1. The Uruguayan economy before 1974

For more than 20 years after the early 1950s, policymakers resorted to import substitution, and wide-ranging controls on domestic goods and financial markets, reflected in high tariffs, other trade barriers, negative interest rates in real terms, and extensive price distortions. Those years were characterized by very low and decelerating real GDP growth, high, volatile, and accelerating inflation rates, a low level of domestic savings, recurrent balance of payments crisis, increasing fiscal deficits, and high unemployment.

The rate of growth of real GDP decelerated from 5.4 percent per annum between 1944 and 1951 to 3.0 percent between 1951 and 1956, 0.1 percent between 1956 and 1967, increasing to 1.6 percent between 1967 and 1974. ^{2/ 3/}

The inflation rate increased continuously, averaging 49.8 percent per year between 1965 and 1970, 62.7 percent between 1971 and 1973, reaching 97 percent in 1973. The unemployment rate exceeded 8 percent between 1965 and 1973.

The balance of payments experienced recurrent difficulties owing to rising fiscal deficits, inadequate credit and exchange rate policies, and some external factors. The fiscal deficit, as a proportion of GDP, increased from an average of 1.9 percent during the 1965–70 period to 3.2 percent over 1971–73. ^4/ The inadequate credit and exchange rate policies, and the persistence of negative real interest rates provided incentives for capital flight leading to a contraction in the size of the regulated financial market over time. During most of these years, multiple exchange rates and restrictive exchange controls prevailed. In March 1972 the authorities adopted a crawling peg, while maintaining the exchange controls. Two main external factors compounded the balance of payment difficulties in 1974: the effects of the first oil shock, and the closure of EEC markets to beef imports, the main Uruguayan export.

In July 1974, a new economic team responded to the weak economic performance and growing imbalances by implementing a package of deregulation policies.

A brief description of the reform process follows, with more attention being paid to the reforms in the financial sector in Section III. ^5/

2. The 1974 economic reform: phase I. 1974–78 ^6/

The economic reform, implemented in Uruguay in 1974, included measures to liberalize international trade and capital transactions, to remove distortions in domestic markets, and to eliminate fiscal imbalances. The liberalization of capital transactions with the rest of the world proceeded at the fastest pace and occurred before other major policy changes. While domestic financial markets were also liberalized relatively fast, the rest of the liberalization program proceeded at a slower pace and suffered transitory reversals. By the middle of 1977, capital transactions were virtually free, while remaining exchange restrictions on current transactions were relatively minor. In contrast, trade restrictions remained severe and complex, with imports subject to a multiplicity of charges and exports subject to an array of taxes and subsidies which severely distorted production patterns.

In September 1974, capital flows were liberalized with domestic residents permitted to hold dollar accounts with domestic banks for the first time; moreover, exchange rate controls were eliminated, and Uruguayans were also free to buy and sell assets denominated in external currencies without restrictions. Capital transfers were freely permitted through the financial market. A liberal foreign investment law was also approved in 1974, which required government authorization only for investment in specified activities. Foreign investors were guaranteed the transfer of capital and profits through the financial market. ^7/ During this phase of the reform, the exchange rate for financial transactions was freed, while the exchange rate for transactions in the goods market followed a passive crawling peg.

Although tariffs remained high and a number of restrictions on current payments or multiple currency practices prevailed, trade transactions were partially liberalized during this phase. From the middle of 1974 to early 1975, all quantitative restrictions on imports were removed. Imports of capital goods were further liberalized in early 1975 when they were exempted from special deposit requirements. On the exports side, taxes on wool and beef exports were considerably reduced during 1974.

Further significant changes occurred in the trade and exchange system between 1975 and 1978. In 1975, import deposits, and exchange taxes on beef exports were eliminated and many trade taxes were reduced. In 1977, the system of import surcharges was further streamlined leading to more uniform rates of effective protection. However, at the same time, a tax (retención) on beef exports was reintroduced, and export subsidies (reintegros) on nontraditional exports were reduced. In January 1978 a new tariff structure replaced the complex system of import duties and related charges with a basic rate of 20 percent and multiple rates ranging from zero percent to 150 percent. Export taxes continued to increase during 1978 while subsidies on nontraditional exports continued to be reduced.

During this phase, the authorities began to remove the comprehensive price controls in the economy, but at a slow pace. During 1974, controls were progressively eliminated or released on selected products and services, most of which were not included in the consumer price index. After a temporary reintroduction of price controls in 1975, all prices of goods considered to be competitively produced and not included in the CPI were freed in early 1976. Although price liberalization continued, by the end of 1978 46 percent of products in the consumer price index (CPI) basket were still subject to price control.

Tax reforms were also implemented to reduce tax evasion and simplify the tax system. Late in 1974, the tax on earned income, and the inheritance tax were abolished while the basic rate on the value-added tax was increased from 14 percent to 18 percent. During 1975, the coverage of the value-added tax was widened and its basic rate was increased to 20 percent, while the tax rate on income from industry and commerce was raised from 20 percent to 25 percent. During 1977 and 1978, the authorities strengthened tax administration and introduced severe penalties for tax evasion together with indexation of late tax payments.

The stabilization strategy included an anti-inflationary policy based on reductions in fiscal deficits and in monetary growth. The fiscal position of the Central Government improved substantially owing to a strong rise in revenue collection and expenditure restraint. As a result, the overall deficit of the Central Government fell from the equivalent of 4.5 percent of GDP in 1975 to 1.3 percent of GDP in 1978 (Statistical Appendix Table 1). However, monetary aggregates continued to grow rapidly (Chart 1). The stabilization plan also included an expenditure-switching policy, consisting mainly of a real devaluation followed by a passive crawling peg, and the promotion of nontraditional exports.

View Full Size

URUGUAY: MONEY AGGREGATES GROWTH RATES

(In percent, per annum)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 2.

• Download Figure
• Download figure as PowerPoint slide

Economic performance clearly improved, despite an unfavorable evolution of the terms of trade. A period of rapid growth (Chart 2) replaced the pre-reform stagnation: real GDP grew at an average rate of 3.9 percent between 1974 and 1978. Inflation (Chart 3) declined from over 107 percent per annum in the last quarter of 1974 (CPI) to 41 percent by the end of the third quarter of 1978. The balance of payments strengthened substantially, with overall surpluses of US$167 million in 1976, US$102 million in 1977 and US$286 million in 1978 (Chart 4). This was due partly to large inflows of foreign private capital, attracted by the removal of exchange controls, the introduction of foreign currency deposits in local banks, the uncertain political situation in Argentina, and the high interest rates in Uruguay. The current account also improved, reflecting mainly the strong growth of nontraditional exports. Both unemployment and real wages fell.

View Full Size

URUGUAY: REAL GDP GROWTH RATES

(In percent, per annum)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: IMF, International Financial Statistics.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

URUGUAY: PRICE AND WAGE INFLATION

(In percent, per annum)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 4.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

URUGUAY: BALANCE OF PAYMENTS

(In millions of U.S. dollars)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 5.

• Download Figure
• Download figure as PowerPoint slide

3. Phase II: 1979–82

The second phase of the economic reform had three key elements: (1) the implementation of an active crawling peg, (2) the starting of a process of unification of import tariffs, and (3) an important tax reform. In addition, deregulation of foreign trade, commodity price and financial markets continued.

The strategy underlying this phase reflected disappointment with the stabilization measures taken in Phase I, which had failed to lower inflation to an acceptable level. To deal with this problem, the new strategy relied on using the exchange rate to influence inflationary expectations. On October 17, 1978, the dual exchange market was de facto unified when the Central Bank announced that it would buy and sell any amount of foreign exchange at rates that would be adjusted from time to time. On October 26, 1978, the Central Bank indicated for the first time the future rate of devaluation through the sale of three-month treasury bills redeemable in domestic currency or U.S. dollars; the redemption value (principal and interest) in U.S. dollars was indicated on the bills. On December 28, 1978, the Central Bank published a preannounced schedule of devaluations (informally known as the tablita). ^8/ This schedule indicated the daily exchange rates for the Uruguayan peso (in terms of U.S. dollars) through March 1979. On March 27, 1979 the schedule was extended through the end of 1979, and on July 10 through the first quarter of 1980. Thereafter, the schedules were announced six to nine months in advance.

