Information about Western Hemisphere Hemisferio Occidental
Journal Issue
Share
Article

Brazil: Selected Issues and Statistical Appendix

Author(s):
International Monetary Fund
Published Date:
January 2001
Share
  • ShareShare
Information about Western Hemisphere Hemisferio Occidental
Show Summary Details

VIII. Restructuring Brazil’s State-Owned Financial System1

A. Introduction

1. This section reviews recent government efforts to restructure banks owned by state governments. The problems of state banks developed gradually and were associated with the fiscal problems of the state governments. Over time, the state banks became the main creditors of the state governments and, in practice, were forced to roll over credits to state governments and to capitalize the interest on these credits, thus jeopardizing their solvency and profitability. This section is structured as follows: subsection B discusses the fiscal problems of the state governments and the debt restructuring agreements granted by the central government to state governments. Subsection C analyzes the financial problems of state banks prior to the restructuring of the state government debt. Subsection D discusses the Program of Incentives for the Restructuring of the State Public Financial System (PROES), explains how the debt rescheduling agreements for the states negotiated between 1997–99 have facilitated the restructuring of state banks, and presents the main results of the program. Subsection E concludes.

B. State Fiscal Problems and the Debt Restructuring Agreements of the States2

2. The finances of state governments deteriorated in 1995 when a small primary fiscal deficit was registered. The primary deficits widened further in the following two years (Figure 8.1).3 The roots of the fiscal problems of the states were varied. As a result of the 1988 Federal Constitution, the states and municipalities assumed more spending responsibilities, without being fully compensated with additional sources of financing. States and municipalities assumed more responsibilities in the areas of health, education, water and sanitation, and public works as the federal government reduced its relative size in these areas. The increase in resources granted to the states appears to have been far less than those granted to municipalities.4

3. Another factor explaining the state fiscal problems in the second half of the 1990s is that, during the years of high inflation, state governments frequently tended to delay payments to take advantage of the eroding effect of inflation on the real value of nominal spending. This reliance led to unsustainable fiscal policies and practices that were unfolded with stabilization. In addition, the fiscal disequilibria of states were also caused by the lack of attention paid to expenditure control, that was reflected especially in the large size of the payrolls (on average between 70 percent and 80 percent of net revenue); the frequent practice by the state treasuries of assuming various debts of independent state agencies, including state-owned enterprises; and the neglect of the tax system, which manifested itself in the excessive tax concessions and exemptions and in high rates of tax evasion.

4. The deterioration of the fiscal situation of the states appears to have been related to the inability or the unwillingness of the federal government and the senate to exert forceful control over the growing indebtedness of states, and the periodic bailing out of the states by the federal government. In 1989, the federal government bailed out the states for a first time by formally assuming their external debt. In 1993, the federal government again helped the states by assuming their debt owed to federal financial institutions that the states had not serviced for a number of years. The expectation of federal bailouts contributed to the maintenance of unsustainable fiscal policies among states.56

5. Apart from the factors just mentioned of a more structural nature, the management of the finances of state governments became increasingly difficult because of large increases in salaries granted to state employees at the end of 1994, that the new state governments had to honor in 1995; the sharp rise in domestic interest rates in early 1995, following the Mexican crisis in 1994; and a stronger effort by the federal government to close the new borrowing opportunities available to the states.7 The consolidated primary fiscal balance of states and municipalities swung from a surplus of 0.9 percent of GDP in 1994 to a deficit of 0.2 percent of GDP in 1995, and to a deficit of 0.6 percent of GDP in 1996. The financial situation of states also became more difficult in 1995 due to the rise in scheduled debt service payments for those states which had renegotiated debt with the federal government in 1989 and 1993, as amortization payments began falling due following the expiration of the grace periods associated with both renegotiations. For the states having substantial contractual debt, such as São Paulo, Rio de Janeiro, and Minas Gerais, the sharp rise in interest rates in March 1995 created severe problems in terms of meeting debt service.

6. The resulting cash flow problems led the states to turn increasingly to short-term revenue anticipation loans (AROs) from commercial banks at market interest rates, to accumulate arrears with respect to payroll and payments to suppliers, to roll over almost all of the securitized debt and capitalize the accrued interest (both authorized by the senate), and, in many cases, to accumulate arrears in the servicing of loans to state banks. Due to the rapidly deteriorating fiscal situation, the federal government authorized the Caixa Econômica Federal (CEF) to provide emergency lines of credit to states in 1995.8 In exchange for these loans, which carried market interest rates and maturities up to the end of 1998, 17 states signed agreements with the federal government in 1996 that promised to implement fiscal adjustment policies consistent with obtaining zero operational balances, including by reducing payroll spending to 60 percent of net revenues by 1998; to privatize state assets; to improve and modernize their systems of tax administration and public expenditure management; and to refrain from contracting new AROs. The CEF disbursed about R$2.5 billion under this program between late 1995 and 1996.

7. The results, however, fell short of expectations. First, some of the objectives of the fiscal adjustment programs were unrealistic: for instance, the states were required to adopt measures to ensure actuarial equilibrium in their state pension systems by the end of the first quarter of 1996. Yet the fiscal adjustment programs were slow to get off the ground because the federal government was still in the process of developing its capacity to monitor the finances of the states. At the same time, because the entire amounts of the loans were disbursed upfront, the urgency to implement reforms on the part of the states waned once the immediate crisis of meeting payroll expenses had been taken care of. The voluntary retirement programs also met with only limited success.

