2 The Argentine Banking Crisis of 1980
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

Abstract

In March 1980, one of the largest private banks in Argentina—Banco de Intercambio Regional (BIR)—failed; within a few days, the Central Bank had to intervene three other major banks, two of which were subsequently liquidated.2 Thus began a serious crisis of the Argentine financial system, which resulted in the liquidation of 71 financial institutions over the next two years and caused far-reaching changes not only in the financial system but also in economic policies. The restructuring process is still continuing. Most of the authors who have analyzed these developments have concentrated on the broad macroeconomic aspects; a few others have dealt with selected features of the Argentine financial sector in the context of the crisis. This chapter integrates these two sets of analyses and focuses more closely on the financial sector by emphasizing the regulatory aspects that previous studies have largely ignored.

I. Introduction and Overview of the Literature on Argentina’s Financial Crisis

In March 1980, one of the largest private banks in Argentina—Banco de Intercambio Regional (BIR)—failed; within a few days, the Central Bank had to intervene three other major banks, two of which were subsequently liquidated.2 Thus began a serious crisis of the Argentine financial system, which resulted in the liquidation of 71 financial institutions over the next two years and caused far-reaching changes not only in the financial system but also in economic policies. The restructuring process is still continuing. Most of the authors who have analyzed these developments have concentrated on the broad macroeconomic aspects; a few others have dealt with selected features of the Argentine financial sector in the context of the crisis. This chapter integrates these two sets of analyses and focuses more closely on the financial sector by emphasizing the regulatory aspects that previous studies have largely ignored.

This section provides a brief overview of selected literature on the Argentine financial crisis. The next section presents evidence on the macroeconomic and general business environment, and Section III analyzes the crisis itself. Section IV presents the conclusions. The research strategy followed in the paper reflects the belief that an appropriate analysis of the Argentine banking crisis has to look beyond the financial system, because a main cause of the crisis was the deterioration of the system’s loan portfolio—a deterioration that stemmed in part from developments in the general economic environment and their effects on business conditions.

Overview of the Literature

The Argentine financial crisis has been discussed mainly in the context of the economic policies implemented between 1977 and 1981, with only a few studies focusing on the financial system.3 This section briefly reviews the different approaches that have been followed in the literature. Fernández discusses three causes of financial crises: inadequacies of free market economies, inappropriate monetary policy, and inherent instability of the financial system.4 For him, the latter best explains Argentina’s financial crisis: under a fractional reserve system with state deposit insurance, financial institutions in trouble can delay failure by resorting to “liability administration.” Liability administration is described as the policy of an institution that, facing a cash shortfall caused by nonperforming loans, must raise interest rates to attract new deposits to replace those maturing. An explosive situation can result.

According to Fernández, this situation arose during the Argentine financial crisis as firms defaulted on their bank loans. These defaults were due to enterprise failures, which, in turn, were caused by frustrated expectations over macroeconomic policy and over the relationship between interest rates and the rate of change in the price of each enterprise’s product. Finally, Fernández finds merit in Simons’s proposal for financial reform, under which sight deposits would have a 100 percent reserve requirement to preserve their liquidity while time deposits would be replaced by bank acceptances or shares whose value would be market determined, that is, time deposits would resemble mutual fund shares.

Commenting on Fernández’s paper, Feldman (1983) suggests that the financial crisis should not be interpreted in isolation but should be viewed as a reflection of the real sector crisis that resulted from the growing incompatibility between real domestic interest rates and the rates of return of investments in domestic assets. Furthermore, Feldman suggests that high domestic interest rates largely reflected risk premiums owing to devaluation expectations that exceeded the Government’s preannounced, and realized, devaluation rates.

Using the framework developed by Minsky (1977), Dreizzen (1984) constructs several indicators to study the financial situation of a sample of industrial firms. He concludes that between 1977 and 1983 the financial structure of firms became increasingly fragile and had to endure strong destabilizing shocks, most of which originated in the financial market. Among the shocks that took place at different times in that period, he cites restrictive monetary policy, bank failures, and devaluations.

Another paper, by Petrei and Tybout (1985), examines firms’ indebtedness by analyzing a sample of industrial firms during the period 1976–81. They suggest that these firms originally obtained huge financial subsidies because of “real currency appreciation and unconstrained access to foreign credit, then again in 1981 because of an exchange insurance program.” Moreover, the accelerated growth in the debt/capital ratio in 1980, coinciding with lower profit rates for the firms in the sample, contributed to the development of the crisis.

Arnaudo and Conejero (1985) compare the performance of the three banks that failed in 1980 (Banco de Intercambio Regional, Banco de Los Andes, and Banco Internacional) with one another and with the average performance of private domestic banks. They conclude that several indicators warned of the impending failures of these three banks.

Damill and Frenkel (1987) further develop some of the points raised in the earlier literature. They suggest that the negative real rates of 1979 prompted firms to borrow more, thereby increasing their fragility. This fragility was exacerbated by the short maturities of loans, which made firms particularly vulnerable to “exogenous shocks,” such as higher interest rates induced by changes in exchange rate expectations. The consequent loss of profitability of some sectors increased the share of nonperforming loans in bank portfolios. Moreover, some banks that went bankrupt in 1980 had spectacular rates of growth, which the authors attribute to “speculative elements.”

The literature just reviewed suggests that the economic environment, as well as factors intrinsic to the financial system, led to the crisis in the Argentine financial system.

II. The Economic Environment

The 1976–82 period saw many economic changes in Argentina. The economic team that came to office in March 1976 inherited a high rate of inflation, a serious balance of payments problem, and a substantial fiscal deficit. The new team not only adopted short-term measures to cope with this situation but also carried out some important structural reforms, the most radical of which took place in the financial sector. Many of these measures and reforms were blamed for the serious difficulties that surfaced when the financial crisis erupted. As a result, beginning in March 1981, economic policy was significantly modified and most of the reforms begun in 1976 were scrapped.

Table 1.

Phases of Reference Cycles, 1961–82

article image
Source: Arranz and Elías (1984).

To facilitate assessment of the contribution of changes in the economic background to the financial crisis, this section presents some evidence on the behavior of the economy between 1976 and 1982: (1) the evolution of GDP and its major components; (2) the evolution of monetary and credit aggregates; (3) the behavior of some important prices (interest rates, the exchange rate, and some asset prices); and (4) the evolution of enterprise debt.

Evolution of GDP and Its Major Components

Consideration of the evolution of GDP and its components is relevant for the analysis of the financial crisis on at least two grounds. First, a significant body of literature produced by Kindleberger, Minsky, and others asserts that financial crises are an integral part of the business cycle, because they are a necessary consequence of the previous boom. Second, a downturn in economic activity can reduce firms’ sales and profits in the affected sectors, compromising the liquidity and solvency of these firms and, therefore, their ability to service their debt.