Corbo, de Melo and Tybout (1986) have summarized the rationalization of this plan as provided by the Rodríguez (1982) model:

• Interest rate parity obtains continuously because of the absence of controls on capital flows and the assumption of perfect asset substitutability. The law of one price holds for tradeables, and the rate of change in the prices of nontradeable goods is a function of inflationary expectations—which are assumed to form adaptively—and of excess demand for nontradeable goods. The model predicts that the implementation of the “tablita” should immediately reduce nominal interest rates and, to a lesser extent, inflation. The decline in real interest rates should first stimulate demand, creating an excess demand for nontradeable goods and thereby inducing a temporary appreciation of the real exchange rate. As inflation falls, both the real interest rate and the real exchange rate should increase, approaching their long-run equilibrium from below. The economy should stabilize without undergoing the recession associated with traditional contractionary measures.” ^9/

Tariff reductions were accelerated during this phase. Import duties on a large number of imports were reduced or eliminated during the second half of 1978 and in 1979. Moreover, in December 1978, the Government announced a timetable to unify the numerous import taxes to a global tariff, and to reduce this global tariff gradually to an average level of 35 percent over the period January 1, 1980-January 1, 1985. ^10/

Export subsidies were also reduced. On March 29, 1979 the system of export prefinancing (preanticipos) was eliminated. However, export subsidies for many products continued to be granted in the form of tax credit certificates.

In November 1979, a major tax reform went into effect. This reform included new taxes on agricultural income and sales, the merger of several excise taxes, the rationalization of a number of other taxes, and the substitution of higher value-added tax rates for several social security taxes. This tax reform was mainly oriented to increase the international competitiveness of Uruguayan products by lowering labor costs and eliminating taxes on exports of agricultural products.

Price deregulation continued during this phase and the role of the price control agency gradually shifted from controlling prices to monitoring price movements and market behavior to protect consumer interests. Nevertheless, by March 1981, the prices of 24 percent of goods and services comprising the consumer price index remained subject to control.

The foregoing measures succeeded for a time. The economic performance was impressive until the beginning of 1981. GDP growth exceeded 6 percent over the period 1979–80 (Chart 2). Moreover, after accelerating during 1979 the inflation rate receded since early 1980 until the third quarter of 1982 (Chart 3). The balance of payments was consistently in surplus (Chart 4), exports were greatly diversified, and official international reserves rose to a comfortable level. Balance of payments surpluses resulted mainly from capital inflows and tourism from Argentina. The peso appreciated in real terms with respect to the currencies of all major trading partners except Argentina (Statistical Appendix Table 6). Moreover, capital inflows fueled two booms, one in agricultural land, the other in real estate. ^11/ Confidence in the peso strengthened and substantial amounts of private savings flowed into the domestic banking system.

Certain policy inconsistencies and several external factors resulted in the steady deterioration of Uruguay’s economic and financial performance beginning in late 1980. Demand for Uruguayan exports weakened as a consequence of the recession in the world economy, and the policy adjustments undertaken by Argentina and Brazil. Moreover, the deterioration of the terms of trade and the rising world interest rates adversely affected the current account of the balance of payments. The public sector financial position also deteriorated as a result of a decline in the tax base, a large increase in social security payments ^12/ and a surge in the burden of servicing the public foreign debt. ^13/ Finally, the Argentine devaluations of March 1981 fueled the growing expectations that the Uruguayan tablita was becoming unsustainable.

These factors led to a crisis in confidence in the tablita which resulted in major capital outflows and capital flight—which were evident in large changes in errors and omissions items of the balance of payments (Chart 5)—and a deep recession. Real GDP growth decelerated to about 2 percent in 1981, and in 1982 GDP declined almost 10 percent (Chart 2). Finally, in November 1982, the government dropped its commitment to maintain the tablita and a major devaluation followed (Chart 6).

View Full Size

URUGUAY: CAPITAL ACCOUNT BALANCE AND ERRORS AND OMISSIONS IN THE BALANCE OF PAYMENT

(In millions of U.S. dollars)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 5.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

URUGUAY: NOMINAL EXCHANGE RATE

(Pesos per U.S. dollar, end of period)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: IMF, International Financial Statistics.

• Download Figure
• Download figure as PowerPoint slide

1. Liberalization of international financial transactions

As noted in Section II.2 above, in September 1974 domestic residents were allowed to trade in any kind of assets denominated in any currency, since exchange controls were eliminated altogether, and banks were authorized to receive deposits denominated in foreign currency. Moreover, the abolition of the personal income tax in July 1974 eliminated the need to report these transactions. Also, nonfinancial firms were allowed to borrow in foreign currency at home or abroad with no limitation, and nonresidents were allowed to repatriate both earnings and capital without any limit.

• As Larraín notes,

• “this deregulation is all the more remarkable since before September of 1974 it was illegal for Uruguayans to hold dollars” and “stories are told about people sent to jail after being caught holding a hundred-dollar bill”. ^19/

2. Interest rates

The economic reform that began in 1974 included various adjustments to the ceilings on interest rates, culminating eventually in full liberalization of interest rates.

Effective September 25, 1974, the maximum interest rate (payable in advance) on bank peso loans was raised from 24 percent to 32 percent per annum. At the same time, the ceiling on bank commissions (also payable in advance) was reduced from 10 percent to 6 percent, while the tax on financial transactions (which could be passed on to borrowers) was kept at 6 percent. These measures increased the effective cost to borrowers (including the said tax) from a maximum of 66.6 percent to 78.6 percent per annum. The annual interest rate for loans in foreign currency had already been increased in July 1974 from 14 percent to 16.5 percent. The BROU kept preferential interest rates for public sector loans, and for a few priority sectors.

Interest rates on peso deposits continued to be fixed by the Central Bank until March 1976. Effective October 1, 1974, an annual interest rate of 10 percent was set on demand deposits which previously were non-remunerated. However, this measure was reversed a few months after. At the same time, the interest rate for savings deposits was increased by 10 percentage points to 18 percent per year. For time deposits the annual rates were increased by 15 to 23 percentage points ranging from 30 percent (for 3- to 5-month deposits) to 48 percent (for deposits of 12 months and over). Rates payable on deposits in dollars were free throughout the period.

During 1976, the authorities attempted to reduce interest margins, improve the interest rate structure and encourage efficiency in banking operations. From January 1, commissions were eliminated and all interest rates were made payable at maturity, whereas before they were payable in advance. From April 1, all rates of interest on deposits were freed, but maximum lending rates continued to be set by the Central Bank. These changes enabled banks to pay positive real interest rates on deposits for the first time in many years. Nevertheless, the BROU continued to charge promotional rates on loans for nontraditional exports and for meat-packing; these loans were rediscounted with the Central Bank. However, from April 1, 1976, the annual rate for all new central bank rediscount facilities, which had ranged from 8 percent (for rediscounts to the BROU) to 12 percent (for regular facilities to commercial banks), was raised to 17 percent. In September 1976, rediscount was restricted only for emergency lending to banks facing temporary liquidity problems and carried a penal interest rate.

The ceilings on lending rates were increased again in late 1977, and 1978. In addition, in May 1979, the tax on bank loans was substantially reduced. Finally, on September 12, 1979, all interest rate ceilings were eliminated.

3. Reserve requirements

The level and structure of reserve requirements varied throughout the period, and the requirements were generally reduced, simplified and unified between 1975 and 1979. This reflected a changing balance between the needs of raising revenue for the Government, inducing a certain portfolio composition, and promoting the liberalization program. Both peso and foreign currency deposits were subject to reserve requirements, whose level depended on the maturity of deposits, and which could be partially satisfied with government bonds during certain periods.

For a chronology of adjustments in the reserve requirements, see Appendix I.

To assist banks that were placed in a difficult financial position because of the new interest rates announced in September 1974 (which implied a smaller spread for the commercial banks), effective October 1, 1974, the Central Bank reduced the reserve requirements. Early in 1975, the monetary authorities established a 40 percent marginal reserve requirement on the excess of peso deposits over their average level in December 1974.

A major reform in regulations on reserve requirements took place at the beginning of 1976 when government debt become eligible to meet required reserves regulations. An important change was to disallow the use of any assets except cash in vault or deposits with the central Bank for the fulfillment of reserve requirements. At the same time, however, it was decided that private commercial banks would have to keep in their portfolios an amount of government securities equivalent, as a minimum, to a certain proportion of their deposits. In June 1977 the requirement to holdd government debt was eased, while reserve requirements were increased considerably.

Late in 1977 the authorities began to equalize the treatment of reserve requirements on local and foreign currency deposits. From January 1, 1979, all reserve requirements, both on local and foreign currency deposits, were unified at 20 percent of outstanding deposits at the end of 1978. All increments of deposits above that amount would have a zero reserve requirement. Finally, in May 1979, all basic reserve requirements were eliminated. After this measure, the only reserve requirement which remained in force was a marginal requirement of 100 percent which applied to the amount of deposits in local currency exceeding 16 times a bank’s capital and reserves. In June 1979, this limit was temporarily increased to 30 times capital and reserves, but was extended to foreign currency deposits, lines of credit, guarantees, and other obligations.