The 1997 rescheduling of the state debts (Law 9496/97)

8. Law 9496 authorized the central government to restructure the state debts in exchange for implementing newly-negotiated fiscal adjustment programs. The new fiscal adjustment programs of the states aimed at generating primary surpluses that would permit timely debt servicing, after the debt burden was reduced to sustainable levels through restructuring. Debt/revenue ratios of the states were targeted to fall from an average value of 2.2 percent in 1996 to 1.0 percent between 15 to 30 years time depending on the terms of the rescheduling agreements. Furthermore, 11 percent to 15 percent of the states’ net revenues (own revenues plus transfers from the federal government less transfers to the municipalities) were to be committed for servicing the newly and previously restructured debt. Any debt service not covered by these percentages of net revenue was to be capitalized by the central government.

9. The debt restructuring agreements contained certain aspects of forgiveness on the existing securitized debt and a potential interest rate subsidy on the total restructured debt. The debt forgiveness arose because the capitalization of the accrued interest up to the time of the signing of the contract was added using a subsidized interest rate to the outstanding stock of debt at a specific past date. The difference between the overnight interest rate (i.e., the rate at which the central government borrows), and the interest rate on the restructured debt up to the time of the signing of the renegotiating debt contract between the state and the federal government was then assumed by the federal government.9 The securitized debt was then consolidated with other state debts and the newly-restructured state debt carried the same interest rate used to capitalize the accrued interest between the cutoff date and the time of the signing of the rescheduling agreement.10

10. The renegotiation of states debt under Law 9496 resulted in a significant subsidy to states in 1997 and 1998 as the Selic rate was generally above 6 percent plus IGP-DI during the period. In 1999, however, the potential interest rate subsidy on the total restructured debt did not materialize because the annual interest rate charged (27.4 percent, i.e., IGP-DI of 19.98 percent plus 6 percent) exceeded the Selic rate (25.6 percent).11

11. As of December 1999, the amount of debt restructured under Law 9496 equaled about R$121.7 billion out of a total state debt of R$211.1 billion (Figure 8.2).12 The total state debt was composed of debt owed to the National Treasury (R$180.7 billion, including the R$121.7 billion under Law 9496), debt owed to commercial and multiple banks (R$9.4 billion, out of which R$8.1 billion to federal banks, R$l.l billion to state banks, and R$0.2 billion to private banks), debt owed to the central bank (R$5.1 billion), external debt (R$8.6 billion), and others (R$7.2 billion) (Figure 8.3). The bulk of the state debt assumed by the federal government under Law 9496 was previously held by the state banks (73 percent) and by the federal banks (24 percent) (Figure 8.4).13 Previously rescheduled debt was excluded from this restructuring round.

Figure 8.2.Brazil: Law 9,496 and Total Debt of States

(In millions of Reais as of December 1999)

Source: Central Bank of Brazil and staff estimates.

1/ AC-Acre; AL-Alagoas; AM-Amazonas, AP-Amapá; BA-Bahia; CE-Ceará, DF-Dístrito Federal; ES-Espirito Santo; GO-Goiás; MA-Maranhão; MG-Minas Gerais, MS-Malo Grosso do Sul; MT-Mato Grosso; PA-Pará; PB-Paraiba; PE-Pemambuco; Pt-Piaui; PR-Paraná; RJ-Rio de Janeiro; RN-Rio Grande do Norte; RO-Rondônia; RR-Roraima; R$-Rio Grande do Sul; SC-Santa Catarina; SE-Sergipe; SP-São Paulo; TO-Tocantins.

Figure 8.3Brazil: Total Debt of States

(As of December 1999)

Source: Central Bank of Brazil/DEDIP - Public Debt Department

Figure 8.4.Brazil: Creditors of State Governments that Benefitted from the Rescheduling under Law 9,496

(As of June 2000)

Source: Treasury Department of the Ministry of Finance

C. The Financial Problems of State Banks

12. State banks were negatively affected by specific problems peculiar to them. The management of state banks was frequently changed for political reasons, as state administrations changed, something that produced discontinuities and deficiencies in the management of these institutions. Most state banks were technically deficient at the various levels of decision making, and frequently failed to introduce internal controls. This resulted in poor banking practices that brought about large credits to state governments and a high level of loans in arrears in several state banks; large holdings of collateral with little prospects of recovering the associated loans; and portfolios concentrated in loans and securities with low returns and/or low liquidity that were carried on the books at nominal values. In addition, large amounts of potential tax credits for future profits were kept on the asset side of the balance sheets, although they were unlikely to be fully realized in the future.

13. On the operating side, state banks had a structure of fixed costs that was incompatible with their revenue performance. They also usually had larger operating costs than private banks and loss making branches and agencies were kept open. Large liabilities were incurred through benefits granted to their employees, particularly pension benefits. State banks had to operate under a series of legal restrictions typical of the public sector that hampered their decision making in employment and other areas and had to finance developmental activities that frequently were not commercially attractive. The state banks also were exposed to large contingent liabilities, principally from labor demands. Finally, the success of the Real Plan in curbing inflation reduced the gains obtained by banks from the bank deposit float.

14. State bank problems are well illustrated in a comparison of their financial performance with that of private banks for the period after the launching of the Real Plan in July 1994. In general, state banks’ financial performance was considerably worse than that of private banks, and they lost ground to private and federal banks during the period analyzed. After the launching of PROES, that aims at privatizing, liquidating, or converting state banks into nondeposit-taking developmental agencies, the shrinking in size of state banks accelerated.

