Fernand Braudel Institute of World Economics

As Brazil’s latest surge of inflation was peaking in early 1994, on the eve of the launching of the Real Plan, street urchins were playing in gusts of paper money floating in the air at a traffic light along the Avenida Marquês de São Vicente, threading its way in the windy, reddish São Paulo twilight past decaying factories and warehouses, beside the river that is now the great city’s main sewage channel. At intersections throughout São Paulo, droves of children beg for money from motorists, converging on cars waiting for traffic lights to change. One driver, in a cynical joke, got out of his car to toss into the air a bagful of bills that the boys chased, briefly examined and then tossed into the air again as if to join in the game. One boy, an 11 year-old with a plastic pacifier in his mouth, was leaping to grab at the bank notes, each with a face value of 1,000 cruzeiros, equal to roughly US$15 when issued in 1990 but by now rendered worthless by the relentless pressures of chronic inflation. A few years ago this bag of cruzeiros would have fed their families for several weeks.

With the Real Plan, Brazil’s government changed the name of its money on July 1, 1994 for the fifth time in eight years. By the magic of decree, in previous plans, it had erased three zeros from all monetary and accounting denominations. Since 1985, 13 Finance Ministers succeeded each other as the money changed names: the cruzeiro novo became the cruzado, the cruzado became the cruzado novo, the cruzado novo became the cruzeiro and the cruzeiro became the cruzeiro real. The cruzeiro real, which preceded the real, lasted only 11 months. Gustavo Franco, a Central Bank director with a Harvard doctorate in economics, explained that the succession of currencies has become routine: “It’s like changing a baby’s diapers.” For two years now, the change of diapers has worked. The real was introduced after a surge of prices in the old currency, so the government could launch the Real Plan from a plateau of high prices. Driven down by tight money, high interest rates and promises of fiscal austerity, monthly inflation fell from 48% in June 1994 to below 1.5% in recent months. Inflation now is low but the institutions of chronic inflation are alive and making enormous claims on government resources. Since the Real Plan was launched, massive federal support has gone into sustaining operations of five of the 10 biggest banks that were since merged into other institutions or are still in deep trouble, in one of the most complex banking crises in financial history. Brazil’s big trouble now is a spreading cancer of bank failures and state bankruptcy. The main focus of this cancer is Banespa, the State Bank of São Paulo, until recently Brazil’s second-largest bank, which the Central Bank of Brazil reluctantly took over on the last business day of 1994, a few days after world financial markets were shaken by devaluation of Mexico’s peso. Banespa has become the biggest failure, in terms of lost assets, so far in the annals of world banking. Shortly thereafter, a diabolical front-page cartoon by Paulo Caruso appeared in the newspaper O Estado de São Paulo, showing São Paulo’s outgoing Governor, Luiz Antônio Fleury (1991-95), as the giant gorilla King Kong, seated atop the monumental office tower of Banespa in the old financial district of the world’s third-largest city, making obscene gestures at the population. The obscene gestures on the Banespa tower, once Latin America’s tallest building, display the incest between bankers and politicians that threatens to wreck Brazil’s financial system. If unchecked, these obscenities will persuade leaders that the only escape from economic collapse lies in a revival of chronic inflation. A society organized for decades to accommodate and propagate chronic inflation cannot move smoothly or painlessly into a stable regime of low inflation. Temporary reduction of monthly inflation may come quickly. Voters and foreign bankers may applaud. But economic stabilization is an act of popular sovereignty, rooted in the very purpose of democracy, that activates a long and lasting regeneration of public institutions to strengthen the values of equity and stability. Ending inflation is always and everywhere an act of courage. “Society always is ahead of the politicians in demanding stabilization,” Central Bank President Gustavo Loyola said at a seminar at the Fernand Braudel Institute of World Economics. “The gains from stabilization are spread widely, but the costs are specific to certain privileged groups. In 1994-95 we crushed inflation but not the institutional mechanisms of inflation.” Banespa’s aging tower is as conspicuous on the decaying downtown skyline as was Brazil’s economic growth, the fastest among big countries in the century before 1980. Brazil then entered a phase of economic stagnation, crippling fiscal commitments and escalating chronic inflation from which it is still struggling to emerge. The tower rises beside a spider-web steel bridge, teeming with pedestrians, that spans the concrete panorama of the Valley of Anhangabaú at the core of this metropolis of 17 million people. São Paulo led Brazil’s modernization because the state’s interior is one of the world’s richest farming regions and because its big-city economy is driven by sophisticated communications and