4. Entry barriers

Under the 1965 law, no new banks were allowed, and a ceiling was set on the total number of branches. This law did not ban the establishment of new banking houses, but since they could not raise deposits, this was not a very important exception.

The first major change in entry regulations took place in 1976 when the Central Bank of Uruguay allowed banking houses to receive foreign currency deposits from nonresidents. This was a main factor in the rapid expansion of these houses after 1976, whose number increased from 2 in 1976 to 23 in 1981.

In November 1981 (Law 15207), the entry prohibition in the banking sector was lifted. New banks were allowed, in a number not to exceed each year 10 per cent of the number existing in the previous year; the ban on branch expansion was also lifted. ^20/

5. Prudential regulations

Uruguay’s financial system was subject to a limited array of prudential regulations, most of which were in force for the whole reform period. No formal deposit insurance scheme existed at that time; banking supervision consisted primarily of control of capital and reserve requirements and on-site inspections; the accounting framework and off-site analysis were weak.

Maximum liability to capital ratios limited banks’ expansion. Liabilities exceeding the maximum ratio were subject to a 100 percent reserve requirement. In February 1974 it was established that domestic deposits should not exceed an amount equal to 15 times each bank’s capital. That ratio was raised to 16 some months later, but only for local currency deposits. The maximum limit for bankers’ acceptances was set at 50 percent of the bank capital. In 1979, the liabilities to capital ratio was set at 30 times for banks and 20 times for banking houses, and was extended to cover both domestic and foreign currency deposits.

Minimum capital requirements were in force, varying according to the type of institution (higher for banks, lower for banking houses) and the location (higher for offices in Montevideo, departmental capitals and the cities of Las Piedras, Pando, and Punta del Este). These requirements were updated periodically, and were in force all through the period.

Maximum financing limits existed until 1979. When the reforms started in 1974, a bank’s credit to a single client could not exceed 20 percent of the bank’s capital and reserves (except for export prefinancing). In 1975, this limit was raised to 25 percent. A 1977 regulation forbade financing above 25 percent of the bank capital to corporations whose boards were linked to the banks’ boards or managements.

In March 1979, all regulations on maximum financing and risk concentration were eliminated. They were reinstated in December 1980, when they were set at 25 percent for individuals or single firms, and at 35 percent for conglomerates or groups of firms. They applied to disbursed and non-disbursed loans, endorsements, etc. In August 1981, these limits were abolished and were replaced by a requirement that each borrower be subject to a detailed analysis, using a uniform methodology. In September 1982, lending to managerial personnel of firms was prohibited.

Foreign exchange exposure was not regulated. Thus, banks could borrow in dollars and lend in pesos without any limits.

Asset immobilization limits were set at 60 percent of capital and reserves in 1976, and raised during 1977 first to 70 percent and then to 100 percent.

6. Banking supervision and early warning indicators

The Central Bank Annual Reports provide information on the number of on-site bank supervision visits each year. The information for 1975–83 is as follows:

Year	Number of Visits
1975	36
1976	40
1977	97
1978	82
1979	66
1980	74
1981	…
1982	99
1983	96

View Table

Year	Number of Visits
1975	36
1976	40
1977	97
1978	82
1979	66
1980	74
1981	…
1982	99
1983	96

These data suggest that the Central Bank carried out a number of inspections per year that exceeded the number of banks and banking houses in operation. However, these numbers should be interpreted with caution. As such, they do not provide any information on the quality of the inspection or the criteria used for the audit, the amount of information made available to the auditors, or the penalties that could be applied.

• According to Larraín (1986), supervision

• “has been mainly confined to the traditional control of capital and reserve requirements. Balance sheets and profit and loss statements are also reviewed to judge if they are appropriately drawn up so that they accurately reflect the financial situation of the bank”. ^21/

An interesting comment on this issue is made by J. Gil Díaz, President of the Central Bank of Uruguay from 1974 to 1982:

• During my tenure…there was no legal norm that allowed the Central Bank to regulate the way in which the accounting information of financial institutions should be presented and evaluated.” ^22/

Similarly, there does not seem to have been any early warning indicators, or other systematic or informal off-site analysis of banks in the Central Bank. ^23/ Nevertheless, the Central Bank management might have been aware of banking difficulties well in time, as Gil Díaz states:

• …the Central Bank of Uruguay, within the limits of its capacity, discharged its responsibility of controlling the national banking system. Through its supervision, auditing, analysis of periodical information and special information requests, it soon detected that a group of institutions was weaker than the rest. ^24/

1. Financial intermediation and the growth of the financial sector

Charts 7 and 8 illustrate the financial deepening process in Uruguay after 1974 by depicting the ratios of bank assets and liabilities relative to GDP.

View Full Size

URUGUAY: FINANCIAL INTERMEDIATION

(Liabilities of the banking system as a percent of GDP)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 9.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

URUGUAY: CREDIT OF THE BANKING SYSTEM

(As percent of GDP)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 9.

• Download Figure
• Download figure as PowerPoint slide

The interest rate reforms and capital account liberalization generated important portfolio shifts and capital inflows with significant consequences for the Uruguayan financial sector since 1974. The ratio of Ml to GDP declined from about 12 percent in 1974 to less than 8 percent in 1981, while the ratio of M3 (M1 plus savings and time deposits in foreign and domestic currency) to GDP increased from about 19 percent to about 43 percent over the same period.

The sequencing of the economic reform was also reflected in the evolution of the composition of M3. During the first years after the economic reform began to be implemented liberalization of foreign currency deposits in Uruguay was quickly completed while important regulations continued to prevail for domestic currency transactions. During this period, foreign currency deposits increased much faster than peso deposits. The ratio of foreign deposits to GDP increased about 6 times from 1974 to 1977 (with a large proportion held by nonresidents), while the ratio of domestic currency time and savings deposits to GDP increased only 33 percent over the same period.

After 1977, when interest rate ceilings were gradually removed and the authorities began to equalize the treatment of local and foreign currency deposits with regard to reserve requirements, domestic currency deposits became more attractive to investors. The ratio of savings and time peso deposits to GDP increased 2.5 times from 1977 to 1980 while the ratio of foreign currency deposits to GDP declined from 1978 to 1980. The factors that adversely affected the Uruguayan economy and contributed to the loss of confidence in the tablita reversed these trends after 1980. ^25/

The question arises whether this growth of the financial sector translated into an increasing availability of credit to the private sector or whether funds were invested abroad or used to finance fiscal deficits.

Chart 8 shows a fast growth in credit to the private sector between 1974 and 1982. The remarkable reduction in fiscal deficits reduced government borrowing needs between 1974 and 1980. Most of the increase in private sector credit corresponds to credit in foreign currency, which, as a share of GDP, increased about five times over the period.

The problems faced by the Uruguayan economy in the early 1980s, and particularly by its financial sector, tightened credit conditions. Moreover, fiscal deficits, that had been negligible before 1980, absorbed a larger share of domestic credit. Thus, credit to the private sector, and particularly credit in domestic currency decreased after 1982 as a share of GDP. ^26/

Hanson and de Melo (1985) have stressed that a substantial share of private credit was in the form of consumer credit, whose share in total bank credit grew from 4 percent in 1979 to 12 percent in 1981. And that:

• Another part went to finance, and made possible, two asset price bubbles. The first was the agricultural land boom after the fourfold increase in domestic beef prices between August 1978 and August 1979. … The second bubble—which followed immediately—was the real estate boom ignited by Argentine purchasers of real estate in Punta del Este. ^27/

2. Effects on savings, investment, and the efficiency of investment

The question arises as to whether the financial deepening described in the previous subsection resulted from an increase in savings or simply reflected portfolio shifts (financial operations which were carried out in informal markets being absorbed by the formal financial sector under the new policy regime), and capital inflows (attracted by the new conditions prevailing in the formal financial sector but also responding to conditions prevailing in neighboring countries). This subsection also discusses the effects of the liberalization on investment behavior.

Chart 9 presents data on aggregate savings and investment. Also, de Melo and Tybout, ^28/ provide the following period average data:

View Full Size

URUGUAY: SAVINGS AND INVESTMENT AS A PERCENT OF GDP

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 10.