15. An analysis of the quality of state banks’ credit operations reveals that the ratio of nonperforming loans (NPLs) to total credit increased sharply from end-1996 through 1998 and only started declining in 1999 (Table 8.1).14 The sharp increase in the relative size of state banks’ NPLs from June 1997 resulted in large part from a more realistic classification of credit operations that followed the inception of PROES.15 The comparison with the private banking system shows that only between June 1995 to June 1996 the quality of private banks’ credit operations was significantly inferior to that of state banks. This is a period marked by the bankruptcy of some of the large Brazilian private banks, which motivated the implementation of PROER, the central bank program to restructure private banks in difficulties; it is also possible that during that time the private banks were using more rigorous accounting standards than the state banks.16

Table 8.1.Brazil: Credit Operations of Private and State Banks(In millions of reais and in percent)
State Banks
CREDITDec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
a. Credits in arrears and in liquidation (excluding accrued interest and fines on credits in liquidation)6731,4212,2472,5932,7763,0291,7651,8682,5511,7711,515
b. Total Credits26,00732,36839,20243,51446,88651,78417,11015,20014,56112,33911,537
c. Credits in arrears + liquidation (percent of total credits)2.6%4.4%5.7%6.0%5.9%5.8%10.3%12.3%17.5%14.4%13.1%
Private Banks
CREDITDec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
a. Credits in arrears and in liquidation (excluding accrued interest and fines on credits in liquidation)17053,53911,65111,9574,5485,3506,4185,3755,2665,5994,865
b. Total Credits64,30773,06180,72984,28390,35397,674100,80596,15897,124103,494112,068
c. Credits in arrears + liquidation (percent of total credits)2.7%4.8%14.4%14.2%5.0%5.5%6.4%5.6%5.4%5.4%4.3%
Sources: Central Bank of Brazil; and IMF staff estimates.
Sources: Central Bank of Brazil; and IMF staff estimates.

16. Semi-annual operating revenue of state banks averaged about R$17 billion over the period December 1994–June 1997 and rose during this period to the equivalent of 30 percent of private banks’ operating revenue, as the semi-annual operating revenue of private banks decreased from R$84 billion in December 1994 to R$52 billion in June 1997 (during the restructuring period that followed the implementation of the PROER) (Table 8.2). After June 1997, the operating revenue of state banks declined steadily as a percentage of the private banks operating revenue as the number of state banks was reduced and the operating revenue of private banks recovered.17

Table 8.2.Brazil: Semi-Annual Revenue and Profitability of the National Financial System(In millions of Reais)
OPERATING REVENUE
Dec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
Private banks84,42579,75562,53656,05651,86752,23473,76176,77879,896234,792100,530
Federal banks33,47332,50931,45424,15826,13523,46648,05137,89941,21067,84441,182
State banks16,80918,17918,84716,54516,78216,1318,2008,0867,4348,5676,352
Financial system134,707130,443112,83796,75994,78491,831130,012122,763128,540311,203148,064
NET PROFIT
Dec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
Private banks2,3892,3879591502,4353,5585742,9025,9529,3098,069
Federal banks766-2,178-1,104-7,3761,1551,3822,0191,4197741,438967
State banks-527-230-607-155204450-1,315-5,309-1,245-38368
Financial system2,628-21-752-7,3813,7945,3901,278-9885,48110,3649,104
Source: Central Bank of Brazil: Department of Supervision - DEFIS.
Source: Central Bank of Brazil: Department of Supervision - DEFIS.

17. The comparison also is not favorable to state banks regarding net profits. State banks’ profits were generally negative over the entire period through December 1999, except for the semesters of December 1996 and June 1997. This negative performance was accentuated after June 1997, when state banks operating revenue plunged to just R$7.7 billion on average (around 8 percent of private banks’ operating revenue). Private banks’ semi-annual net profits, after dropping in the second half of 1995 and the first half of 1996, recovered and rose steadily reaching over R$8 billion in each of the semesters of 1999.18

18. Following the pattern of net profits, the net profit rate of state banks (net profits divided by net worth) has been generally negative, while that of private banks positive. State banks’ net profit rate deteriorated substantially in 1998, as their problems became more into the open, and only moved into positive territory in the second half of 1999 (Table 8.3). The net profit rate of private banks that had fallen in 1995–97 as a result of the making of provisions, improved from the second half of 1998.

Table 8.3.Brazil: Breakdown of the Profit Rate(In Percent)
NET PROFIT RATE - net profit/net worth
Dec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
Private banks8.727.553.730.576.379.731.416.4610.6916.9412.58
Federal banks3.84-10.50-4.59-22.455.556.407.685.212.905.243.36
State banks-12.15-4.72-11.74-3.322.936.45-39.24-166.74-264.89-13.891.93
Financial system5.09-0.04-1.37-11.595.758.271.82-1.316.6212.189.45
NET MARGIN - net profit/operating revenue
Dec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
Private banks2.832.991.530.274.696.810.783.787.453.968.03
Federal banks2.29-6.70-351-30.534.425.894.203.741.882.122.35
State banks-3.14-1.27-3.22-0.941.222.79-16.04-65.66-16.75-4.471.07
Financial system1.95-0.02-0.67-7.634.005.870.98-0.804.263.336.15
TURNOVER RATIO - operating revenue/assets
Dec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
Private banks39.8829.4721.9815.5314.8612.4817.6016.1118.3348.8821.17
Federal banks20.7218.0415.4311,2011.678.8414.2211.5211.7918.6411.23
State banks24.2920.8218.0614.4613.3611.7012.5013.9014.7819.1915.99
Financial system30.4524.2419.0414.0013.5711.1715.8014.2215.3835.0116.80
LEVERAGE RATIO - assets/net worth
Dec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
Private banks772.42855.971,107.311,380.84912.721,144.611,032.771,060.44783.10874.41740.77
Federal banks810.87869.20847.95656.441,076.301,229.811,285.621,208.791,310.631,326.671,275.51
State banks1,594.721,791.362,017.562,447.431,804.861,976.231,958.251,827.5810,704.891,619.441,126.06
Financial system856.31940.431,079.491,085.251,058.341,261.901,171.601,146.461,009.321,044,27914.33
Source: Central Bank of Brazil: Department of Supervision - DEFIS.
Source: Central Bank of Brazil: Department of Supervision - DEFIS.