financial enterprises and by Latin America’s strongest industrial base, catered to by world-class shops and restaurants. Yet some of its poor migrants dwell in caves dug into the foundations of thruway overpasses. The vast periphery of the metropolis, sprawled over 3,100 square miles, is the scene of surging violence and adult mortality thanks to the collapse of public health and protection under the impact of state bankruptcy. Many of these people hope that the Real Plan has given Brazil a new start with in a process of political reorganization to save them from the sins of the past. But King Kong appears as a monster of self-destruction, heeding no warning and wearing the clothing of political convenience. Fearing financial and political crisis arising from Banespa’s liquidation, Brazil’s federal authorities so far have committed huge sums to keep Banespa afloat. The $24 billion hole in Banespa’s balance sheet far exceeds the recorded losses of the second and third-largest failures in banking history: the $10 billion looting that caused the collapse in 1991 of the Bank of Credit and Commerce International (BCCI), owned by Pakistani financiers and the ruler of Abu Dhabi, and the $14 billion in losses on bad loans and investments that by 1995 wiped out the capital of Crédit Lyonnais, the biggest bank outside Japan, owned by the French government, with assets 13 times greater than Banespa’s. History’s biggest bailout yet may be the roughly $25 billion, roughly one fourth of Brazil’s federal revenues, committed but not yet spent by Brazil’s federal authorities, to keep Banespa afloat, twice what the United States spent ($12.5 billion) in last year’s rescue of Mexico. The current banking crisis is the kind of stabilization hangover that countries experience when emerging from long periods of high inflation. Since Banespa’s collapse, the Bank of Brazil, one of the country’s oldest and most pervasive official institutions, declared the biggest one-year loss ($4.3 billion) in the annals of world banking for 1995, and lost another $7.8 billion in the first half of 1996. In August 1995 the press revealed that Banco Nacional, Brazil‘s sixth-largest bank, padded its loan book with $6.7 billion in phony assets, the biggest bank fraud in world financial history. The King Kong that wrecked Banespa and bankrupted the state government was compound interest. The inflation tax, from which governments profited, has been replaced by a stabilization tax: high interest. Governments rob the public by printing too much money, thus reducing the purchasing power of their nominal obligations. Banks reap the same advantage by using sight deposits, price-shaving on retail money market investments and delays in payments and check clearances under high inflation as a source of cheap funding, which they lend to the government at high interest. The government pays high interest, much of it to foreign investors, to sustain its deficits and keep the financial system going under high inflation. When inflation stops suddenly, as in the Real Plan, high interest becomes a stabilization tax to attract foreign funds to back the local currency, while the government forgoes the inflation tax by running a tight money policy. Old habits and institutions, geared to high inflation, no longer are viable, which is why Banespa and other banks have collapsed and state governments have gone bankrupt. Since the collapse of Banespa, the search for a scapegoat, and avoidance of the consequences, has engaged leading politicians involved in this disaster. In the quagmire of accusations, tortuous negotiations and paralyzing lawsuits among politicians and central bankers, everybody’s fallback position is to transfer Banespa to federal ownership, adding to political and inflationary pressures on an overburdened government and further endangering Brazil’s program of economic stabilization. A decision on Banes-pa’s future may come before December 30, 1996, when the Central Bank’s legal authority to run the bank expires. The controversies surrounding Banespa’s collapse focus mainly on $650 million borrowed by Governor Orestes Quércia (1987-91) from Banespa in late 1990 against future tax receipts, known as an ARO (antecipação de receita orçamentaria), at the end of his four-year term to finance the election of Fleury, his protégé. Since reelection of a President, governors and mayors is banned by Brazil’s Constitution, arranging the election of a politician’s successor is a symbol of his power and prestige and helps to mobilize support in future campaigns. For Quércia, planning to run for President in 1994, this support was crucial, as it is for the current mayor of São Paulo, Paulo Maluf, a former Governor and perennial Presidential candidate, who is running up big municipal debts to boost the election campaign of his Finance Secretary, Celso Pitta, to strengthen Maluf’s next run for President or Governor in 1998. “The situation reminds me of what happened at the end of Quércia’s term, when the state became a theater of public works projects, some of them still incomplete,” observes Roberto Macedo, former dean of the University of São Paulo Economics Faculty. “The goal was to elect his candidate, Fleury, little-known at the time, just as Pitta was a few weeks ago. So the state went bankrupt, the public works stopped and the problem was passed on to the new Governor, Mário Covas.”