Table 1, which presents the phases of reference cycles for Argentina from 1961 through 1982, shows that the financial crisis followed an economic expansion: the failure of the Banco de lntercambio Regional in March 1980 came immediately after the expansionary period that had lasted from March 1978 to February 1980.5

The national accounts data confirm that the financial crisis came after a period of economic expansion: although GDP fell by 3.4 percent in 1978, it grew at rates substantially above historical performance in 1977 and 1979; between 1976 and 1979 the growth rate averaged 3 percent. The crisis cannot be dated so easily with reference to investment, however, because investment kept growing at a rate much above historical averages even after the crisis started; only in 1981 did investment begin to fall sharply. Finally, there were sharp differences in the performance of many sectors; in particular, the financial sector grew much faster than total GDP between 1977 and 1980 but shrank sharply between 1981 and 1983. Moreover, in any given year, rates of growth varied substantially across sectors. These sharp fluctuations in performance undoubtedly required various economic sectors to make substantial adjustments, which may have affected their debt-servicing capacity.6

Behavior of Monetary and Credit Aggregates

Table 2 presents data on the behavior of monetary and credit aggregates in the 1976–82 period. These data are useful not only to complete the general background but also to analyze whether monetary factors could have caused the financial crisis.

The table presents the data in nominal terms, from which two main conclusions can be drawn: (1) that monetary and credit aggregates grew very fast throughout the period, and (2) that time deposits significantly increased their share in M2, jumping from 36 percent in 1976 to 71 percent in 1981.

The data in Table 2 suggest that the crisis cannot be attributed to monetary causes: money grew in nominal terms at a high rate both before and after the crisis, although the rates of growth of M1 and M2 fell sharply in 1980, once the crisis was under way. The data for domestic credit show a similar pattern.7

Further evidence on monetary developments is provided in Table 3, which presents data on the behavior of the ratios of currency/money and of excess reserves/liabilities subject to reserve requirements. According to the monetarist hypothesis, changes in these ratios caused monetary contractions in the United States that resulted in financial crises and recessions.8 This hypothesis implies that the rise in these ratios depressed the money multiplier, inducing a fall in the money supply or in its rate of growth.

Table 2.

Main Monetary and Credit Aggregates, 1976–82

(In thousands of australes)1

article image
Source: Central Bank of Argentina, Boletín Estadístico, various issues.

The austral is the currency unit introduced in Argentina in 1985 to replace the peso (

article image
1 = $a 10 million).

Includes outstanding balance of “Cuenta de Regulación Monetaria’ (Interest Equalization Fund).

Includes loans in foreign currency.

Table 3.

Ratios of Currency to Deposits and Excess of Bank Reserves to Reserve Liabilities, 1976–82

article image
Sources: Central Bank of Argentina, Boletín Estadístico, various issues; Gaba (1981); and International Monetary Fund, International Financial Statistics, various issues.

The series begins with the reintroduction of fractional bank reserves in June 1977.

Period average.

End-of-quarter data.

The data do not support this hypothesis for the Argentine financial crisis. As shown in Table 3, the currency/deposit ratio was lower by the end of the first quarter of 1980 (when the Banco de Intercambio Regional was closed) than in the same period of the previous years; although this ratio tended to rise during the rest of 1980, it still remained below the levels of 1976 and 1977. This situation suggests that the crisis undermined the public’s confidence in the safety of deposits only briefly. A study of the period from June 1977 to June 1981 finds evidence of a shift in the intercept of the estimated function for the currency/deposit ratio (Demaestri (1982)). This shift was positive for April and May and negative for June 1980. Another study of the period from May 1978 to March 1982 (Dabos and Demaestri (1983)) found that the demand for currency, in real terms, became less sensitive to the deposit interest rate beginning in March 1980 (i.e., before the liquidation of the first major bank but after the liquidation of one of the largest financial companies).

The excess reserve ratio fluctuated widely between the end of the second quarter of 1977 (the start of the financial liberalization) and the end of 1982. Even after the outlier observations for the second quarter of 1977 and for the whole of 1982 are excluded, the ratio remained volatile: for instance, it dropped from 1.88 to 0.20 between the second and third quarters of 1978.9 Despite this volatility, the data for 1979 are well within the historical range of values, suggesting that bank reserve behavior was not a contractionary influence on the money supply.10 More important, the fact that the behavior of bank excess reserves in 1980 shows no departure from the historical pattern suggests that the crisis did not reduce the willingness of banks to invest in yield-earning assets.

The Behavior of Interest Rates

Several authors have blamed interest rate behavior during and after the liberalization experience for the business failures that occurred between 1977 and 1982. The purpose of this section is to present evidence that can help to evaluate the merits of that explanation.11

Nominal interest rates jumped dramatically when interest rate controls were lifted at the end of the first half of 1977: both deposit and lending rates more than doubled from the first to the second half of 1977. It is important to gauge the volatility of interest rates before the crisis, because a high volatility could frustrate business planning and undermine the financial health of firms. The monthly variability of interest rates within each year, measured by the coefficient of variation, was lower during 1979 and 1980 than at any other time between 1977 and 1982.12 This lower variability was probably caused by the policy of preannounced devaluations between December 1978 and March 1981, which encouraged interest rate arbitrage by reducing exchange rate risk.

The data in Table 4 show that annual average lending rates were positive in real terms during the liberalization period. Also, their highest value (4.87 percent a month, using the wholesale price index—WPI) was attained in the last quarter of 1979; moreover, the annual average of lending rates peaked in 1980. Therefore, high lending rates helped to precipitate and aggravate the crisis by making debt servicing more difficult. However, although these rates were positive on average, and sometimes very high, they were negative during many quarters. Thus the effect on each firm depended not only on the sign and size of its net financial position but also on the pattern of this position through time.

The behavior of real lending rates raises several issues. First, did the high loan rates reflect high deposit rates as well as high spreads? Second, why did enterprises not switch to foreign borrowing? Third, why were enterprises prepared to borrow at rates much higher than the marginal rate of return on investment?

Table 4 indicates that, although lending rates fluctuated in line with variations in deposit rates, the large and volatile spreads contributed significantly to the observed high lending rates. In order to explain the high spreads, Gaba (1981) broke them down into the cost of reserve requirements, the cost of excess reserves, and the gross financial yield for the bank (after taking into account the effect of non-interest-bearing deposits). He found that, except for the second half of 1977, when the high reserve requirement accounted for the largest share of the spread, by far the largest component was the gross financial yield required to cover administrative costs and profits.

Some studies have identified high administrative costs as a significant component of the spreads.13 Although these costs, shown in Table 5, are sizable, a large residual component of the spread remains to be explained. One explanation of the residual is that banks had some monopoly power, which they exploited by charging rates that included a monopoly rent. On the surface, this explanation is appealing: bank services are differentiated products that offer scope for imperfect competition (e.g., it is costly for borrowers to switch banks). A corollary of this explanation is that more competition should lower spreads. In this regard, the liberalization measures, which not only freed interest rates but also eliminated barriers to entry, should have undermined the monopolistic position of the banking system and thus lowered spreads. But it could be argued that whereas eliminating entry barriers reduced monopolistic rents, dismantling interest rate controls allowed banks to better exploit whatever monopoly power they retained. Thus the net effect of the 1977 reform on the rent component of spreads appears to be ambiguous. Another possible explanation is that high spreads reflected risk premiums on loans, which rose because of uncertainties about the course of economic policy and because banks knew very little about many of their new clients.14 This explanation helps to explain why borrowers were willing to pay high real rates,15 but it is inconsistent with the explanation that borrowers were willing to pay high rates because they expected a government bailout—because if bankers shared that expectation they would have viewed their loans as low-risk assets and hence charge a low-risk premium.16 A detailed analysis of the causes of high spreads is beyond the scope of this paper.17

Table 4.