• Download Figure
• Download figure as PowerPoint slide

	1955–63	1964–73	1974–78	1979–83
Domestic savings/GDP	11.9	12.5	11.0	11.7
Private savings/GDP	…	11.3	9.6	9.6
Foreign savings/GDP	2.9	-1.7	2.5	2.9
Private investment/GDP	…	9.6	12.1	12.6

View Table

	1955–63	1964–73	1974–78	1979–83
Domestic savings/GDP	11.9	12.5	11.0	11.7
Private savings/GDP	…	11.3	9.6	9.6
Foreign savings/GDP	2.9	-1.7	2.5	2.9
Private investment/GDP	…	9.6	12.1	12.6

A straightforward reading of this evidence provides no support to the view that the domestic savings rate rose in response to interest rate deregulation. Although gröss domestic savings seem to have been only slightly lower during the reform period, private savings fell during that period. Moreover, both gross domestic savings and gross national savings showed a declining trend after 1976, following two years of rapid increase (Chart 9).

De Melo and Tybout (1986) analyzed the influence of the financial liberalization on saving rates, and concluded that the effect of real interest rates on the savings rate was statistically not significant, although the savings rate shifted upward in the post-reform period. Foreign capital inflows and real exchange rates had the dominant impact on savings.

With respect to investment, the average value of private investment ratio rose in the reform period. ^29/ This could also be attributed, as de Melo and Tybout (1986) point out, to the exceedingly low return on capital in the prereform period, as documented in Harberger and Wisecarver (1977).

However, the shift in period averages masks two distinct trends apparent in the yearly data. Investment ratios (in relation to GDP) increase steadily until 1980 (from 7.7 percent for private fixed capital formation and 11.6 percent for gross domestic investment in 1974 to 11.8 percent and 17.4 percent in 1980, respectively), with a marked decline thereafter (to 5.2 percent and 9.3 percent, respectively, in 1984).

De Melo and Tybout (1986) document structural shifts in the investment function (an upward shift of the intercept term and more responsiveness of investment to interest rates and real exchange rates) but this is not interpreted as evidence in support of the McKinnon hypothesis. Rather, they interpret the finding of significant accelerator-type effects to be a sign that Uruguayan savings were not constrained, despite the presence of ‘financial repression’ in the prereform years. Larrain (1986) points out that the rationalization and decrease of profit taxes as well as the elimination of quotas on imports of capital goods were major determinants of the shift in post-reform investment function.

On the issue of the efficiency of investment, Hanson and de Melo (1983) find a 40 percent rise in the ex post incremental output/capital ratio from 1967–1974 (0.18) to 1975–1981 (0.25). They argue that the improvement in efficiency probably reflected not only the improved allocation of credit but also the improved utilization of capacity, the rapid growth in less capital-intensive industries (which benefitted from the goods and financial market reforms) and the easing of restrictions on capital goods imports. ^30/

Thus, even though the empirical evidence do not support a strong responsiveness of domestic savings to financial liberalization, the economic reforms implemented in Uruguay in the mid-1970s raised the level and efficiency of investment.

3. The behavior of interest rates

a. Nominal interest rates

Nominal interest rates fluctuated sharply as ceilings were lifted and economic conditions changed (Charts 10 and 11).

View Full Size

URUGUAY: ANNUAL NOMINAL INTEREST RATES (PESO)

(In percent)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 11.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

URUGUAY: ANNUAL NOMINAL INTEREST RATES (DOLLAR)

(In percent)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 11.

• Download Figure
• Download figure as PowerPoint slide

In general, nominal lending and deposit peso interest rates increased during the first years of the reform, but the ceilings continued to be binding until late 1977. Interest rates were relatively stable during the second phase (1979–82), particularly since late 1979. They jumped when the stabilization policy collapsed and inflation started to regain momentum. Nominal interest rates on dollar operations increased during the whole reform period, particularly during the second phase (under the active crawling peg regime). However, interest rate ceilings on dollar lending rates were also binding until late 1977. Nominal interest rates on dollar operations became highly volatile between the second half of 1979 and late 1982 when they started to fall steadily.

With regard to the second phase of the reform, Favaro (1985) also points out two stylized facts in relation with nominal rates. First, the term structure of interest rates remained upward sloping; (i.e., interest rates on short-term deposits were below the rates on longer maturity deposits). Also, despite changes in economic circumstances and in the institutional scenario, domestic interest rates experienced little fluctuations.

The authorities attempted to reduce the spread between lending and deposit rates since 1976. However, only after 1979 did spreads decline markedly (Charts 12 and 13), when the tax on bank loans was substantially reduced and basic reserve requirements were eliminated. Increased competition in banking also contributed to the decline in spreads. ^31/ Spreads, particularly those on peso rates, widened considerably in late 1982 and early 1983. This measure of spread does not indicate intermediation costs or performance of the banking industry, as the net spread—net of the impact of taxes, reserve requirements, etc.—would do. ^32/

View Full Size

URUGUAY: SPREAD IN INTEREST RATES (PESO)

(In percent per annum)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 11.1 Prime loan rate- interest rate on time deposits of up to 6 months.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

URUGUAY: SPREAD IN INTEREST RATES (DOLLAR)

(In percent per annum)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 11.1Price loan rate- interest rate on time deposits under 6 months.

• Download Figure
• Download figure as PowerPoint slide

However, there is not enough available information to analyze the evolution of costs of intermediation. Moreover, constantly evolving regulations on portfolio and reserve requirements on financial institutions make it difficult to construct a series on net spreads from the available information on gross spreads. Nevertheless, the available data show that the removal of different regulations greatly contributed to the reduction of financial spreads.

b. The spread between peso and dollar rates

With free convertibility of the peso and no interest rate ceilings, it was expected that interest parity conditions would obtain, due to asset market arbitrage. Therefore, the domestic interest rate would be determined by the world interest rate, adjusted for exchange rate risk.

The ex post spread between the rates on peso deposits and peso equivalent rates on dollar deposits was high and varied throughout the period (Chart 14). It widened after the adoption of the tablita, remained about constant in 1980, and dropped “sharply during the period of exchange guarantees extending over seven months starting in February 1981”. ^33/

View Full Size

URUGUAY: INTEREST RATE DIFFERENTIALS¹

(In percent per annum)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 11 and 12.¹ Peso prime lending rate- peso equivalent of dollar prime lending rate.

• Download Figure
• Download figure as PowerPoint slide

The evolution of the spread between peso and dollar rates raises two issues: (1) why the rate of interest on assets in pesos exceeded the rate of interest on assets in foreign currency, even when the assets were traded in the same domestic market (so that we can abstract from country risk considerations) and (2) why this ex post realized premium kept changing over time.

The most common explanation for the existence of the premium is that of expectations of devaluation, i.e., the lack of credibility of the public in the exchange rate policy. If this is true, then the variation in the premium could be traced to the factors that affect those expectations. Moreover, a positive spread implies that expected devaluation was higher than realized devaluation.

Hanson and de Melo (1985) tested a model of interest rate determination along these lines, based on the uncovered interest rate parity using monthly data for November 1978 to December 1981. They concluded that the long-run elasticity of the peso deposit rate with respect to the rate on foreign currency deposits, 0.98, is not significantly different from the unitary elasticity predicted by the interest parity theorem, suggesting a constant ex-ante spread. Moreover, the ex post rate of devaluation seems to have had an insignificant effect on the ex ante spread.

Favaro (1985) found the expectations of devaluation hypothesis unsatisfactory and proposed an alternative explanation, based on a model of financial cost minimization which introduces the existence of real costs of adjusting a given net foreign asset position. He emphasized that, under certain policy rules, limited arbitrage opportunities may result in a wide gap between the nominal interest rate and the rate of inflation and hence in movements of the real rate of interest. He also concluded that differences between domestic and foreign interest rates do not necessarily mirror expectations of devaluations, but reflect the existence of real adjustment costs of the debt structure owing to uncertainty in the policy regime and incomplete futures markets.

c. Ex-post real interest rates

Real interest rates (ex-post) showed a high volatility between 1976 and 1985 (Charts 15 and 16). This volatility of real interest rates generated financing difficulties for the Uruguayan firms. Declining financial costs when real rates were negative (up to mid-1980) led to overindebtedness which generated a large debt service burden for Uruguayan firms when real rates went up to high levels in the next few years (up to mid-1982).

View Full Size

URUGUAY: EX-POST REAL INTEREST RATES (PESO)

(In percent)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 13.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

URUGUAY: EX-POST REAL INTEREST RATES (DOLLAR)

(In percent)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 13.

• Download Figure
• Download figure as PowerPoint slide

Blejer and Díaz (1986) conclude that external factors—foreign interest rates and tradable goods prices—were the major factors affecting the behavior of real interest rates in Uruguay, while monetary disequilibria and changes in the nominal exchange rate had insignificant effects on real rates. Also, they conclude that the exchange rate risk did not affect real interest rates.

4. The financing of nonfinancial sector firms

A study by de Melo, Pascale and Tybout (1985) discusses the interplay of real and financial shocks in the Uruguayan economy using financial statements of industrial firms. This study distinguishes three phases in the development of the financial crisis.