Shortly after his inauguration in January 1995, Covas accused his two predecessors, Quércia and Fleury, of “scorched earth tactics” in bankrupting the state and making it “ungovernable” as a result of systematic plundering of its revenues and borrowing capacity. At the end of 1993, all the state’s assets were worth only one-fifth of its obligations. Most of these assets are in 28 state enterprises, including two banks and related financial companies, four power companies with huge dams and the biggest electricity distribution network in Latin America, five transport companies (including a railroad), a water-sewage utility and a housing corporation.

A year later, the recovery value of these assets fell to one-tenth of state debts under the explosive impact of compound interest. When he took office on January 1, 1995, Covas discovered that the state had only $400,000 in its bank accounts to meet a payroll of $800 million within five days. In his final weeks in office, in the custom of outgoing Brazilian officeholders, Fleury inflated the state payroll for the new governor by granting 118% salary increases to police, judges and state attorneys. Overdue debts with contractors and suppliers totaled $4.2 billion. All over the state, 3,800 public works projects were paralyzed for lack of funds, only 439 of which had reached 80% completion. Indignation at the incapacity of public institutions to meet the needs of the population suffused the new governor’s inaugural speech, evoking images of “workers who walk to work in the dark because they lack money for a bus fare; hands of prisoners reaching out from behind bars of overcrowded and violence-ridden police station cells; anguished mothers, carrying small children, unable to get attention at public clinics;....the sick abandoned in corridors of public hospitals; children alienated by impoverishment of schools in crisis.” However, instead of addressing these problems, Governor Covas, a close political ally of President Fernando Henrique Cardoso, became obsessed with recovering control of Banespa for the state government. After a year of tortuous negotiations, a $15 billion bargain was struck to enable the state to pay its debt to Banespa. The federal Treasury would lend $7.5 billion to São Paulo and the National Bank for Economic and Social Development (BNDES) would “advance” another $7.5 billion toward privatization of the state’s railroad and three airports. However, approval by Brazil’s Senate was required, which took another six months. By then, compound interest had raised the debt by another $3 billion to $18 billion. Then Covas, a skillful negotiator, said the deal was off because São Paulo had no way to pay the extra $3 billion. A source close to the negotiations reports that “a consensus is developing among Covas and his associates against reassuming control of Banespa,” which would involve firing thousands of people and closing scores of branches, now seen by them as being politically inconvenient. Brazilian politicians find a million ways to say that all practical solutions are politically impossible. Confirming this folk wisdom is the failure of Cardoso and the resistance of Congress in efforts to end privileges of public stakeholders —civil servants, pensioners, state banks and governments, municipalities, huge federal financial institutions such as the Bank of Brazil and the Caixa Econômica Federal (CEF)— that parasitically bleed government of the resources needed to operate a complex society and to invest in long-term processes of modernization. The economic benefits of liquidation —freeing public resources tied up in unviable claims and bankrupt positions to be invested more productively— are politically dangerous and never discussed. So monetary pressures bred by escalating domestic public debt, by bank rescues and by the tangled web of exaggerated government guarantees threaten the hopes invested in economic stabilization. Brazilian politicians, including Cardoso’s reformist government, have trouble perceiving limits to creating money and credit to fudge problems of moral and fiscal cohesion. Twice in the past decade, in the Cruzado Plan (1986-87) and the Collor Plan (1990-91), monthly inflation approached zero, but then rebounded to near hyperinflation levels after the government failed to make needed adjustments in fiscal policy and structure. The trouble with the Real Plan is that cutting monthly inflation was such an easy victory, at little cost compared with stabilization efforts in other countries, that political leaders neglected to create the psychological climate of a stabilization plan in which a nation fights desperately for its survival as an organized society. When Fernando Henrique Cardoso was Finance Minister (1993-94), Professor Thomas Sargent of the University of Chicago, a historian of hyperinflations, wrote him an open letter in which Sargent argued:
Persistent high inflation is always and everywhere a fiscal phenomenon, in which the Central Bank is a monetary accomplice....A Central Bank cannot by itself stop an ongoing inflation against the will of a fiscal policy authority determined to run persistent budget deficits. Indeed, a Central Bank determined to ‘go it alone’ and to fight inflation with tight money in the face of persistent deficits can achieve only temporary gains in the battle against inflation, and at the cost of making inflation worse in the future. This outcome results because, in the face of a persistent fiscal deficit, a Central Bank can achieve go-it-alone tight money only by forcing the fiscal authority to issue increasing amounts of interest-bearing debt: without a fiscal adjustment down the road, the monetary authority will eventually be forced to generate more inflation down the road in order to raise the inflation tax....Credibility is not something that a small number of people, even Presidents and Ministers, can manipulate. In a democracy, a fiscal policy credible for supporting a stable currency must be arranged so that ‘the votes are there’ for levying enough explicit taxes to cover whatever expenditures have been voted.