Real Interest Rates, 1974-82

(In percent a month)

article image
Source: Central Bank of Argentina.

The formula used for deflation was: Real rate = (nominal interest rate - rate of inflation)/ (1 + rate of inflation).

Corresponds through the first semester.

Table 5.

Administrative Costs of Financial Institutions, 1981–821

article image
Sources: World Bank (1984), and author’s own estimates.

Administrative costs a year as percentage of total loans.

The weights used are the shares in total loans as of end-December 1982.

Although the high spreads made domestic loans appear more expensive than foreign loans, firms did not switch to foreign sources of credit, in part because the preannounced devaluation schedule (described in the next subsection) did not eliminate the uncertainty over the course of the exchange rate. In this regard, several authors18 have suggested that a perceived exchange risk discouraged foreign borrowing. This reasoning is consistent with the fact that the domestic deposit rate exceeded the yield in pesos of dollar deposits abroad until February 1981,19 thus indicating that borrowers and depositors shared similar expectations over the course of the exchange rate. Another reason is that probably only large enterprises had direct access to foreign credits (Petrei and Tybout (1985)). The evolution of interest rate differentials is depicted in Table 6, where the U.S. prime rate was taken as the representative rate for borrowers and the treasury bill rate as the representative rate for deposits. Analyzing a similar data set, Blejer (1982) concluded that the Argentine financial market was informationally efficient but that an uncorrelated time-varying risk premium was present between June 1977 and August 1981.

Several hypotheses have been offered to explain why firms borrowed at relatively high real interest rates. Besides the high-risk and bailout hypotheses already mentioned, the explanations include distress borrowing by firms in difficulties and speculative borrowing induced by devaluation expectations.20

Table 6.

Differentials Between Domestic and Foreign Interest Rates, 1977–82

(Quarter averages; in percent a month)

article image
Sources: International Monetary Fund, International Financial Statistics, various issues; and Central Bank of Argentina.

Adjusted for the actual depreciation of Argentina’s currency.

Liberalization of Foreign Sector

The policies followed in the external sector, especially exchange rate policies, had important effects on the soundness of the financial system, directly by influencing the capital flows and the value of the foreign debt of firms, and indirectly by dramatically changing many relative prices in the economy—in particular, asset prices.

The lackluster record in the fight against inflation prompted the authorities to use the exchange rate as a stabilization instrument, beginning in December 1978 and ending in March 1981.21 For this purpose the Central Bank periodically published a schedule of daily devaluations for given periods of time. This strategy was intended to make tradable prices follow international prices (nontradables would follow a similar path, assuming some substitutability between tradables and nontradables). Moreover, the preannouncement of the rate was expected to reduce uncertainty and to help make domestic interest rates move in line with foreign rates.

The first exchange rate schedule covered the period through August 31, 1979. Several others followed, carrying a declining rate of devaluation. On October 1, 1979 the third schedule was issued: it set a 2.8 percent devaluation for January 1980, to be reduced by 0.2 percentage point a month through the end of 1980. In September 1980, however, the authorities revised the schedule and announced that the devaluation rate that had been set for October (1 percent) would be maintained for November and December. This policy led to a substantial real appreciation of the peso, which began to affect the exchange market.22

After some failed attempts at restoring confidence in the exchange policy, a new economic team scrapped the policy of preannounced devaluations and sharply devalued the peso (by about 23 percent). In addition, the team stated that, in the future, they would follow a policy of frequent minidevaluations. Despite these statements, further reserve losses prompted a new devaluation of about 23 percent on June 2. Twenty days later, the foreign exchange market was split in two: a commercial market, with the rate to be set by the Central Bank, and a financial market, with the rate to be determined by the market.23 Transactions to be carried out in the latter market included most new financial transactions, sale of a specified fraction of the proceeds of some exports, and other transactions that were not allowed to be made through the commercial market. In order to reduce the effect of past devaluations on private foreign debt and to encourage renewal of foreign loans, the Government compensated borrowers for the effects of the June devaluation on loans that were rolled over for at least one year. To encourage further rollover of foreign debt, the Government also established an exchange insurance facility; to qualify, new loans or loan renewals had to have a minimum maturity of one and a half years. In the last quarter of 1981, the scope of the exchange insurance facility was broadened and a swap facility was established for six-month operations.

The administration that came to power in December 1981 returned to a more liberal exchange system, unifying the exchange markets, eliminating the exchange insurance and swap facilities, liberalizing sales of foreign currency, and announcing that the peso would be allowed to float. In April 1982, however, drastic exchange controls were imposed to cope with the problems caused by the conflict with Britain over the Falkland Islands/ Malvinas. Subsequently, a new administration, inaugurated in July 1982, reintroduced dual exchange markets, one commercial and the other financial. The Central Bank set the rate in the first market, but in the second it intervened solely as a buyer, at a predetermined rate. Gradually, however, operations initially assigned to the commercial market were transferred to the financial market, and in November both markets were merged.

These policies substantially affected the real effective exchange rate, as depicted in Table 7. These data show a real appreciation of the domestic currency of about 31 percent (export-weighted index) or 35 percent (import-weighted index) between December 1977 and March 1980 (when the financial crisis surfaced). This appreciation continued until the end of 1980 to a cumulative total of 40.5 percent (export-weighted index) or 42.5 percent (import-weighted index); in January 1981, the effective exchange rate began to fall as a result of the faster devaluation of the peso, a process that continued for the rest of the period covered in Table 7.

The uncertainty introduced by shifts in exchange rate policy was compounded by the changes in the degree of openness of the economy to capital and trade flows. Between 1976 and 1982, the openness of the economy to capital flows changed several times. The first period, from April 1976 to March 1981, was characterized by a dismantling of restrictions and a higher degree of integration of the domestic and international financial markets. The problems that surfaced in 1981 and 1982 in the foreign sector increased the risk of foreign borrowing for both borrowers and lenders and reduced that integration; this reduction was aggravated by the increased exchange rate volatility already mentioned, by the dwindling reserves, and by the measures taken in 1982 to restrict debt repayments.

The liberalization of capital flows was coupled with liberalization of trade between 1976 and 1981, but the trade liberalization was never so comprehensive as financial liberalization. Two main pieces of trade liberalization were a program of phased tariff reductions to be carried out between 1979 and 1984 and the elimination of import prohibitions. However, the newly inaugurated administration abandoned the tariff reduction program on April 1, 1981.

Table 7.

Evolution of the Exchange Rate, 1977–82

article image
Source: Central Bank of Argentina.

Based on monthly averages of the nominal exchange rate adjusted by price (WPI) and exchange rate movements in Argentina’s major trading partners.

From June 22, 1981 to December 24, 1981 and from July 5, 1982 to November 1, 1982 the foreign exchange market was split into a commercial and a financial market. For effective exchange rate calculations, the rate actually applicable to trade transactions (i.e., commercial or mixed rate) was used.

Wage Policy

The wage policy followed during much of the liberalization period aimed at adjusting salaries more or less in line with inflation, as measured by the consumer price index. This policy, combined with the exchange rate regime just described, dramatically increased real labor costs, particularly in dollar terms: real wages in terms of dollars increased by almost 177 percent between December 1978 and December 1980,24 while real wages in terms of the CPI and the WPI increased by only 23 and 54 percent, respectively.25,26 The result was a loss of competitiveness of Argentine tradable goods, which deteriorated the foreign sector position and the economic solvency of producers of those goods.