19. The factors behind the trend in the net profit rate of banks can be analyzed by looking at how much of the revenues materialize in profits (measured by the net margins: net profits divided by revenue); the extent that assets generate revenue (measured by the turnover ratio: ratio of revenue to assets); and the amount of assets that has been mobilized by the net worth of the bank (measured by the leverage ratio: assets to net worth) (Table 8.3). The net margin ratio of state banks was generally negative during the period analyzed, worsening after December 1997, and only recently improving somewhat. Although always positive, private banks’ net margin varied strongly over the entire period, and have increased recently.19 Private banks, in general, were more efficient in generating revenue in relation to the size of their assets than state banks. The private banks’ turnover ratio was higher than that of state banks all along the period. At the same time, state banks, in general, were more leveraged than private banks (twice as much as private banks).20

D. The Program of Incentives for the Restructuring of the State Public Financial System—PROES

20. The PROES program was established in August 1996 through Provisional Measure 1514 in order to reduce the role of state governments in the banking system and curb credit expansion to states and municipalities. Under the PROES arrangements, the central government financed the restructuring of state banks. State bank claims on impaired assets were exchanged for central government bonds, with the state governments becoming, in turn, debtors to the central government. The new debt of the states to the central government was consolidated with other state debts and restructured under the Law 9496 agreements.21 The quid pro quo for such aid was that state governments had to agree to either liquidate state banks, privatize them, transform them into a nondeposit-taking financial institution (a development agency), or restructure the balance sheets of state banks and ensure that they would be run on a commercial basis, as a condition to keeping them open. Under any of the options chosen, the exchange, to a large extent, of nonperforming assets (state bonds) for performing assets (federal bonds) facilitated the restructuring of these banks.

21. As an initial stage toward privatization under PROES, the state banks might also be federalized. In this case, the federal government signs a buy-sell contract with the state government. After that, the federal government is accountable for the management of the bank and for preparing the bank for privatization, or if this is not possible, for liquidation. Privatization revenues of federalized banks are used to amortize state restructured debt owed to the federal government. A notable example of federalization is the case of BANESPA, one of the two state banks of the state of São Paulo and the largest state bank overall. The loans to the state government were the main assets of BANESPA. As a result of the difficulties caused by those nonperforming loans, during the second half of 1994, BANESPA resorted increasingly to short-term liquidity assistance from the central bank, until the central bank decided to place it under intervention on December 31,1994.22 BANESPA was eventually federalized, and the state of São Paulo finished paying off the balance of its “conta gráfica” by transferring the bank to the central government. After several delays in its privatization because of judicial challenges, the government plans to privatize BANESPA before the end of 2000.

22. If the state government decides to keep a state bank opened and maintain control over it, only 50 percent of the costs of the restructuring program are financed by the federal government. This is also the case in the instances when state governments have more than one financial institution and they decide to keep control over one of the state financial institutions, allowing the others to be privatized or liquidated.23

23. As of August 2000, the federal government had issued R$55.4 billion in LFT-A and LFT-B securities24 under PROES arrangements (equivalent to around 5.8 percent of the 1999 GDP) and some R$3.1 billion are still to be issued (Table 8.4).25 The value of bonds issued under PROES operations amounts to about R$91.9 billion through August 2000, when the accumulated interest income linked to the Selic rate is taken into account.

Table 8.4.Brazil: PROES—National Treasury’s Securities Issues—LFT-A and LFT-B(As of August 2000)
StateTotal issued

(In millions of reais) 1/
Issue dateTo be issued

(In millions of reais)
Acre1313/29/994
Alagoas0427
Amapá2512/29/980
42/24/990
Amazonas3138/2/9935
518/25/9935
Bahia1656/1/9850
1,4336/25/9850
Ceará9855/27/99175
Espirito Santo26011/25/981/0/00
Goiás4765/27/9940
606/20/0040
Maranhão3012/15/980
3021/13/990
Mato Grosso1921/22/994
3366/15/98105
6166/16/98105
2,2806/24/98105
Minas Gerais3297/2/98105
1728/6/98105
9038/19/98105
605/4/00105
Pará1271/22/990
Paraná2,6873/5/990
1376/16/990
73512/1/990
1,63912/15/990
Perrnambuco3298/15/9820
9168/27/9820
Piauí692/24/0059
Rio Grande do Norte1013/18/990
412/22/990
Rio Grande do Sul2,38012/10/980
17312/10/980
Rondônia5495/20/984
Roraima402/18/990
Santa Catarina1983/29/992,130
685/5/992,130
São Paulo33,57912/23/970
2,54812/24/970
Sergipc411/18/990
Total55,4443,053,053
Source: Central Bank of Brazil, Public Debt Department, DEDIP.

Face value.

Source: Central Bank of Brazil, Public Debt Department, DEDIP.