Cardoso and Covas have steadfastly avoided the hard choices posed by state bankruptcy and the collapse of Banespa. Meanwhile, Quércia and Fleury, in political disgrace because of the aura of corruption enveloping their administrations, mounted a campaign in the press and the courts to save their reputations and to discredit the Central Bank’s handling of Banespa.
Banks are prime sources of election finance in Brazil. Their survival and continuing liquidity become a tragic obsession for politicians, involving colossal waste of resources. Along with Banespa, the Central Bank took over four smaller state banks and, months later, two big private banks, Econômico and Nacional, amid revelations of fraud and political scandal. Brazil now toils in a political and financial quagmire engulfing much of its public and private banking system. Inflation is low now, but Brazil’s oversized banking system, concentrated in the public sector and engineered to profit from chronic inflation, is in deep trouble. Until now, Brazil has sailed along on a crest of expanding world liquidity while avoiding institutional changes demanded by stable prices. Failure to deal effectively with the overhang of fiscal and banking problems will breed inflation in the future.

Brasil´s federal system is
under great strain. Other states
reproduce São Paulo´s bankruptcy
on a smaller scale.

São Paulo’s state government now owes roughly $72 billion, 16% more than Brazil’s entire debt to foreign banks ($62 billion) before it got relief under the Brady Plan of the U.S. Treasury. Central Bank auditors, poring over the giant carcass of Banespa in early 1995, found $13 .5 billion of bad debts on its books, dwarfing its nominal capital of $1.9 billion. Of these bad debts, $9.5 billion was owed by Banespa’s owner, the state government. Brazilian banking laws bar concentration of loans in a single borrower and self-dealing by controlling interests, but state banks were freed from these limits from 1975 to 1990, subject to approval of each loan by the Central Bank. In December 1990, as Senator from São Paulo, Cardoso proposed a special resolution enabling Quércia to borrow the $650 million ARO from Banespa, which was rolled over repeatedly in the 1990s as it ballooneda into a debt of $3 billion as high interest was capitalized. With the Central Bank running Banespa, the $9.5 billion state debt to its own bank grew to $20 billion. Capitalized interest is accruing at a rate of $700 million monthly. Another $4 billion in bad debts is owed by the private sector, mainly from insider loans originally funded by Banespa’s foreign borrowing. Of $57 billion in loans outstanding, the state government owes $24 billion to its two banks, Banespa and Nossa Caixa, Nosso Banco [Our Fund, Our Bank], an old savings institution that recently became a universal bank, with 12,000 employees and 550 branches, currently is ranked 10th in assets among Brazilian banks. The state pays $20 million monthly on its $5 billion debt to Nossa Caixa under a previous renegotiation, while capitalized interest grows this debt by $100 million each month. Negotiations with the Central Bank on how to merge Banespa and Nossa Caixa into a single state bank have begun, but no viable formula is in sight on how to deal with the state debts. Of $20 billion in state debts to Banespa, two-thirds are from the bank’s guarantees of old loans to defaulting state agencies and corporations that were rolled over continuously for more than a decade at what became astronomical real interest rates in the 1990s. The final blow, dramatizing the political and financial vulnerability of Banespa, was Quércia’s $650 million ARO, now amounting, because of compound interest, to one-third of these debts. São Paulo owes another $10 billion in unpaid judicial awards. Creditors are suing for a federal takeover of the state government to secure payment of these awards under a rarely used constitutional provision.
Top Central Bank officials were divided over whether or not to liquidate Banespa. The Central Bank interventors appointed to run Banespa never received a reply to their requests for authorization to cut back the scale of operations. In the 20 months since the Central Bank took over Banespa, political pressures —from mayors, unions and legislators— prevented the new management from reducing significantly the number of employees and branches. Of its two million accounts, 1.7 million are of state employees. Four-fifths of Banespa’s funding consists of deposits of the state government.