Changes in Asset Prices

The purpose of this section is to investigate whether a speculative bubble could have played a role in the crisis.27 A speculative bubble that drove up asset prices could have induced people to borrow in order to purchase those assets. When the bubble burst, the price of assets would suddenly have become lower than the value of the counterpart loans.28 This situation would have made borrowers unable—and unwilling—to repay their debts, while banks would have found that foreclosing did not allow them to recover their credits in full.

Table 8 presents the prices of some assets deflated by the wholesale price index (WPI) and by the peso/dollar exchange rate. These deflators are particularly useful because of the existence at the time of substantial public holdings of financial assets and liabilities linked to the WPI or the U.S. dollar. Thus, the ratios in Table 8 also illustrate the opportunity cost of holding real assets relative to holding such financial assets (or liabilities). The first two columns, which correspond to the price of cars, illustrate the effect of the appreciation of the peso over much of the liberalization period (as well as the restrictions on car imports). Whereas cars appreciated by only 18 percent vis-à-vis the wholesale price index between the second quarter of 1977 and the first quarter of 1980, they appreciated by almost 170 percent vis-à-vis the U.S. dollar over the same period. Furthermore, in this period the rate of appreciation vis-à-vis the WPI was negative in several quarters, whereas the appreciation vis-à-vis the dollar was continuous except for two quarters (last quarter of 1977 and second quarter of 1979). The appreciation vis-à-vis the dollar extended until the end of 1980, just before the policy of preannounced devaluation began to change in February 1981.

Table 8.

Selected Relative Prices of Assets, 1977–84

article image
Source: Central Bank of Argentina.

Apartments appreciated steadily both with respect to wholesale prices and to the dollar (third and fourth columns of Table 8): between the second quarter of 1977 and the first quarter of 1980 apartments appreciated by 34 percent vis-à-vis wholesale prices and by 207 percent vis-à-vis the dollar; by the end of 1980 these rates of appreciation had reached 46 percent and 307 percent.29 The last two columns correspond to cattle prices.30 Again, although cattle appreciated by about 20 percent vis-à-vis wholesale prices between the second quarter of 1977 and the first quarter of 1980, its appreciation vis-à-vis the dollar was much higher—175 percent over the same period.

In summary, Table 8 suggests that no general pattern was evident for asset prices until the second quarter of 1981, when the substantial devaluations of the peso that had taken place during that quarter caused asset prices to plummet in terms of dollars; the dollar prices of assets remained substantially below the 1979–80 values throughout the period under analysis. Apartments depreciated both in terms of dollars and of the wholesale price index: the latter fall began in the first quarter of 1981, and the real price of apartments never returned to the high levels of the period that began in the second quarter of 1979 and ended in the first quarter of 1981.

Table 9 presents series on the index of value of the stock market. This index reflects the value of the outstanding shares of all firms listed in the Buenos Aires Stock Exchange valued at the latest available price. The first column shows the average value of the index over each month; the other two columns present the deflated values using the wholesale price index (second column) and the exchange rate index (third column). The series displays a high degree of variability;31 nevertheless, the index appreciated, both in terms of the wholesale price index and of the dollar, between the last quarter of 197732 and the first quarter of 1980—which is the peak period of the two series. The behavior of the stock value index differs sharply from the behavior of the relative price of the other assets just discussed. Stock prices dropped in both nominal and relative terms (compared with wholesale prices and the U.S. dollar) immediately after the start of the financial crisis (March 1980); in contrast, the dollar price of the assets included in Table 8 continued to increase while their price relative to wholesale prices showed no significant change in the next few quarters after that date.

Table 9.

Stock Market Behavior

(Base: January 1978 = 1)

article image
Sources: FIEL, Indicadores de Coyuntura, various issues; and Bolsa de Comercio de Buenos Aires, Boletln, various issues.

The evidence in Tables 8 and 9 does not support the hypothesis that a bursting speculative bubble caused the first episodes of the crisis (i.e., the bank failure and interventions of March-April 1980). However, the value of enterprises declined dramatically after the crisis began, partly as a result of the uncertainty induced by the crisis. This decline deepened the financial crisis by reducing the value of collateral and making it more difficult for enterprises to substitute capital market financing for banking loans.

The Combined Effect of Changes in Interest Rates and Relative Prices

The variability of interest rates and relative prices, including the prices of assets, complicates the analysis of the effects of lending interest rates on various sectors during the liberalization period. Gauging these effects helps in evaluating the merit of some demands for debt relief, which were based on the assertion that extended periods of high interest rates had driven many borrowers into insolvency. A way to analyze this issue is to compute the cumulative effect of real lending rates on a loan, using different price indices. This is done in Table 10 for selected prices. Figures in that table correspond to the ratio between the value of a loan33,34 and the price index indicated at the top of each column. Therefore, if a ratio rises between two dates, the debt burden of the borrower rises, measured as the units of the price index that would be needed to repay the loan. Alternatively, in the case of an asset, the rise in the ratio indicates what yield of the asset, compounded and net of depreciation, would have allowed borrowers to keep their wealth unchanged between the chosen dates.

The data in Table 10 show substantial volatility in most of the series: in particular, this is so when wages and cattle prices are chosen as denominators (first and sixth columns). The data also indicate that the debt burden, measured by the wholesale price index or by the stock price index, had begun to increase about two quarters before the first episodes of the crisis and continued to do so until the end of the first quarter of 1982. The next quarter the debt burden began to fall because of a rise in stock prices and a fall in real interest rates (the latter resulted from the establishment of interest rate ceilings in July 1982). For the same reasons, the debt burden measured in dollars also went down dramatically after the first quarter of 1982, using the exchange rate as the denominator (column 3).35

Enterprise Debt

The previous discussion suggests that economic agents had to face important shocks during the financial liberalization, which included significant changes in key relative prices—including assets—and, during some periods, high borrowing costs. The vulnerability of firms to such shocks depends, inter alia, on the ratio of debt to total assets (“gearing ratio”); a high ratio not only magnifies the effect of interest rates on firm profits but also indicates that the firm has a low capital base to absorb losses.

Petrei and Tybout (1984) have analyzed the evolution of the gearing ratio and similar indicators between 1976 and 1981. Their analysis is based on financial statements of 155 publicly traded industrial corporations, classified into three major categories: exportable goods producers, importable goods producers, and nontradable goods producers. Their work suggests that, except for exportables, the gearing ratio increased over the liberalization period, whereas liquidity fell and foreign debt increased as a proportion of total assets. They conclude:

During the late 1970s, firms appear to have substituted dollar debt for peso debt, keeping their overall leverage stable. But beginning in 1980, when earning rates fell sharply, firms steadily increased their reliance on debt finance. So this year and thereafter, some of the increase in firms’ financial riskiness may have been due to distress borrowing. The beginning of the upward leverage trend corresponds to the emergence of banking sector crises, and may well have been a causal factor, (p. 50)

Table 10.

Real Interest Rates for Selected Price Indices, 1977–84

(Base: January 1978 = 1)

article image
article image
Sources: Raw data provided by the Central Bank of Argentina and author’s own estimates.