A first phase, of opportunities for nonoperating earnings ^34/ is described as the time when “the seeds of disaster were sown”. ^35/ In this phase (up to the late 1970s), with the opening of the capital account and the economic recovery under way, firms (especially in the exportable sector) took exposed positions in dollars to expand this capacity and to take advantage of currency arbitrage opportunities. Real borrowing costs were highly negative, owing first to the controlled interest rates and then to borrowing in foreign exchange. The authors explain that at the time, rapid dollar indebtedness may not have seemed inordinately risky because operating earnings were clearly improving, prevailing interest rates posed no major debt service problem and government reserves seemed adequate to maintain the exchange rate regime indefinitely”. These expectations would prove unwarranted in the end.

The second phase, of real side problems and rising financial cost, corresponds to the time when “the desire to survive … replaced the lure of easy money as the motive behind increasingly risky financial structures”. ^36/ When expectations of the abandonment of the Argentine tablita policy and devaluation generalized, confidence in the sustainability of the Uruguayan exchange rate regime waned. Interest rates started to climb, which induced firms to borrow in dollars in the hope that the exchange rate regime could be maintained. The authors explain that although operating earnings were stable, reflecting Argentine demand, net earnings dropped, reflecting increasing financial costs. Corporations stepped up their borrowing, partly to offset the reduction in internal sources of funds, but also, surprisingly, to finance increased fixed investment and the continuance of large dividend payments. This situation affected especially the exportable sector (with export subsidies reduced or eliminated). Thus, the scenario of the financial crisis was set, with overindebted firms left in an illiquid position and highly exposed in dollars.

The final phase corresponds to the financial crisis itself. The confidence crisis triggered by Argentina’s devaluation rapidly deteriorated the situation in Uruguay, and led to the final collapse of the exchange rate regime in November 1982, with the abandonment of the “tablita” and a 100 percent devaluation. This shock devastated the firms heavily indebted in dollars and made an important part of bank loans nonperforming, with many banks becoming technically insolvent.

Mezzera and de Melo (1985) study the importance of different shocks based on interviews of managers in manufacturing, agricultural and exporting firms. These subjective assessments also confirm the importance of financial shocks (impact of devaluation, rising costs of working capital, etc.)

5. Market structure

Describing the situation previous to the banking crisis of 1965, Daly (1967) noted that the great expansion of bank branches represented nonprice competition in a context in which negative real deposit rates, liberal rediscounting policies, and real estate speculation allowed banks to sustain high profits, despite high operating costs and a small volume of deposits.

The 1965 law severely limited the installation of new banks and opening of new branches. Spiller (1984) noted that these restrictions to entry and the easy access to information on competitors’ behavior (distributed by the Central Bank) promoted the development of cartel relationships.

Spiller and Favaro (1984) study the effects of the 1977 decree which allowed banks to raise deposits from nonresidents. ^37/ The study focuses on interaction among oligopolistic firms. Their main conclusion is that the interaction among the four dominant banks was reduced after the legal change, and that firms in the fringe group (small firms with a market share of less than 2 percent) did not expect retaliation from dominant firms.

The effects of the 1981 law, which liberalized entry somewhat, have not been similarly studied. The effects of the banking crisis on the structure of the banking system has, however, overshadowed the effects of that law and other measures. Hanson and de Melo (1985) point out that 20 of 22 Uruguayan banks have changed hands since 1981. As the financial crisis worsened and the commercial banks were being intervened or capitalized directly or indirectly by the Banco de la República Oriental del-Uruguay (BROU), the concentration in the sector has increased sharply, with the BROU and the banks owned by it holding more than 70 percent of the banking system deposits in 1986.

6. Currency substitution

The share of foreign currency assets and liabilities in the Uruguayan banking system increased sharply between 1974 and 1985 (Charts 17 and 18).

View Full Size

URUGUAY: FOREIGN CURRENCY DEPOSITS AS A PERCENT OF M3

(In percent)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 14.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

URUGUAY: CURRENCY SUBSTITUTION

(In percent)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 14.

• Download Figure
• Download figure as PowerPoint slide

The share of foreign currency deposits jumped from 11.8 percent of M3 and 31.7 percent of total time and saving deposits in 1974 to 65.6 percent and 75.4 percent, respectively, in 1985. On the credit side, whereas only about 18 percent of the credit of the banking system to the private sector was denominated in foreign currency in 1974, this share amounted to 77.4 percent in 1985. Moreover, although no statistics exist on foreign currency deposits held abroad by domestic residents, an idea of the trend of these deposits can be obtained based on the U.S. Treasury Department data on deposits in U.S. banks held by nonbank Uruguayan depositors. These deposits grew steadily from 1974 to 1976, decelerated from 1977 to 1979, and regained strength beginning in 1980 (Chart 19).

View Full Size

URUGUAY: DEPOSITS IN U.S. BANKS BY URUGUAYAN NONBANK PRIVATE SECTOR

(In millions of U.S. dollars)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: U.S. Treasury Bulletin, various issues.

• Download Figure
• Download figure as PowerPoint slide

A high and increasing proportion of foreign currency deposits in Uruguay have been held by nonresidents (mostly Argentineans). The fraction of foreign currency deposits in the private banking system held by nonresidents increased from 16.8 percent in 1974 to 49.3 percent in 1982, declining thereafter to 45.0 percent in 1985 (Chart 20). ^38/

View Full Size

URUGUAY: FOREIGN CURRENCY DEPOSITS IN PRIVATE BANKS HELD BY NON-RESIDENTS

(Percent of foreign currency deposits)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 14.

• Download Figure
• Download figure as PowerPoint slide

These huge portfolio shifts and the sharp process of dollarization are most remarkable and have potentially far-reaching implications for the implementation of monetary policy. Banda and Santo (1983) analyze empirically the domestic money substitutes and their importance for the effectiveness of monetary policy, concluding that the closest substitutes for a narrow money definition are savings and time deposits, denominated in pesos, with a low elasticity of substitution between foreign currency and domestic deposits. They argued that after the capital account was opened, increases in foreign currency deposits reflected speculative capital flows from Argentina rather than a portfolio shift by domestic residents, and that this capital inflow was often monetized—resulting in an increase in Ml. In contrast, Ramirez-Rojas (1985) concluded that currency substitution—defined as the demand for foreign fiat money by domestic residents, was empirically important in Uruguay, and should be taken into account in the implementation of economic policy. ^39/

De Melo (1985) argued that a dollarization of the magnitude observed reflected more than an increase in the demand for fiat money. He suggested that it was caused by a portfolio adjustment between dollars abroad and dollars held in the Uruguayan financial system. He based this contention on the evolution of the ratio of Uruguayan dollar deposits in the United States to dollar deposits in Uruguay. This ratio decreased sharply between 1974 (3.53 percent) and 1980 (0.24 percent) and started rising again afterwards (0.85 in 1983). ^40/ De Melo attributes these movements to changes in expectations about the evolution of the economy, and the maintenance of the exchange rate regime, and after 1982 to the impact of the domestic financial crises and insolvency of Uruguayan banks on people’s confidence.

The turning point probably began during 1980 instead of 1982, as pointed out by de Melo. At that time, deposits in U.S. banks by Uruguayans began to rise again (Chart 19) and at a faster pace than foreign currency deposits in Uruguay ^41/ while, at the same time, the share of foreign currency deposits in the domestic financial market also increased considerably (Charts 17 and 18). This evidence suggests that the external factors which adversely affected the performance of the Uruguayan economy and more expansive fiscal policies began to make the exchange rate policy less credible in 1980, promoting a new round of currency substitution.

1. The financial crisis

As described above, the financial position of Uruguayan firms, especially in the tradeables sector, deteriorated sharply between 1980 and 1982, owing to the increasing peso overvaluation, the rise in interest rates, and the historical build up of debt denominated in foreign currency. The collapse of the tablita and the accompanying devaluation was a major blow for enterprises—especially producers of nontradeables that lacked foreign exchange cover—and triggered the financial crisis by inducing loan defaults which made many financial institutions technically insolvent.

The devaluation of the Argentine peso (78 percent in March 1981) provoked a massive outflow of capital in Uruguay, owing to the lack of confidence in the sustainability of the Uruguayan tablita once Argentina had abandoned its own. The capital outflow and the drop in Argentine demand together reduced the price of assets in Uruguay, ^42/ thus also reducing the value of collateral on bank loans.

Banks’ loan portfolios were clearly affected (as Chart 21 shows). ^43/ The stage for the crisis was set. By the beginning of 1982, many banks were in serious difficulties and there was a threat of a generalized banking crisis. When the peso was finally devalued, firms indebted in dollars were devastated, and to make things worse for banks, the value of collateral had dropped, making foreclosures more difficult.

