or lose deposits; non-monetary obligations change as the central bank creates or absorbs them; credit to commercial banks and to the public sector, or international re- serves, inevitably change as the economy evolves. Let PPI represent the growth in F' resulting from causes other than in- terest capitalization, and so on for the other assets and lia- - 37 - bilities. The overall change in each asset or liability is then * given by = R [12] /Auf = (AU = au [13] 4Y = arI = a?("T [14] = IHO = &T [153 and, in domestic-currency terms, aAf aA' -3 /A,A f. E163 The overall change in the monetary base is given then by where iAB*l Iapex + ~JjI+ LJAPI E A' 17] -38- REFERENC8. 4 Barbone, Luca and Paul Beckerman. "Inflation, Monetary Policy and Quasi-fiscal Deficits: A Simple Model, with an Ap- plication to Argentina," unpublished paper, March 1989. Balilo, Tomas J. T. "The Argentine Banking Crisis of 1980, (unpublished IMF Working Paper, November 1987). Calvo, Guillermo. "The Perils of Sterilization," unpublished IMF Working Paper, March 1990. Canavese, Alfredo J. and Guido di Tella. "Inflation Stabilization or Hyperinflation Avoidance? The Case of the Austral Plan in Argentina, 1985-87," in Bruno, Michael, at al, eds., Inflation Stabilization: The Ex- perience .of_Israel. Arentina. Brazil,. Bolivia .and Mexico, (MIT Press, Cambridge, Mass., 1988), pp. 153-190. Dornbusch, Rudiger and Mario Henrique Simonsen. "Inflation Stabilization with Income Policy Support. Working Paper. National Bureau of Economic Research, 1987. Giorgio, Luis Alberto. "Crisis Financieras, Reestructuracion Bancaria.e hiyerinflacion en la Arentina," working paper, -December 1989. Heymann, Daniel. "From Sharp Disinflation to Hyper and Back: The Argentine Experience, 1985-1989.* Working Paper, United Nations Economic Commission for Latin America, 1989. International Bank for Reconstruction and Development. Arentina: Social Sactoars-in Crisis (IBRD: Washington, 1988). Riguel, Miguel. "Ups and Downs in Inflation: Argentina Since the Austral Plan. Working Paper. World Bank, 1988. iguel, Miguel and Nissan Liviatan. "The Inflation- Stabilization Cycles in Argentina and Brazil." Working Paper. World Bank, 1989. Kiguel, Miguel and Pablo Andres Neumeyer. "Seignorage,. Inflation Tax and the Demand for Money: The Case of Ar- gentina." Working Paper. World Bank, 1989. - 39 - Machinea, Jose Luis and Jose Maria Fanelli. "Stopping Hyperinflation: The Case of the Austral Plan in Argen- tina, 1985-87," in Bruno, Michael, t AI, eds., Infla! tion Stabilizatign: The Experience of Israel Argen- tina. Brazil. Bolivia and Mexico, (MIT Press, Cam- bridge, Mass., 1988), pp. 111-152. Piekars, Julio A. "&l deficit guasi fiscal del Banco Central," paper presented in the Seminar on Central Bank Quasi-Fiscal Operations (Brasilia, August 1987), Center for Latin American Monetary Studies. Rodriguez, Carlos A. and.Aqiles Almansi. "Reforma onetaria y Einanciera en -Hiperinflacion." Center for Macroeconomic Studies of Argentina (CEMA), Serie Documentos de Trabaio No. 67 (August 1989). Sargent, Thomas and Neil Wallace. "Some Unpleasant Monetarist Arithmetic," Federal Reserve Bank of Minneapolis Quar- terly Review, Fall 1981, pp. 1-18. World Bank. "Argentina: Tax Policy for Stabilization and Economic Recovery," Report No. 8076-AR, Novem- ber 8, 1989.