Value of loan deflated by different price indices. The loan takes a nominal value of one in January 1978 and its quarter-to-quarter variations are due to accumulated interest (at the average market lending rate). Nominal values of the loan are deflated by the price indices indicated in the table.

Using the same data base, Dreizzen (1984) constructed an indicator of financial fragility f, based on Minsky’s theories. This indicator is defined as the ratio of debt-service payments to self-generated funds where debt service is defined as debt amortization plus interest payments, and self-generated funds are defined as profits plus asset depreciation and interest, minus taxes. Dreizzen’s sample included 143 “normal” enterprises and 23 that had to renegotiate their debts under judicial surveillance. Therefore it is possible to compare the behavior of these two categories of enterprises. His data suggest that the firms that eventually fell under judicial surveillance had a much higher increase in the index of fragility than normal firms. However, it is unclear whether this index provides an explanation of why some firms had difficulties or whether it just states those difficulties—that is, being unable to service debts with their own resources.

Business Failures

The economic developments just summarized caused business failures, which in turn contributed to the failure of banks and other financial institutions. Table 11 presents some data on the liabilities of failed business firms for the Buenos Aires court district. The table shows quarterly total liabilities in nominal terms (first column), deflated by the wholesale price index (second column), and as a ratio to bank credit to the private sector (third column). Inflation makes it difficult to interpret the first column. However, the last two columns give the same broad picture: business failures increased in real terms every year until 1982, peaking in the first quarter of that year, while the ratio of business failures to total private credit shows the same pattern, except for 1980 when the ratio fell.36 Moreover, the highest real rate of increaseof business failures (76 percent) was in 1980 (i.e., when the financial crisis started), although this development is masked in column 3 by the expansive credit policy followed vis-à-vis the private sector. Industry was the sector most seriously hit in the early stages of the crisis: its share in total liabilities of bankrupt firms jumped from 53.7 percent in 1979 to 83.9 percent in 1980.37

Table 11.

Total Liabilities of Bankrupt Firms

article image
Sources: FIEL, Indicadores de Coyuntura, various issues; and International Monetary Fund, International Financial Statistics, various issues.

Comprises data for firms filing for bankruptcy or for judicial surveillance in Buenos Aires courts in each quarter.

III. The Financial Crisis

The Financial System Before the Crisis

Argentina’s financial crisis was the result of the economic developments discussed in the previous section and of developments within the financial sector itself. A key aspect of the latter was the financial liberalization that was implemented in 1977, which drastically changed the functioning of the financial system. The delay in fully adapting the supervisory apparatus to these changes played a major role in deepening the crisis. Once the crisis started, the Central Bank acted swiftly to prevent the collapse of the financial system. However, the measures adopted over time to deal with the crisis had a long-term influence on the structure of the financial system and on the economy as a whole.

The Argentine financial system underwent radical changes in the 1970s. Before 1977, it was heavily regulated. This regulation had been reinforced by the law of nationalization of deposits enacted in 1973 and by the ensuing complementary measures adopted by the Central Bank. The nationalization law had established that commercial banks would receive deposits only for the account and on behalf of the Central Bank; the banks had to keep these deposits in the form of cash-in-vault or deposits with the Central Bank. In other words, banks had a reserve requirement of 100 percent.38,39 Banks could only lend out their own capital and reserves and the funds received via rediscount. The aim of that law was to empower the Central Bank to allocate credit selectively to different sectors and regions. The Bank also set the interest rates on loans and deposits.40 However, the system of nationalized deposits and regulated interest rates soon proved to be too rigid, particularly when the inflation rate, which had been repressed by price controls in 1973, went up significantly in 1974 and jumped dramatically in 1975. In addition, the lack of an explicit link between deposits and loans reduced bank incentives to attract deposits. Therefore the system had to be made more flexible. In 1975, interest rates on certificates of deposit were freed, and automatic rediscount facilities, linked to the growth of specified deposits, were introduced; interest rates on time deposits were also raised significantly. Despite these measures, the yield on nonindexed financial assets was negative in real terms. As a result, over the nationalization period (1973-77), M2 fell by 50 percent in real terms, time deposits by 56 percent, and Ml by 46 percent. This shrinkage of the financial system was made easier by the existence of indexed government bonds, which not only provided a hedge against inflation but also were very liquid.41

Entry into the financial system also was heavily regulated. The Central Bank had to approve the establishment of new banks and the opening and closing of branches of existing banks.42,43 The approval process included an evaluation of the need for new banking services and the prospective bank’s capital adequacy.

The administration that came to power in March 1976 immediately decided that a profound reform of the financial system was needed. The economic program announced on April 2, 1976 stated the need “to give back to the financial and banking system its flexibility and efficiency [by] eliminating the system of deposit nationalization that is inoperative from the point of view of official credit control and that also conspires against the development and agility of financial activity.”44 The reform legislation was enacted in early 1977; its two main pieces were Law 21495, which authorized the Central Bank to convert the financial system back to a system of fractional reserve requirements, and Law 21526, which provided a new legal framework for financial institutions.45

These laws provided the legal basis that allowed the Central Bank to deregulate the Argentine financial system to an unprecedented degree. The Central Bank freed interest rates, which had been totally or partially regulated since 1935. At the same time, in order to facilitate competition by improving market transparency, the Central Bank enjoined banks from charging commissions and special fees on loans, practices that had been widely used to raise the cost of credit when rates were regulated. The Central Bank also abandoned most selective credit practices, returning the responsibility for credit allocation to the commercial banking system; however, the Bank established a special refinance line to facilitate export financing at preferential rates.

Prudential Regulations and Bank Supervision Before the Crisis

Adapting prudential regulations and bank supervision to the new, more liberal, system created by the financial reform was difficult because the many years of heavy regulation provided no guidance on how the Argentine financial system would function in a more liberal environment.

The Central Bank of Argentina has traditionally enjoyed wide powers over the regulation and supervision of banks and other financial institutions. Such powers were preserved by the financial reform, thus allowing the Central Bank to revise existing regulations. The revised regulations covered four areas: capital requirements, asset immobilization, liabilities/ capital ratios, and ratios between the amount of the loan and capital (of the lender and of the borrower).

Minimum capital requirements varied according to the type and location of financial institutions. These requirements were adjusted annually by the wholesale price index.

The regulations on asset immobilization required that immobilized assets could not exceed 100 percent of the capital and reserves of financial institutions.46 For this purpose, immobilized assets were defined as physical assets, prepaid expenses, all kinds of noncurrent and value-impaired loans, and some other assets of lesser importance. In addition, noncurrent and value-impaired loans could not exceed 5 percent of the capital and reserves of financial institutions.

For all financial institutions, the maximum ratio of financial liabilities to net capital and reserves was set at 25. Limits were also set on the maximum financing, including both loans and bank guarantees, that a given client could obtain. These limits were set at 50 percent of the borrower’s capital and reserves for borrowing from any individual financial institution and at 80 percent for total borrowing from the financial system. Financial institutions could exceed these limits when clients offered real assets as collateral; moreover, loans to promote exports, some loans to state suppliers, and seasonal loans were not included in these ceilings. In addition, total financing to an individual client could not exceed 40 percent of the lender’s equity, and a sublimit of 30 percent was set for actual financing of any kind (i.e., excluding guarantees). There were some exceptions for these limits in the case of foreign trade operations.