View Full Size

URUGUAY: RATIO OF BAD LOANS/NET WORTH IN THE CONSOLIDATED BANKING SYSTEM

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 15.

• Download Figure
• Download figure as PowerPoint slide

2. Policy responses

The authorities attempted a series of measures to overcome the financial crisis, but their response was complicated by the ongoing political transition. Expectations of a general debt relief were generated, and the attitude of “hold off payments, wait and see” was reinforced by the evolution of the legal developments and the workings of the judicial system which tended to favor debtors.

a. The bailout: special credit facilities and the portfolio purchase scheme

As nonperforming loans began to get financial institutions in serious trouble, the Central Bank of Uruguay (CBU) devised a series of relief mechanisms.

In 1982, and particularly during the second half of the year, some emergency support funds were provided by the CBU to troubled financial institutions. The BH received substantial financial assistance since early 1982. Net credit from the CBU to that institution increased over 1,500 percent from September 1981 to September 1982. ^44/

Another source of monetary expansion during 1982 (but prior to the abandonment of the tablita) was net credit to the Central Government. It increased about 400 percent from September 1981 to September 1982. ^45/ Moreover, two special facilities were established to assist private sector borrowers: the export prefinancing scheme and a voluntary refinancing scheme. These schemes involved subsidized interest costs which were mostly borne by the Central Bank and subsequently caused a substantial expansion of the monetary base.

The export prefinancing scheme, abolished in 1979, was reinstated in September 1982 and eliminated when the tablita was abandoned two months later. This amounted to an exchange guarantee and was made available only to nontraditional exporters. Under the scheme, dollar loans from commercial banks against future export revenues were deposited at the CBU (for a six-month term) for their peso equivalent, and when the deposit matured, the CBU would refund the original amount (but for a 10 percent annual charge). After the devaluation, capital gains were reaped by exporters, with the losses absorbed by the CBU. The voluntary refinancing scheme allowed debtors in the agricultural, industrial and commercial sectors to reschedule their debt (up to a maximum of 33 percent of the outstanding private commercial bank credit to those sectors) for a period of five years, with a two-year grace period. Maximum interest rates were to be 2 points above LIBOR (for 180-day maturities) on foreign currency loans, and 90 percent of the average rate charged by banks on peso loans. During the grace period, the borrower would pay only three fifths of the interest cost of the rescheduled loan to his bank, with CBU covering the rest. Thus two fifths of interest cost would be capitalized and repaid as part of the amortization payments, once the grace period expired. However, the most decisive relief measure was the portfolio purchase scheme, that comprised two different operations: the portfolio purchase linked to loans to the CBU (compra de carters vinculada a préstamos al Banco Central), and the portfolio purchase linked to bank intervention (compra de cartera vinculada a negociación de bancos). These two measures provided for the CBU to purchase commercial banks’ nonperforming loan portfolio with dollar- denominated bonds and promissory notes issued by the CBU. In the first operation, the banks were to arrange for medium-term external finance to the CBU in a multiple of the amount of the loan portfolio transferred to the CBU while, in the second case, the CBU arranged for the sale of a bankrupt local bank to a foreign financial institution, and purchased the bad loan portfolio with the issue of bonds and the write-off of previous financial assistance.

Under both operations, which took place from late 1982 to 1984, the CBU acquired assets for the equivalent of US$632 million ^46/ and received financing for the equivalent of US$328 million against the issuance of bonds and promissory notes for US$855 million and cancellation of financial assistance for US$105 million. ^47/ Thus the effects of the operation on the balance sheet of the CBU, abstracting from interest payments within this two-year period, would be:

Changes in the Central Bank Balance Sheet

(In millions of U.S. dollars)

Changes in the Central Bank Balance Sheet

(In millions of U.S. dollars)

Assets		Liabilities
Credit to the private sector (portfolio purchased)	+632	Long-term debt (bonds and promissory notes issued)	+ 855
International reserves (resulting from external financing to CBU)	+328
Claims on commercial banks (cancellation of financial assistance)	- 105

View Table

Changes in the Central Bank Balance Sheet

(In millions of U.S. dollars)

Assets		Liabilities
Credit to the private sector (portfolio purchased)	+632	Long-term debt (bonds and promissory notes issued)	+ 855
International reserves (resulting from external financing to CBU)	+328
Claims on commercial banks (cancellation of financial assistance)	- 105

A disaggregation by currency of portfolios acquired by the Central Bank of Uruguay is given in Charts 22 and 23. It is shown that 74.5 percent of portfolios purchased through the scheme that involved loans to the CBU and 65.1 percent of those acquired through the scheme that involved bank interventions were denominated in foreign currency. ^48/

View Full Size

URUGUAY: CURRENCY COMPOSITION OF LOANS PURCHASED

(As a percent of portfolio purchased linked to loans to CBU)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Central Bank of Uruguay.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

URUGUAY: CURRENCY COMPOSITION OF LOANS PURCHASED

(As a percent of portfolio purchased linked to bank intervention)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Central Bank of Uruguay.

• Download Figure
• Download figure as PowerPoint slide

The loans purchased by the CBU were concentrated in the livestock sector, 36.8 percent of peso portfolios and 25 percent of dollar portfolios, and the industrial sector, 21.8 percent of peso portfolios and 37.8 percent of dollar portfolios (Statistical Appendix Table 17).

The portfolios purchased by the CBU were highly concentrated in terms of the size of the debt. Less than 1.5 percent of borrowers (those with outstanding loans of more than US$1 million) had debts totaling 48.3 percent of the debt purchased, and about 16.5 percent of borrowers debtors (those with debt outstanding of US$50,000 or more) had debts equivalent to almost 92 percent of the total debt purchased (Statistical Appendix Table 18). Unfortunately, no similar figures for total credit are available for comparison. ^49/ In any case, the available information suggests that prevailing prudential regulations on risk concentration and maximum financing limits proved insufficient to preserve loan quality.

The administration and management of the portfolios purchased by the Central Bank of Uruguay kept switching throughout the period, from CBU delegating them to the BROU and the foreign banks involved in the purchase of local banks, to assumption by CBU, to administration by BROU, and finally, to administration by one of the newly nationalized commercial banks. This is partly attributable to the lack of appropriate human and technical resources at CBU for the difficult task of managing such a huge and diversified portfolios. Loan recovery undoubtedly has suffered from the indecisive management policy. It turned out that a good part of the loans was unrecoverable, with far greater costs to the CBU than had been anticipated.

Whether it was appropriate for CBU to use long-term dollar debt to pay for a portfolio partly denominated in pesos gave rise to some debate. It was justified by the Government on the basis of (a) the spreading out of the monetary effects over several years, and (b) banks’ unwillingness to hold long-term peso assets. For instance, the then president of the CBU declared later in parliamentary testimony that monetary conditions prevented a cash payment; in his view, a cash payment would have been immediately used to purchase dollars. He also argued that payment with long-term notes in domestic currency was not feasible either: the nonexistence of this type of assets was proof that they would not have been acceptable. ^50/

The portfolio purchase linked to loans to the Central Bank of Uruguay started in October 1982, when the central bank declared its readiness to purchase part of the loan portfolios of commercial banks, in return for foreign currency loans equivalent to 200 to 300 percent of the portfolio purchased. This proportion varied according to the quality of the loans purchased, of which at least 66 percent had to be of good quality—as judged by the CBU—or to be guaranteed by the selling bank. The operation was financed by the issue of promissory notes with seven years’ maturity and three semesters’ grace, which carried an interest rate of 1.5 points over the LIBOR, with repayment scheduled to be in 11 semestral installments, equal and consecutive.

The CBU did not select the portfolio, but did assess it. If according to this assessment, at least two thirds of the portfolio offered was not considered of good quality, offering banks could attempt a recomposition of the portfolio offered to CBU until that threshold was reached. For the election of transferable portfolios, banks would have to take into account borrowers’ capabilities to meet the following conditions: (1) a minimum interest rate of two points over the LIBOR for dollar debts and 90 percent of the average market loan rate for peso debt, and (2) a maximum maturity of seven years, and a maximum grace period of three semesters.

As a result of the operation, the CBU purchased portfolios for the equivalent of US$216 million, receiving fresh loans in foreign exchange for US$328 million (i.e., the credit received from the commercial banks net of the purchased portfolio value) and financing the transaction with the issuance of notes for US$544 million. ^51/

This operation was decided because the CBU was facing an acute shortage of reserves, triggered by the lack of confidence in the tablita and the massive intervention required to sustain the exchange rate. In 1982, international reserves fell by about US$1,000 million. ^{52/ 53/} Bertero (1985) has argued that a feasible alternative would have been the sale of a part of the gold reserves, and her computations show that the CBU could have obtained the same amount of foreign exchange with the sale of about 21–25 percent (depending on market prices) of its gold reserves at international prices.