Face value.

24. As part of PROES, the central bank provided liquidity to state banks by swapping short-term central bank bills (LBC-E) for the federal treasury securities (Table 8.5). Outright purchases of LFT-A by the National Treasury totaling R$29.3 billion at face value have partially refunded the central bank.

Table 8.5.Brazil: PROES—Total LFT-A Swapped for LBC-E(As of December 1999)
StateSwap dateTotal swapped 1/

(In millions of reais)
Maturity range

of the LBC Issued
Outright purchase by the National Treasury

of LFT-A in BCB’s Portfolio
DateTotal purchased 2/
SP1/16/9836,96120-139
BA4/2/9879422-14112/22/99757
MG6/15/9812,16032-8112/22/9912,160
MG7/2/9867843-64
SE7/10/984921-14012/22/9949
MG9/9/9848723-142
SC9/21/9825025-144
PE e BA9/30/9872230-14912/22/99403
SC10/6/9841431-15012/22/99653
BA10/21/9845930-149
RS11/23/989,88532-15112/22/999,748
RS12/16/982,02830-14912/22/992,019
AP1/27/992530-14912/22/9925
MT e PA2/10/9919630-14912/22/99193
RR2/24/99758-121
AP e PR3/12/993,04135-15412/22/993,046
RN3/25/995429-14812/22/9952
SE e SC4/14/9911130-14912/22/99106
RR5/3/99460-12312/22/9911
GO e CE9/29/9963430-14912/22/99109
GO, CE e A11/24/9939330-149
Total69,35229,331
Source: Central Bank of Brazil: Institutional Communication Group - GCI - FOCUS, February 2000.

Present value as of the swapping date.

Face value.

Source: Central Bank of Brazil: Institutional Communication Group - GCI - FOCUS, February 2000.

Present value as of the swapping date.

Face value.

25. The central bank also provided a bridge loan (through the CEF) to Minas Gerais in 1997 until the PROES agreement with the federal government was finalized. In addition, the central bank also financed the assumption by the Banco do Brasil of the deposits of the banks of the states of Acre and Alagoas in 1997. Despite these several operations, the central bank did not play a significant role in the financing of PROES operations as it did in the case of PROER, where it provided the financing to restructure private banks. However, the central bank, in its role as the bank supervision agency, evaluated the financial conditions of the state bank to be restructured and, in some cases, prepared the bank for privatization. Another major dissimilarity with PROER was that under PROES there was no case of a state bank being split into a “good bank/bad bank.” In the case of BANERJ, the state bank of Rio de Janeiro, the bank was indeed split into a “good bank/bad bank,” but its privatization did not take place under PROES. The “good bank” was acquired by Banco Itaú (a private bank) and the “bad bank” was liquidated by the central bank.

26. In August 1996, when PROES was initiated, 35 financial institutions were controlled by state governments, out of which 23 were commercial/multiple banks. As of August 2000, ten financial institutions had been closed, five had been privatized by the states, one is being privatized by the state government (BANESTADO of Paraná), seven had been federalized and are in the process of privatization, and five had had their balance sheet strengthened, while remaining under the control of state governments (Table 8.6). Three state banks have not participated in PROES and two states do not have banks. At the same time, 16 states have been authorized to create development agencies, out of which nine have already done so and started operations (Table 8.6). The development agencies will onlend financial resources advanced by the BNDES or multilateral creditors and will not accept deposits. They are not allowed to lend to state governments or to have access to other resources (e.g., bond issuance in foreign or domestic markets). In some cases, the development agency has inherited the development portfolio of the former state bank.26

Table 8.6.Brazil: PROES - Current Status
OptionTotalFinancial

institution
StateCurrent

status 1/
Date 2/Auction price

(in R$ million)
BanacreAcre503.31.1998-
PradubanAlagoas606.29.1998-
BanapAmapá505.26.1998-
BematMato Grosso512.16.1997-
Extinction10MinascaixaMinas Gerais505.08.1998-
CaixegoGoiás511.13.1998-
BandemRio Grande do Norte505.13.1998-
BDRNRio Grande do Norte505.13.1998-
BeronRondônia502.12.1998-
BanerRoraima103.25.1998-
BanebBahia206.22.1999260.0
BEMGEMinas Gerais209.14.1998583.0
Privatization carriedCredirealMinas Gerais208.07.1997127.3
out by the states6BanestadoParaná210.17.20001,625.0
BandepePemambuco211.17.1998182.9
BanerjRio de Janeiro206.26.1997311.1
Privatization carriedBanespaSào Paulo312.24.1997
BEAAmazonas308.04.1999-
out by the federalBECCeará305.31.1999-
government7BEGGoiás305.31.1999-
BEMMaranhào307.21.2000-
BEPPiaui303.31.2000-
BESCSanta Catarina-08.30.2000-
BanestesEspirito Santo403.31.1998-
BanparáPará403.30.1998-
Restructuring5BanrisulRio Grande do Sul403.31.1998-
Nossa Caixa NossoSão Paulo412.23.1997-
BaneseSergipe403.30.1998-
AcreAcre-03.31.1998-
AlagoasAlagoas-06.29.1998-
AFAPAmapá705.26.1998-
AFEAMAmazonas708.04.1999-
DesenbancoBahia-03.19.1998-
AF de Goiás S.A.Goiás711.13.1998-
Mato GrossoMato Grosso-12.16.1997-
Transformed intoBDMGMinas Gerais-05.08.1998-
development agencies16AF do Paraná S.A.Paraná706.30.1998-
PemambucoPemambuco-06.12.1998-
PiauiPiaui-02.26.1999-
AF do R.G. do NorteRio Grande do Norte705.13.1998-
Sul Caixa – CaixaRio Grande do Sul703.31.1998-
AFDES do Est deRondônia702.12.1998-
AFERRRaraima703.25.1998-
BadescSanta Catarina703.31.1998-
Not participating in3BRBDistrito Federal---
BandesEspirito Santo---
PROESParaibanParaiba---
States without state-2-Mato Grosso do Sul---
owned banks-Tocantins---
Source: Central Bank of Brazil: Public Debt Department–DEDIP.