At the same time, branching regulations were eased. The new law eliminated the requirement that the Central Bank had to approve branch openings of domestic banks. Thereafter, domestic financial institutions simply had to comply with some specific requirements and advise the Central Bank in advance of their intention to open a new branch. But, the Central Bank specified that institutions located outside the major centers would temporarily receive preference for opening branches.47 Local branches of foreign banks remained subject to prior authorization. Moreover, the establishment of new institutions and the transformation of existing ones were facilitated; many formerly nonbank institutions became banks, and thus could broaden the scope of their activities.

These regulations were revised over the years of the reform. Most of the revisions in 1978 and 1979 aimed at increasing the freedom of financial institutions. For instance, in 1978 the Central Bank eliminated the limit on maximum bank lending defined as a proportion of the borrower’s capital, but the limit defined as a proportion of the lender’s capital remained in force.48 In 1979 however, the Central Bank increased the minimum capital requirement for financial institutions, measured in real terms.49

An important change in deposit insurance took place in 1979. Until then, the Central Bank had fully insured depositors and had borne the full cost of the insurance scheme. The scheme that became effective in November 1979 provided limited coverage, and financial institutions were free to join or not. Those that decided to join had to pay a monthly fee equal to 3/10,000 of their average liabilities subject to reserve requirements. The scheme was not a funded insurance system, however, because the coverage provided was independent of the insurance fees collected. Domestic currency deposits of up to 1 million pesos50 were fully insured, whereas those above that amount would be insured only up to 90 percent; the amount that received full insurance would be adjusted monthly with the WPI. Deposits in foreign currencies were not insured.

These measures dramatically changed the perceived riskiness of financial assets, although the perception of risk became widespread only when failures of large institutions began. The public believed that one particular group of institutions provided de facto full insurance, despite the fact that most of them chose not to participate in the Central Bank insurance scheme. This group comprised state banks, usually guaranteed by the corresponding government (national, provincial, or municipal), and branches of foreign banks, which people thought would receive help from their parent banks in case of need. The riskiness of the rest of the system remained to be evaluated by the public, usually on the basis of scant information.

All in all, the prudential regulations in place at the time were fairly comprehensive. Moreover, lack of observance of central bank regulations carried penalties ranging from fines to withdrawal of the charter. However, the Central Bank lacked the supervisory structure to cope with a financial system that was growing fast and was experiencing a dramatic increase in its freedom of action. In addition, the liberalization of the system generated many mergers, transformations of one type of financial institution into another, and an expansion in the number of branches. Many new participants entered the financial system. Moreover, the Central Bank’s supervisory mechanism was biased toward monitoring compliance with regulations rather than analyzing the quality of bank assets—and poor quality assets would become an important cause of failures.

The Central Bank recognized these difficulties and started to study ways to improve its supervision of banks and nonbank financial institutions.51 However, a new comprehensive system of bank supervision became operative only as of January 1, 1981, well after the crisis had begun.52 In the meantime, the Central Bank supervised financial institutions through an analysis of the reports that those institutions had to submit periodically and through on-site inspections. It is hard to evaluate the quality and scope of this supervision, but the annual reports of the Central Bank usually provide some data on the number of inspections carried out over the year. Although the data are fragmentary, they suggest a drop in the percentage of institutions inspected between 1977 and 1981, which began to be reversed only in 1982.

The foregoing discussion suggests that the main supervisory problems were the lack of an appropriate supervisory mechanism and the scant use of the one available, rather than the lack of regulations or supervisory power, which were quite comprehensive. The Central Bank began to work on the first problem soon after the start of the reform, but it began to revamp the supervisory mechanism only in 1981, well after the crisis had begun.

Early Warning Indicators

The question arises as to whether the Central Bank, using available information, could have detected that a crisis was imminent. A way to answer this question is to construct some indicators and test their power in predicting the crisis; this was done in two papers on the Argentine crisis. The first, by Dueñas and Feldman (1980), presents a set of six indicators based on the financial statements (balance sheet and profit and loss statement) that banks have to file with the Central Bank of Argentina: (1) liquid assets/deposits; (2) capital, reserves, and nondistributed profits/risk-bearing assets; (3) problem loans/total loans; (4) risk-bearing assets/income-earning assets; (5) total operating costs/ total operating income; and (6) net income from exchange operations/total net income. The data sample runs from June 1977 to June 1979 and comprises 17 banks, 2 of which were later liquidated by the Central Bank. The study shows that those indicators would have provided early warning about the problems of those 2 banks. Only in January 1981, however, did the Central Bank begin to compute and use such indicators systematically—even though the necessary data had been available for a long time.

The second paper, by Arnaudo and Conejero (1985), also concludes that individual bank failures could have been predicted. Relying only on published sources, the authors considered several indicators to analyze the behavior of the four major banks that collapsed in March-April 1980. Their approach was to compare these banks with one another and with the average behavior of private domestic banks. Several indicators are discussed, including the “unit defensive position,”53 growth of deposits in relation to interest rate premium paid, cost per unit of deposits, share of foreign exchange operations, and branch expansion. Of these, the first three were effective in warning of the impending collapse. These two studies suggest that some indicators could have predicted the onset of the crisis. Other information regularly received by the Central Bank—such as bank reserve positions—could have provided additional warning of individual problem banks.

The studies just described suggest methods for detecting when individual banks depart from the system’s norm for some performance ratios. Although the Argentine crisis initially affected only a few banks, it became evident fairly soon that most of the system had substantial portfolio problems. Therefore, it is useful to analyze whether there were some leading indicators of the condition of the whole banking system.

Table 12.

Share of Problem Bank Loans by Economic Sector, 1974–80

(In percent)

article image
Source: Central Bank of Argentina, Boletín Estadístico, various issues.

Includes all loans currently overdue for more than 10 days, and those that were overdue for more than 10 days in the past and for which special agreements were reached to regularize payments.

Excludes personal loans and loans to unclassified activities.

Table 12 presents data on problem loans, as reported by banks to the Central Bank between 1974 and 1980. In the case of loans to the primary sector and manufacturing, the share of problem loans increased continuously beginning in 1976; the heavy weight of those two sectors in total loans then caused a similar increase in the overall ratio. But the other sectors do not show such a clear pattern until 1980, when every sectoral ratio increased dramatically and the overall ratio jumped by almost 250 percent. These data suggest that until the end of 1979 the deterioration of the portfolio of the banking system as a whole was not particularly alarming, especially given the structural changes that were taking place in the economy. But in 1980 that deterioration became alarming because the economy decelerated (Table 1 suggests that a recession began in February of that year). Moreover, indicators of the firms’ situation, such as earnings, liquidity, and profits, worsened.54 This worsening is also reflected in the jump in enterprise bankruptcies in 1980 (Table 11).