Charts 24 and 25 clearly show that the main sellers of the portfolios in this operation were the foreign commercial banks, with a 60.2 percent of the loans denominated in pesos and 74.3 percent of the loans denominated in dollars. ^54/ For comparison, only 24.8 percent of the peso-denominated assets and 24.6 percent of the assets denominated in dollars of the banking system were held by foreign banks in 1982. ^55/

View Full Size

URUGUAY: PORTFOLIO PURCHASE LINKED TO LOANS, PESO DENOMINATED

(As a percent of portfolio purchased linked to loans to CBU)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 19.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

URUGUAY: PORTFOLIO PURCHASE LINKED TO LOANS, DOLLAR DENOMINATED

(As a percent of portfolio purchased linked to loans to CBU)

Citation: IMF Working Papers 1991, 030; 10.5089/9781451844849.001.A001

Source: Statistical Appendix Table 19.

• Download Figure
• Download figure as PowerPoint slide

The more sizable portfolios acquired by the Central Bank of Uruguay under this operation were sold by the Citibank (the equivalent of US$74.8 million) and the Bank of America (the equivalent of US$49.9 million), totaling 58 percent of the portfolio purchased by CBU (against loans from the banks to CBU for US$297.3). ^56/ It is noteworthy that Citibank and Bank of America were among the creditor banks negotiating the external debt of Uruguay at the time. ^57/ The portfolio purchase transaction with Citibank was closed a month before the demise of the tablita, allowing this institution to make a sizable capital gain. ^58/

The portfolio purchase linked to bank intervention was of a different nature. As the position of local banks worsened, the Central Bank of Uruguay stepped in to forestall bankruptcies by arranging the sale of insolvent local banks to foreign banks. As a condition for buying those insolvent institutions, the foreign banks got the CBU to acquire the portfolio of poor quality loans (for the equivalent of US$416 million) paying for it through the issue of bonds and promissory notes, denominated in U.S. dollars—with seven years’ maturity, two years’ grace, and at an interest one and a half points over LIBOR ^59/ (for a total of US$311 million), and through writing-off the equivalent of US$105 million in emergency financial assistance credits that had been granted to these institutions (no longer needed once they were taken over by the solvent foreign banks).

A total of five takeover operations took place. The ultimate rationale for this operation was the avoidance of a banking panic and the maintenance of the stability of the financial system. There have been claims of irregularities in some of these takeover deals, with the subsequent opening of parliamentary and judicial investigations. ^60/

The authorities had in mind a monetary approach to the theory of financial crisis and sought to implement the unavoidable bailout (after emergency financial assistance failed to save the banks in difficulties) in what seemed the most efficient way. ^61/ The main arguments advanced in favor of this way of implementing the bailout instead of some alternative way like direct intervention of the affected banks were: ^62/

• (1) It allowed a case-by-case treatment, permitting more flexibility. Interventions cannot do this, due to their adverse effect on expectations, which could trigger a generalized banking panic unless all interventions took place at the same time.

• (2) Direct intervention amounts to a monetization of the deposits of insolvent banks with negative consequences for the conduct of monetary policy. In addition, the operating costs of the intervened banks must be borne by the central bank. On the other hand, the adopted scheme allowed to spread the bailout losses over several years, and minimized the losses for the state, since previous financial assistance could be recovered and did not entail bearing the operating costs of the distressed banks. Unfortunately, there is hardly any data on the recovery of the loan portfolio purchased under both schemes.

• According to press reports ^63/ the terms initially offered to the debtors under both operations, i.e., portfolio purchases linked to loans to the CBU, and portfolio purchases linked to bank intervention were: one-year grace, two-year amortization and rates going from 71 percent to 79 percent, for peso debt; and two-year grace, three-year amortization and 13 percent interest for dollar debt.

Some measures were taken in April 1984 to attempt the recovery of part of these nonperforming assets. ^64/ A National Office of Asset Recovery was created to administer the portfolios. Also, borrowers were classified into two groups: high standard and low standard borrowers. High standard borrowers were those who had paid in 1983 at least 60 percent of the interest accrued during six months of that year and would have to pay, before May 15, 1984, 60 percent of the interest accrued during 1983 as a whole. This classification applied to both peso- and dollar-denominated debts.

High standard borrowers benefitted from the following measures:

• (1) The equivalent of 20 percent (for borrowers in local currency) and 40 percent (for borrowers in foreign currency) of the interest accrued during 1983 were written off.

• (2) Maturity periods were extended to five years with one year of grace for portfolios denominated in local currency purchased before 1984; to four years for portfolios denominated in local currency purchased during 1984; to eight years with one year of grace for portfolios denominated in foreign currency purchased before 1984; and to seven years with one year of grace for portfolios denominated in foreign currency purchased during 1984.

• (3) Borrowers in local currency could opt for converting their debt outstanding on December 3.1, 1983 into indexed debt to be repaid in 14 six-month installments at an annual interest rate of 4 percent over the adjusted principal. For each period, the adjustment factor to be applied would be the lower of the change during the period in the exchange rate or of that in the corresponding sectoral price index.

• (4) Borrowers in foreign currency could opt for converting their foreign currency-denominated debt as of December 31, 1983 into local currency indexed debt with the same interest rate and conditions given to debtors in local currency that opted for converting their debt into indexed debt.

In addition, all borrowers who, before September 30, 1984, made prepayments 180 days before the due date would benefit from a write-off equivalent to the prepaid amount, up to a maximum equivalent to 25 percent of the debt outstanding at the time of prepayment. Those debtors making prepayments (180 days before the due date) after September 30, 1984, would also benefit from a write-off equivalent to the prepaid amount but up to 15 percent of the debt outstanding at the time of prepayment.

These measures improved somewhat the debt collection: data on recovery of outstanding debt as of August 31, 1984 ^65/ show that 30 percent of the recovery on dollar debt and 34 percent of the recovery on peso debt took place between May 1, 1984 and August 31, 1984. Nevertheless, the recovery pace remained dismal. By December 31, 1983 the Central Bank of Uruguay had recovered only US$8.7 million (or 1.8 percent) of the dollar debt; and NUr$409.9 million (or 6.4 percent) of peso debt. ^66/ By August 31, 1984, these figures had risen to US$13.7 million for dollar debt and NUr$724.5 million for peso debt. The recovery pace seems to have slowed down afterwards, against the background described in the next subsection.

This poor performance can be attributed to a number of factors of different nature:

• (1) Actual insolvency of debtors, aggravated by debt concentration.

• (2) Unwillingness to pay on the part of debtors, on the expectation of a general debt amnesty.

• (3) Inadequacies of the judicial system, which have tended to favor borrowers and made it difficult to attach their property.

• (4) The belief that the future improvement of economic conditions would increase the real value of collateral, allowing a better recovery performance.

The long-term impact of this purchased portfolio on the accounts of the Central Bank of Uruguay is hard to assess, and there is no public information on issues such as how the loans are being accounted for, which part is in arrears, if interests are being capitalized. It seems, though, that an important part of this portfolio may have been written off. Moreover, the data on total credit of Central Bank of Uruguay to the private sector show sharp falls in the months of December, suggesting that a portion of these loans are written off as operational losses at the end of every year. ^67/

b. The backlash: the Domestic Debt Refinancing Law and the BROU takeovers

In the last few years of the period under study payment habits deteriorated. Growing expectations of a general forgiveness of debt pervaded society with the demands for a “political” solution to the debt problem becoming increasingly aggressive.

Regulations on compulsory selective credit to exporters were reinstated in 1984, and a forward exchange market linked to compulsory credit in pesos to exporters was also created.

Banks were confronted with an increase in arrears as debtors delayed payments in expectation of some sort of debt relief after the change in administration in March 1985 and, rather than lending to the private sector, preferred the safer returns from holding treasury bills. These expectations also affected the recovery of the portfolio purchased by the Central Bank of Uruguay under the schemes described above.