Codes: 1-extinguished; 2-privatized, 3-federalized; 4-restructured; 5-legal extinction in progress; 6-under central bank intervention; 7-authorized by the central bank to operate as a development agency and already operating.

Refers to original contract date. In the case of banks that have already been privatized, indicates auction date.

Source: Central Bank of Brazil: Public Debt Department–DEDIP.

Codes: 1-extinguished; 2-privatized, 3-federalized; 4-restructured; 5-legal extinction in progress; 6-under central bank intervention; 7-authorized by the central bank to operate as a development agency and already operating.

Refers to original contract date. In the case of banks that have already been privatized, indicates auction date.

27. After the restructuring done under PROES, state banks, that accounted for about 17 percent of the financial system assets from December 1994 to the middle of 1997 (calculated from data in Table 8.7), saw their participation in the financial system shrink drastically to about 4.5 percent of financial assets by December 1999. Private banks, by contrast, have maintained their participation in the financial system to about 52 percent over the whole period, while that of federal banks increased. In line with the reduction in their share of total assets, the participation of state banks’ liabilities in the total liabilities of the financial system declined from 18 percent on average over the period December 1994–June 1997 to 5 percent in December 1999, whereas the share of private banks’ liabilities in the financial system stood around 50 percent, and that of federal banks rose to 43 percent. As regards net worth, the participation of state banks in the financial system represented 10 percent over the period December 1994–June 1997, and dropped to less than 4 percent in December 1999. Conversely, the participation of private banks’ net worth in the financial system increased from 52 percent over the period December 1994–June 1997 to 67 percent in December 1999, while the share of federal banks was 30 percent by end-1999.27

Table 8.7.Brazil: Balance Sheet of the National Financial System(In millions of Reais)
ASSETS
Dec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
Private banks211,675270,622284,458360,842349,080418,654419,088476,444435,826480,384474,968
Federal banks161,542180,238203,855215,713223,859265,503338,027328,960349,413363,878366,607
State banks69,19587,311104,348114,393125,618137,84265,62158,19050,31344,64839,716
Financial system442,412538,171592,661690,948698,557821,999822,736863,594835,552888,910881,291
LIABILITIES
Dec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
Private banks184,271239,006258,769334,710310,834382,078378,509431,515380,172425,446410,850
Federal banks141,620159,502179,814182,852203,060243,914311,734301,746322,753336,450337,865
State banks64,85682,43799,176109,719118,658130,86762,27055,00649,84341,89136,189
Financial system390,747480,945537,759627,281632,552756,859752,513788,267752,768803,787784,904
NET WORTH
Dec-94Jun-95Dec-95Jun-96Dec-96Jun-97Dec-97Jun-98Dec-98Jun-99Dec-99
Private banks27,40431,61625,68926,13238,24636,57640,57944,92955,65454,93864,118
Federal banks19,92220,73624,04132,86120,79921,58926,29327,21426,66027,42828,742
State banks4,3394,8745,1724,6746,9606,9753,3513,1844702,7573,527
Financial system51,66557,22654,90263,66766,00565,14070,22375,32782,78485,12396,387
Source: Central Bank of Brazil: Department of Supervision - DEFIS
Source: Central Bank of Brazil: Department of Supervision - DEFIS

28. The decline in size of state banks is reflected also in the sharp reduction in the number of agencies and employees that has been taking place continuously since 1994 (Table 8.8). The number of agencies was reduced from 3,844 in 1994 to 1,586 in 1999 while the number of employees dropped from 147,898 in 1994 to 40,997 in 1999.