The Crisis

Although many of the problems that led to the financial crisis had been discussed for some time, until 1980 the evolution of the financial system had suffered no major setbacks. The first overt signs of a financial crisis were the failure of one bank, the Banco Intercambio Regional, on March 28, 1980; the intervention of three others by the Central Bank almost immediately afterward;55 and a significant reshuffling of deposits within the financial system. The failed bank had, in a few years, become one of the largest private banks in the country, with a network that had increased from 46 branches in 1977 to 96 in 1979. Its failure caused important losses to its depositors: those that had dollar deposits lost everything56 and those that had peso deposits lost the fraction uncovered by deposit insurance (at least 10 percent) plus the cost of having the insured fraction immobilized until the Central Bank actually paid off insured deposits.57 This failure made the public acutely aware of the fact that bank deposits had some risk, which varied widely from one institution to another. People withdrew funds from institutions whose solvency they questioned and deposited the funds with institutions the depositor considered solvent. In addition, operators in the interbank money market became more cautious because participants would have little chance of recovering loans made to a failed financial institution. Among the institutions whose solvency was questionable were those that had expanded enormously in a short time, largely because they paid high interest rates on deposits—in other words, institutions whose performance resembled that of the Banco de Intercambio Regional. The institutions that were considered solvent were those backed by the Government at any level (national, state, or municipal) or by a foreign parent institution, or, to a lesser extent, those domestic private banks that had a well-established reputation. Deposits could be reshuffled rapidly owing to their short maturities—time deposits were concentrated in maturities no longer than 30 days. This reshuffling and the drying up of interbank funds helped to propagate the crisis: The Central Bank had to intervene three banks within a month of the Banco de Intercambio Regional’s failure. Although the economic situation of these banks might have required some central bank action, the run on their deposits made urgent intervention the only policy option open to the Bank.

The Authorities’ Reaction to the Crisis

The authorities had to act on several fronts to cope with the crisis. First, they had to take emergency measures to avoid a bank panic; second, they had to search for longer-term solutions to the private debt problem and its effects on financial institutions; and third, they had to find ways to restructure the financial system.

Emergency Measures. When the Banco de Intercambio Regional was closed, the Central Bank had to address the solvency problem of those institutions that faced massive deposit withdrawals. Its first measure was to create a new credit facility to aid financial institutions whose deposits were falling. This step was taken on April 3, 1980, when it became evident that existing lender-of-last-resort facilities were inadequate to cope with the size of the withdrawals. The second measure was to increase by a multiple of 100 the maximum size of fully insured deposits retroactively to November 18, 1979, which was the date when the reduction in deposit insurance had come into effect.58 The third measure was the authorities’ intervention, on April 28, 1980, of the three banks that had suffered the biggest drain in deposits (Banco Internacional, Banco Oddone, and Banco de los Andes).

These measures gradually succeeded in stabilizing the situation. Aggregate deposits fell in real terms immediately after the start of the crisis, but by August they had already exceeded the March levels. However, the distribution of deposits among financial institutions changed in favor of state and foreign institutions, a pattern that persisted over time (Table 13).

Solutions to the Private Debt Problem. Once the immediate danger of a panic had been averted, the authorities turned their attention to the problem of private debt.59 The first actions attempted to provide relief by encouraging the lengthening of loan maturities, most of which had been very short (i.e., below one year). In November 1980, the Central Bank announced that it would be prepared to make advances to financial institutions at a much longer term (one year) than prevailing deposit maturities. By reducing the maturity risk of financial institutions, these loans would encourage those institutions to extend the maturities of their own loans. However, this measure was in operation only in November and December 1980. A second effort was made in April 1981 when a new scheme of central bank advances to financial institutions was announced. Financial institutions were allocated funds, first by auction and later by direct allocation, up to 12 percent of their deposits, at market-related rates. These funds had to be used to refinance existing business debt—mainly debts of the agricultural, manufacturing, and construction sectors. This scheme was abandoned in November 1981, when a special law established a more ambitious refinance program with the following characteristics:

  • (a) Banks and other financial intermediaries should refinance 50 percent of the liabilities of the manufacturing sector and 40 percent of the debt of other industrial sectors (excluding personal and mortgage debts) outstanding at the end of August 1981. The refinanced portion of the debt would be payable over seven years, with a three-year grace period. Borrowers would pay a yearly interest of 3 percent, and the principal would be fully indexed to the Financial Adjustment Index (Indice de Ajuste Financiero—IAF), which was computed by annualizing the interest rate paid by a sample of banks on 30-day deposits (tasa testigo).

  • (b)Banks and other financial intermediaries were authorized to rediscount with the Central Bank the full amount of the refinanced debt. This rediscount carried a zero rate of interest and was fully indexed to the IAF.

  • (c)Along with the rediscount, the banks had to lodge with the Central Bank a government bond for an amount equivalent to the rediscount. This bond had a maturity of seven years with a three-year grace period, was not transferable or negotiable (except between banks), and paid a yearly interest rate of 6 percent (tax free), with full indexation of the principal to the IAF.

  • (d)Private financial institutions were free to participate or not in the scheme. Participating institutions had to provide refinancing to all client firms requesting debt consolidation. Participation was compulsory for all official banks.

  • (e)Participating private institutions had the option to contribute 1.5 percent of the amount refinanced to a guarantee fund administered by the Central Bank. The guarantee fund applied to 75 percent of the refinanced debt, and provided banks with protection against bankruptcy or default by firms receiving refinance. If a firm defaulted, the guaranteed portion of the principal would be frozen, interest free (although it would continue to be indexed to the IAF), for the remainder of the seven-year period, at the end of which the guarantee would be made effective. The rediscount of that transaction and the equivalent bond amount would be canceled.

  • (f)Firms benefiting from the debt consolidation would have to increase their capital by 10 percent of the refinanced amount (IAF-adjusted) over a two-year period. Any distribution of dividends, moreover, would have to be matched by an equivalent cancellation of the (indexed) refinanced debt at the moment the distribution took effect. By March 1, 1982, this refinancing program had provided funds for 4,709 billion pesos—equivalent to about 2 percent of total loans to the private sector.

Table 13.

Distribution of Deposits Among Groups of Institutions at Selected Dates, 1980–84

(In percent)

article image
Source: Central Bank of Argentina, Boletin Estadistico, various issues.

The mechanisms just outlined were, in essence, refinancing schemes with an element of subsidy, but they still preserved the principle of free rates. Near the end of 1981 the authorities diluted this principle by setting a so-called reference rate and establishing that any interest differential over that rate had to be deposited with the Central Bank. This measure did not have much effect because the reference rate was rapidly phased out when the administration changed soon thereafter. This administration—in office between December 1981 and July 1982—restored free interest rates and tried to follow a free-market approach to economic management. However, the conflict over the Falklands/ Malvinas caused the authorities to intervene more actively in the economy—though not in the financial market—and eventually led to the downfall of the administration.

The administration that came to power in July 1982 believed that “there is a manifest disproportion between the magnitude of enterprise and household liabilities, both in pesos and dollars, and also the liabilities of the public sector itself, in relation to the value of real assets, especially productive assets (farm land, urban industrial, and commercial real estate, machinery, etc.).”60 To solve this problem, the authorities decided to introduce a drastic reform in the financial sector, the major aim of which was to generate negative real interest rates for a limited time in order to erode the value of existing bank loans and deposits.61 At the same time, new financial assets were created in order to give the public some protection against future inflation, and thus encourage the financial savings needed for new lending. In broad terms, the strategy called for a transitional period during which the financial system would have three main segments: a regulated segment, whose relative importance would fall over time as negative real rates shrank the value of its assets and liabilities until the segment disappeared altogether, and two other segments, one with free rates and the other indexed, whose combined relative importance would grow over time.