• According to a local observer, the demise of the tablita influenced notably the deterioration of respect for any kind of jurisdiction. This was aggravated by the belief that the private banks had accumulated huge profits during the period of the tablita. As a result, the issue of the inability of corporations to pay began to be considered as having a political origin and later, with the return of democratic political activity, these problems were added to the revisionist and demanding-of-damage compensation claims attitude developed since then. ^68/

This climate is reflected in the figures ^69/ on the fraction of loans in arrears and unpaid over the total liabilities of domestic debtors with private banks:

Loans Unpaid and in Arrears

Loans Unpaid and in Arrears

Date	Total Loans (in percent)
September 1980	1.5
September 1981	3.0
September 1982	11.0
September 1983	18.0
September 1984	25.0
September 1985	34.0
March 1986	42.0
June 1986	56.0
September 1986	59.0

View Table

Loans Unpaid and in Arrears

Date	Total Loans (in percent)
September 1980	1.5
September 1981	3.0
September 1982	11.0
September 1983	18.0
September 1984	25.0
September 1985	34.0
March 1986	42.0
June 1986	56.0
September 1986	59.0

The demanded “political solution” was offered by the Domestic Debt Refinancing Law (Law No. 15786), passed in November 1985, and the decrees complementary to the law (decree 83/986 on refinancing regime for the agricultural sector and 84/986 on the refinancing regime for industry and services), issued in February 1986. ^70/ The purpose of the law was to alleviate the debt burden of financially viable firms and provide the legal environment for an orderly rescheduling of that debt, in order to allow those firms new access to borrowing. The law in itself favored debtors, and provisions in the law allowed solvent firms to postpone payments. Moreover, the law included a transitory moratorium for all debtors that applied for classification under the law. Evaluation and classification of debtors according to the provisions of the law was left to a specially appointed Financial Analysis Commission, whose work was delayed with the result that loans were not serviced in the meantime. ^71/ Financial intermediaries were squeezed by the nonperformance of assets, and one after another domestically-owned banks became technically insolvent, and were taken over by the BROU.

The law’s refinancing schemes were optional for private debtors. For debtors to the banking system, a two-stage process was established. First, the Financial Analysis Commission has to determine those debtors that were eligible for refinancing. Then, commercial banks had to proceed with the refinancing agreement, which entered into effect if the debtor accepted it. However, if the debtor disagreed or was delinquent in servicing the refinanced debt for more than six months, he was subject to judicial procedures to liquidate his/her assets. The law aimed at providing debt relief to economically viable firms in financial difficulties, excluding from its provisions firms considered solvent, firms considered “nonviable” (except for some special cases), and foreign-owned firms. The provisions of the law were complicated, containing different criteria, circumstances, and terms of refinancing. ^72/

The eligibility criteria to obtain refinancing differed across sectors. “Solvent” firms were excluded, as noted. For the agricultural sector, solvency was defined in terms of maximum indebtedness per hectare, that varied across subsectors. Firms exceeding that maximum limit (NU$4,000) had access to refinancing. For industry, firms were considered solvent if the sales to liabilities (with the financial system) ratio was greater than 2.5, except for small firms which were considered solvent if indebtedness per employee was less than 20,000 pesos. For the commerce and services sector, firms were solvent if liabilities were less than 60 percent of assets, except for small firms (indebtedness less than 25,000 per employee in this case). “Nonviable” firms in the agricultural sector were those whose indebtedness per hectare exceeded a certain maximum (which varied according to the subsector); in the industrial sector a sales to liabilities ratio of less than 0.9 (priority activities) or 1 (other activities), make a firm nonviable; in the commerce and service sectors, nonviable firms were those with a negative operative margin (special activities) or that operated with negative margin and had debt exceeding NUr$2 million (general activities). All other firms that were not “solvent” or “nonviable” were “viable.” Small agricultural and industry nonviable firms were also eligible, as were debtors that refinanced their liabilities according to previous norms issued by the Central Bank of Uruguay.

In general, the conditions and terms of refinancing were contingent on the categorization of debtors. Debtors in the agricultural sector were classified in categories A, B, C, D, E, F, G, depending on the subsector, the size of the unit, and the amount of outstanding debt per hectare. For the industrial sector, categories included small firms, priority activities, nonpriority activities, and debtors that had rescheduled their liabilities under previous debt-relief programs and that had met the corresponding installments. The categories for commerce and services sectors were small firms, special activities, general activities, and debtors that had refinanced their liabilities under previous debt-relief programs. ^73/

The amount subject to refinancing was computed in two steps. First, the total amount outstanding as of June 30, 1983 was computed. For this, interest was capitalized up to January 1983 at the terms originally contracted; and after that date, penalty rates should not exceed the market loan rate for domestic debt, and the preferential rate for dollar debt. Second, the outstanding debt as of October 15, 1985 was computed taking account of the outstanding debt as of June 30, 1983 previously calculated and using pre-established interest rates: for the favored debtors (those which originally contracted loans at preferential interest rates), it was the basic rate charged by the BROU for domestic debt, and 12 percent for dollar debt; for all others, the normal rate for peso debt and the preferential rate for dollar debt. The grace period was between one and three years, depending on the sector of activity, while the payment period was between five and ten years, also depending on the sector and categorization of the debtor. The interest rates were generally lower than the market rate (from 33 percent to 90 percent of the average market rate) for peso debt, depending on the sector and category, and the market rate for obligations in dollars.

Interest payments were in some cases only a fraction of the required amounts, with the unpaid portion subject to capitalization. A fraction of the interest unpaid between June 1983 and October 1985 (depending on sector and category) could be deferred to the latter years of the repayment period. A premium of up to 20 percent for prompt payment was established for some sectors and categories. The amortization was quarterly, with amounts increasing over time as a fraction of the outstanding debt, with the precise schedule depending on sector and category.

Also, firms were not allowed to pay dividends until debt was reduced to one third of the original amount, and dividend distribution required approval of the Financial Analysis Commission and a majority of creditors (dividends were not to exceed 20 percent of profits in any case).

The debt-relief scheme involved cumbersome procedures, which allowed borrowers legally to delay servicing their debts. One of the first measures of the new parliament had been approval of a law suspending court attachment of debtors’ property for all debt contracted after July 1978 for a period of 35 days (Law 15741, of April 10, 1985). ^74/ This period was subsequently extended through November 20, 1985. The law of domestic debt refinancing extended this period for another 60 days. But, more importantly, it established that after this extension lapsed, the moratorium would continue for all debtors applying for refinancing for as long as their application was under study.

The Commission had to determine whether a debtor was subject to automatic refinancing under the provision of the law (i.e., was a viable firm or otherwise eligible firm) or not (nonviable firm). The procedure was likely to lead to delays. A debtor could submit his application to anyone of its creditors (but only to one). Then, all these applications had to be centrally processed, and all the creditors notified of all applications presented by all the firms against which they held any financial asset, after a comprehensive list had been produced. Taking into account the number of debtors (tens of thousands) and the applications presented (estimated to be about 9,000, according to press releases), ^75/ a formidable amount of paperwork was involved. ^76/ It is not surprising then that the works of the commission were delayed for more than a year, until mid-1987.

As a result, bank losses continued to increase, and a new phase of the crisis arrived, marked by the takeover of banks by the BROU resulting in progressive “de facto” nationalization of the banking sector.

The first episode of government participation in the banking sector took place much earlier, at the beginning of 1984, when the Banco del Plata was liquidated, with deposits being reimbursed.

At the beginning of 1985, before further takeovers by the BROU, the three largest private banks in Uruguay were the Banco Comercial (10.2 percent of deposits), the Banco Pan de Azucar (9.9 percent) and the Banco de Italia (4.7 percent). ^77/

In May 1985, the Banco de Italia was taken over. It was the local branch of an Argentine bank that had been intervened by the Central Bank of Argentina. The local branch was solvent, though, according to press reports. ^78/ The CBU ordered the intervention and the BROU assumed the majority of the equity.

In July 1985, the Banco Pan de Azúcar was taken over. It had become insolvent, and its head office in Chile was also in process of liquidation. The BROU assumed the equity capital without indemnization.

In April 1986, Banco Pan de Azúcar and Banco de Italia merged, under the name of Banco Pan de Azúcar. About 83 percent of the equity of the new institution is held by the BROU. ^79/

In March 1987, the Banco Comercial by then technically insolvent was recapitalized by BROU in an operation ordered by the CBU. Its capital was raised from NUr$750 million to NUr$10,000 million (with a priority right for the acquisition given to former shareholders), and roughly 90 percent of the capital was held by the BROU at the end of the operation. ^80/

In June 1987, the Banco Pan de Azúcar—in which BROU had assumed the majority of shares—acquired Banco La Caja Obrera, following a run on the latter’s deposits (which was the last domestically owned private commercial bank in Uruguay). As a result, 75 percent of deposits ended in government-owned banks, amounting to a sort of “de facto” nationalization of the banking sector.

This nationalization was a highly debated topic in Uruguay, with some arguing in favor of a “de jure” nationalization, integrating the purchased banks on the official bank system, and others in favor of a reprivatization.