Table 8.8.Brazil: State-Owned-Banks - Branches and Employees
STATE BANK.BranchesEmployees
199419951996199719981999199419951996199719981999
BRB - BANCO DE BRASILIA S.A.5353475656553,5144,8123,8653,5183,6303,652
BANCO DO EST. DE GO S.A.164175166168166-4,9654,3363,7673,1092,909-
BANCO DO EST. DE MT-EM LIQ. EXTRAJ44303030--1,4051,2681,164---
BANCO DO EST. DO AC S.A.161515151-78965457848342-
BANCO DO EST. DO AM S.A.3836363636361,4081,4501,3231,2881,2611,146
BANCO DO EST DE RONDONIA S.A.29282828--1,4201,101992979--
BANCO DO EST. DO PA S.A4037373737371,9721,4881,3061,057852751
BANCO DO EST. DO MA S.A.9090867979771,8742,3481,665855714601
BANCO DO EST. DO PI S.A.566777319290278262237234
BANCO DO EST. DO CE S.A.7785857871-3,4113,4032,5931,6731,480-
PARAIBAN-BANCO DO EST DA PB557788403480474442426426
BANCO DO EST. DE PE S.A.56535252--3,8503,8752,2972,095--
BANCO EST.DE ALAGOAS-LIQ.EXTRAJ.282323---1,1011,1891,128---
BANCO DO EST. DE SE S.A.4949474849501,7971,4811,0071,0621,0771,014
BANCO DO EST. DA BA S.A.174174169169170-5,6534,5204,1933,9172,867-
BANCO DO EST. DE MG S.A.491518501473--11,01211,3099,6797,969--
BANCO DE CREDITO REAL. DE MG S.A.858585---2,8273,0762,955---
BANCO BANESTES S.A.1041041048584844,4164,5043,7273,4833,5493,520
BANCO ESTADO RJ S/A-EM LIQ.EXT.227221194---12,73711,5297,420---
BANCO DO EST. DE RR S.A.509911-2162261961252-
BANCO EST. AMAPA-EM LIQ.EXTRAJ.222---110100110---
NOSSA CAIXA-NOSSO BANCO S.A.49548248248348548515,54414,30713,40312,32112,28712,313
BANCO DO ESTADO DE SAO PAULO S/A612612610569--38,86437,85434,87230,002--
BANCO DO EST. DO PR S.A.39239339239139137712,65213,88012,88612,00411,5789,641
BANCO DO EST. DE SC S.A.255254255255256-5,9918,6668,4838,0487,474-
BANCO DO EST. DO RS S.A.30331231531637037010,6489,1838,9888,5298,0817,699
Total3,8443,8513,7833,3832,2671,586147,898147,329129,349103,22158,46640,997
Source: Central Bank of Brazil: Department of the Organization of the National Financial System - DEORF.
Source: Central Bank of Brazil: Department of the Organization of the National Financial System - DEORF.

E. Conclusions

Brazil has been successful in addressing the problems of state banks that were largely associated with the financial problems of state governments. The restructuring of state governments debt established the necessary conditions to restructure the state banks. Through PROES the Brazilian authorities have been able to close, privatize, or convert state banks into nondeposit taking development institutions that cannot lend to state governments. Of the original 35 state banks, only five state banks have been maintained open after their balance sheets were strengthen with the help from PROES. The cost of PROES has been significant, but not high for international standards, and the program has closed a major source of quasi-fiscal losses in Brazil.

References

    Afonso, José Roberto Rodrigues, 1997, Descentralizaçao Fiscal, Efeitos Macroeconómicos em Funçao da Estabilização: O caso (peculiar) do Brasil, CEPAL.

    • Search Google Scholar
    • Export Citation

    Alleyne, Trevor, 1998, The Post-Real Fiscal Crisis in the States, in SM/98/27, Recent Economic Developments of Brazil, (Washington, D.C.:International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Bevilaqua, Afonso S., 2000, State-Government Bailouts in Brazil, Texto para Discussão 421, Departamento de Economia, PUC-Rio (Brazil).

    Carneiro, D., Werneck, R., andGarcia, M., 1993, Strengthening the Financial Sector in the Brazilian Economy, Texto para Discussão 301, Departamento de Economia, PUC-Rio (Brazil).

    • Search Google Scholar
    • Export Citation

    Central Bank of Brazil, 2000, Focus newsletter (Brazil).

    Central Bank of Brazil, 2000, Situação dos Bancos Estaduaís-Quadro Resumo (Brazil).

    Fonseca, D.S., Rentabilidade dos Bancos Comerciais: Um Exercicio de Decomposição, Monografia de Graduacão, Departamento de Economia. PUC-Rio (Brazil).

    • Search Google Scholar
    • Export Citation

    Giambiagi, Fabio, and AnaCláudia Alem, 1999, Finanças Públicas: Teoria e Prática no Brasil, pp. 262–64 (Brazil).

    Maia, Geraldo, 1999, Restructuring the Banking System-the Case of Brazil, BIS Policy Papers: Bank Restructuring in Practice, pp. 112–115 (Brazil).

    • Search Google Scholar
    • Export Citation

    Serra, José, and JoséRoberto Rodrigues Afonso, 1999, Federalismo Fiscal à Brasileira: Algumas Reflexões, Revista do BNDES(Rio de Janeiro, Brazil).

    • Search Google Scholar
    • Export Citation
1

Prepared by Geraldo Maia and Lorenzo Pérez.

2

This section draws on Alleyne (1998).

3

Figure 8.1 contains information on the combined primary balances of the states and municipalities. Prior to 1999, separate information is not available for the states and municipalities. In 1999, the primary surplus of the states was 0.16 percent of GDP compared with a surplus of 0.21 percent of GDP for the states and municipalities together.

4

Several authors have pointed out that the 1988 Constitution was clearer in the redistribution of revenue powers than in the redistribution of spending obligations and that, with the adoption of the constitution, the central government abandoned its spending obligations in certain areas without a formal transfer of responsibilities to states and municipalities. In these circumstances, subnational governments were forced to meet the unfilled demand for services created by the withdrawal of the central government. See Giambiagi and Além (1999); and Serra and Afonso (1999).

5

Under a first scheme authorized in 1989 by Law 7976, the central government formally assumed much of the external debt of the states. In exchange, the states entered into debt with the central government for an equivalent amount in domestic currency, but with a longer maturity and an interest rate equivalent to that being paid by the federal government. The second scheme, permitted by Law 8727 of 1993, renegotiated the states’ debt owed to federal financial institutions. The new debt of the states to the federal government had a maturity of 20 years, with a real interest rate equal to the average real interest rate of the original debt (between 6 percent and 8 percent). A debt service ceiling of 11 percent of net revenues (i.e., own revenues plus transfers from the federal government less transfers to the municipalities) was established, with any excess being capitalized. Yet another form of federal bailout for the states was a series of swapping agreements between the central bank and state banks under which the central bank swapped central bank bonds for nontradable state government securities held by state banks.