Although the financial system that emerged from these reforms was supposed to evolve back to a free system, later adjustments moved it in the opposite direction: the regulated segment grew in size while the free segment languished. Finally, in August 1983 the Central Bank prohibited the acceptance of any new deposits for the free segment, which was phased out as outstanding deposits matured.

The effects of the reform on the real value of the outstanding debt and on the size of the financial system were dramatic: negative real regulated rates eroded the outstanding real value of bank loans and reduced the demand for financial assets. Table 10 shows that the real value of debt deflated by the WPI had fallen by almost 37 percent between the second and fourth quarters of 1982 and by 62 percent by the end of 1984; all the other deflators in that table give a similar picture. Moreover, the size of the banking system shrank significantly below its prereform size, as measured by the real value of loans and deposits (Table 14). Table 14 shows that the real claims of the system on the private sector outstanding by the end of 1982 represented only 85 percent of those outstanding in June 1982, a proportion that fell to 70 percent by the end of 1983 and to 63 percent by the end of 1984. For total claims (i.e., including claims on the Central Government) the corresponding figures are 65 percent for the end of 1983 and 60 percent for the end of 1984. Moreover, the proportion of total claims financed by private sector peso deposits declined from 39 percent in June 1982 to 26 percent three months later, and then began to increase—but without reaching the levels that existed during most of the financial liberalization experience.

Table 15 presents quarterly data on deseasonalized monetary aggregates in real terms, for the period beginning in the first quarter of 1977 through the fourth quarter of 1984. Two major conclusions can be drawn from that table. First, although the crisis caused monetary aggregates to fall in the second quarter of 1980, there was a recovery in the third quarter, when M2 reached its peak for the whole period included in Table 15. Also, the fall in the demand for monetary aggregates became increasingly pronounced between the fourth quarter of 1980 and the second quarter of 1981. These two facts suggest that uncertainty about economic policies (related to an impending change in administration) was more important than the earlier bank failures in undermining people’s willingness to hold peso-denominated assets. Second, only after the reintroduction of interest rate controls in the third quarter of 1982 did quasi-money and M2 fall below the levels they had reached immediately before the liberalization.

Table 14.

Evolution of the Banking System

(In thousands of australes)

article image
Source: International Monetary Fund, International Financial Statistics.

Deflated by the WPI.

Table 15.

Monetary Aggregates in Real Terms, Deflated by the CPI, 1977–84

(In thousands of australes)

article image
Source: Central Bank of Argentina, Boletín Estadístico, various issues.

Quarterly average of end-of-month balances.

M1 = Currency in circulation plus demand deposits of the private sector.

Quasi-money: Savings and time deposits and “other deposits” of the private sector (accrued interest is a major component of “other deposits”).

M2 = M1 plus quasi-money.

As the system shrank, banks had to rely less on the public for their funding and more on the central bank rediscount, which became a major source of funding.62 Also, as the importance of bank lending fell, the market sought alternative financial sources. Some, like bank acceptances, went through official channels; others, like inter-enterprise loans, did not. Lack of information on these alternative sources prevents us from analyzing them in this paper. As a general proposition, however, a financial market with regulated and unregulated segments entails a loss of economic welfare and macroeconomic control, compared with a free, integrated market. In particular, financial savings will tend to be lower and the cost of capital higher, on average, because of the market imperfections created by the segmentation.

Besides relieving the burden of domestic currency debts, successive administrations also took measures to relieve the burden of foreign currency debts. One of the first measures was to compensate some borrowers for the increase in the domestic currency value of their debts that resulted from the devaluation of June 1981. This compensation covered debts incurred or renewed between January 1 and May 29, 1981 that matured until the end of that year, provided that the debts were rolled over for at least one year. In addition, the Central Bank established an exchange insurance scheme that subsidized borrowers insofar as the premium charged was below the actual rate of devaluation.63 Over time, the scheme was enlarged and made even more attractive as Argentina’s reserve position deteriorated and the authorities tried to induce borrowers to renew their foreign credits.

Measures to Restructure the Financial System. The Central Bank acted to facilitate the restructuring of the financial system not only by liquidating failed financial institutions but also by encouraging mergers and sales and by increasing the Bank’s own flexibility to deal with problem institutions.

In October 1980 the Central Bank established a special line of credit to finance a fraction of the cost of mergers and purchases of financial institutions. The financed fraction could vary in each case and the interest rate on this line of credit was below market rates.64 Later, the Central Bank created a formal service to help financial institutions meet partners for mergers or purchases.

In 1982, a new law broadened the powers of the Central Bank to deal with problem financial institutions. The purpose of this law was to give more flexibility to the Central Bank, which, under the old law, had very few options besides liquidating insolvent financial institutions. Liquidation has always been the last-resort remedy provided by the Argentine banking legislation to deal with institutions in serious difficulties; in addition, liquidation has been the most severe penalty that could be inflicted on a financial institution. When the Central Bank approves the liquidation of a financial institution, a liquidator—usually from the Central Bank staff—is appointed to replace the chief executive officer and the board of directors of the institution. The liquidator’s first task is to pay the insured part of maturing deposits with funds advanced by the Central Bank.65 The liquidator does not make new loans but is responsible for collecting those outstanding as they mature. The liquidator also has broad powers in administering the institution on a day-to-day basis, subject to the approval of senior central bank management. The liquidator’s task is completed once court approval has been obtained for the final disposition of the institution’s assets and liabilities.

The liquidation process is quite costly. First, announcement of the liquidation of a large financial institution can create a confidence crisis. Second, both depositors and borrowers have to transfer their business to another institution. Third, funds advanced by the Central Bank for the liquidation are a source of monetary expansion that may be difficult to sterilize. Fourth, the physical and human resources that were employed by the liquidated institution may lay idle for a relatively long time. Fifth, borrowers tend to give relatively low priority to repaying loans made by a liquidated bank, because they prefer to honor their commitments to operating banks, which can provide new credits; this attitude introduces a rigidity in credit distribution in favor of the clients of the liquidated institution. Sixth, the cost escalates if the liquidated institution has acquired assets from bankrupt borrowers that must be administered by the liquidator. Few liquidators have expertise in administering assets that belonged to the bankrupt borrower (e.g., a steel mill), but many liquidators have had to deal with such a situation.

Because liquidation proved to be a rather costly solution to the problems of financial institutions, the new law was intended to reduce the need for liquidation by allowing different solutions to be applied at the discretion of the Central Bank.66 For instance, the Central Bank could authorize one sound financial institution to administer a troubled institution for a given period of time, at the end of which the first institution had the option of purchasing the second.67 In addition, the Central Bank was granted explicit powers to intervene troubled financial institutions; the Bank had exercised such powers since the crisis started but their legality had been challenged. The new law also expanded the Central Bank’s choices in liquidation cases: for example, the Bank could decide to liquidate an institution without revoking the institution’s charter.

The financial crisis and the actions of the Central Bank added impetus to the restructuring of the financial system that was already under way when the crisis began. The number of nonbank financial institutions fell sharply beginning in 1979, largely as a result of mergers, transformation into banks, or absorption by banks; also, 97 institutions were liquidated between May 1977 and December 1983. The number of banks increased almost continuously until 1981, but this increase was more than offset by the decrease in the number