6

Afonso has suggested that part of the willingness of the federal government to “federalize” the debt of the states was a realization that its own macroeconomic policies, for example sharp rises in real interest rates, often were the factors that precipitated the fiscal crisis in the states. The fact that the states were able to exercise considerable influence in the national congress also weakened the capacity of the central government to restrict the growing indebtedness of the states. See Afonso (1997).

7

In previous administrations, large salary increases had been granted by outgoing governors, but their impact had been rapidly eroded through inflation. The increases in the wage bill granted in end-1994, which ended up in real increases of 56 percent in Rio de Janeiro, 50 percent in Minas Gerais and 27 percent in São Paulo in the 12 months through July 1995, however, were enormous compared with the real increase of just under 10 percent in the ICMS tax revenues over the same period. As a result, the ratio of payrolls to net revenues rose substantially in many states.

8

There were three lines of credit: one to be used to pay wage and other outstanding arrears; another to finance voluntary retirement programs; and a third one to refinance outstanding AROs.

9

According to Law 9496, the stock of outstanding securitized state debt was taken at March 31, 1996 and accrued interest was capitalized by using an annual interest rate of 6 percent plus inflation (as measured by the IGP-DI index).

10

In fact, the newly-restructured debt was divided into two portions. The first portion, the so-called “conta gráfica” corresponded in most cases to 20 percent of the restructured debt and currently is scheduled to be fully amortized by the end of 2002 by the states that have not done so already (Provisional Measure 2044, as of August 2000). The remaining 80 percent is to be amortized over 15 to 30 years at an interest rate of 6 percent plus the rate of inflation as measured by the IGP-DI index.

11

Both rates are calculated on a monthly accrual basis.

12

This is the total outstanding debt of states as of December 1999. The restructured debt also includes the debt restructured under the PROES program. Following the practice adopted by the Central Bank of Brazil in its publications, we include in the PROES debt the banking debt of the state of São Paulo to its state banks (BANESPA and Nossa Caixa–Nosso Banco).

13

The data used to build Figure 8.4 do not include the debt of PROES.

14

Beginning in the second half of 1997, when PROES began and state banks started to disappear, there is a break in the series of economic and financial indicators for state banks in Tables 8.1, 8.2, 8.3, and 8.7. During this time, the federal banks were also object of some degree of restructuring (for instance, Banco Meridional was privatized, and Banco do Brasil was capitalized by R$8 billion in 1996). Overall, the market share of the federal banks increased over the period (Table 8.7), in part due to the fact that some large state banks were first federalized to prepare them for privatization. See the section below on PROES for more details.

15

See Central Bank of Brazil – Focus, February 8, 2000.

16

For information on PROER, see Maia (1999).

17

The operating revenue of private banks increased steadily after June 1997 reaching R$235 billion in June 1999 as the private banks profited strongly from currency depreciation and the hike in interest rates that took place in early 1999, due to their large investments in indexed government bonds. For example, the interest income of private banks increased from R$25.4 billion in the semester ending in December 1998 to R$103.9 billion in the semester ending in June 1999.

18

A rise in the level of provisions for doubtful loans as a result of soaring delinquency rates accounts for the low level of private banks’ net profits in December 1995 and June 1996. In December 1997, the low level of profits is mainly due to the recognition of losses by the Bamerindus bank, after seeking assistance under PROER.

19

It is interesting to note that the net margin of the financial system as a whole, deteriorated in the early 1990s, the period just prior to the private bank crisis in Brazil. Data collected for a sample of public and private banks for 1991 that included, among others, Banco do Brasil, BANESPA, BANESTADO, and the large private banks show an average net margin ratio of 0.07, significantly higher than the ratio observed after December 1994. See Fonseca (1992), as cited by Carneiro, Werneck, and Garcia (1993).

20

The very large figure for state banks’ leverage in December 1998 is due to the application of stricter accounting procedures that resulted in a negative net worth of R$1.8 billion for BANESTADO (the state bank of Paraná) (see Table 8.7).

21

PROES debt could be amortized over 15 to 30 years at an interest rate of 6 percent plus the rate of inflation as measured by the IGP-DI index. In practice, however, the National Treasury has exchanged states debt for federal government securities of up to 15 years, notwithstanding the authorization to issue 30-year securities. See Central Bank of Brazil – Focus, February 8, 2000.

22

See Bevilaqua (2000) for a discussion of this period.

23

The state of São Paulo had two state banks: BANESPA and Nossa Caixa-Nosso Banco, whereas the state of Minas Gerais had four: BEMGE, Credireal, BDMG, and Minas-Caixa.

24

LFT-A and LFT-B are federal securities of 15-year maturity that pay the Selic rate. The basic difference between them is that the LFT-A is amortized monthly and pays a premium of 0.0245 percent over the Selic, while the LFT-B is redeemed by a bullet payment at the end of the contract.

25

The R$55.4 billion includes R$36.1 billion in bond issues to refinance the banking debt of the state of São Paulo to its BANESPA and Nossa Caixa-Nosso Banco.

26

The number of institutions in August 2000 does not coincide with that of August 1996 because some state banks are being “counted twice,” in the sense that a state bank may have been extinguished, while at the same time a development agency was established.

27

The sharp drop in state banks’ net worth in December 1998 is mainly explained by adjustments noted above concerning doubtful loans of BANESTADO, which resulted in a negative net worth of R$ 1.8 billion for this bank.

Other Resources Citing This Publication