“Banespa basically has been bust since 1980,” said one former director. “By then it had lent more than its entire capital to a single customer, the state government, which also was its owner and which didn’t repay its loans. At the end of 1990, the Central Bank banned all further AROs, which dealt with flow of new loans but not the existing stock that was being rolled over continuously. Private banks made huge profits by funding Banespa’s unpaid state loans on the overnight interbank market as the state debt rose from $4 billion to $20 billion since December 1990.”
Banespa stayed alive for so long because, like other banks, it enlarged its branch network in long periods of high inflation to benefit from the “float” of delays in payment and clearance while lending the funds to the federal government at high interest. In capturing this inflation tax, Banespa had extra privileges as the State of São Paulo’s paymaster. Of its three million personal and corporate accounts, 60% were held, as legally required, by state employees, contractors and suppliers in order to get paid. Another $1.4 billion in judicial escrow accounts was deposited by law at Banespa. Also, accounts of state and municipal governments and public corporations were freed from Central Bank reserve requirements. With $28 billion in assets at the end of 1994, 611 branches and 53,000 people on its payroll (including pensioners), Banespa’s physical network is roughly the size of Citibank’s domestic U.S. operations with less than one-tenth of Citibank’s assets and capital.
All the options are painful. The Central Bank’s ideological choice is privatization, which is not viable. Privatization would drain Banespa’s privileged source of cheap deposits as a public bank. A privatization proposal was rejected by leading private bankers, who face a shakeout in Brazil’s oversized banking industry and have no incentive to take on the grim task, avoided by politicians, of firing thousands of employees and closing hundreds of branches. The remaining options are liquidating the bank, seen by the main actors as politically impossible, and federalization, a recipe for more bailouts and an unending source of printing money to pay bad debts.
The problem will not go away. Banespa is the leading example of Brazil’s fiscal problems that must be solved before inflation is beaten. In early 1995 the World Bank estimated at $140 billion the total debts of Brazil’s 27 state governments, which together now run $24 billion in deficits annually. Most are desperate for new loans. A new federal loan program has been announced to bail out state governments from their debts to state banks on the condition that they close, privatize or recapitalize their banks. Finance Ministry officials say the program reschedule the $24 billion in state bank debts over 30 years.Calculations of the fiscal cost of these bailouts are scratches in the sand, with no provision for defaults on rescheduled loans.
Brazil’s federal system is under great strain. Other Brazilian states, bred in the same political and financial culture, reproduce São Paulo’s bankruptcy on a smaller scale. Debts are rolled over perennially and payments almost never made. A soft budget constraint becomes no budget constraint. The stock of debt swells and the rate of interest no longer matters because debts are not paid. But when inflation stops, then money really matters. A nervous breakdown forced the Governor of Mato Grosso into seclusion for two months earlier this year after as the state government fell five months behind in paying salaries, which absorb more than 80% of revenues of many states. New federal loans were granted only after Mato Grosso’s legislature agreed to privatize the state bank and power company, but the state quickly became one of 12 states asking for 90-day debt relief during the current municipal election campaign. The 19 states that were bailed out with $2.1 billion in new federal loans at the end of 1995 have paid only $268 million in debt service as capitalized interest accumulates. Also five months behind in paying salaries is the Northeastern state of Alagoas. Its payroll consumes 117% of its $20 million monthly revenues, leaving nothing to service a $1.1 billion state debt at 8% monthly interest. The division of Alagoas’s payroll mirrors the distortions of income distribution in Brazil. The 3,500 highest-paid state employees together get as much as the 51,000 poorest paid; some retired police colonels receive pensions of $20,000 monthly, which are not uncommon among state governments. The state legislature and judiciary set their own salaries and staffing levels. Alagoas’s legislature has 3,000 employees, though only 500 can fit in its building. The federal government is lending US$65 million to Alagoas, which also is trying to borrow US$ 160 million abroad on the Euromarkets.

Brazil is a continental archipelago of communities speaking the same language and flying the same flag, with big differences in achievement and shifting centers of power. The regression of modern institutions is concentrated in big cities like Rio de Janeiro and São Paulo. Vast swaths of the country have the trappings but not the effective substance of modern institutions, even though they are linked to the outside world by modern systems of communications and transport. Some states and municipalities are solving their problems, mainly excess employment and debts. Some, like Ceará and Paraná, began to deal with these problems a decade ago. The state governments of Minas Gerais, Bahia and Rio Grande do Sul have taken big steps to reduce their fiscal burden. Even São Paulo increased revenues by 27% in 1995 and balanced its non-debt budget for the first time in several years. “We had 210,000 employees when I took office and now we have 180,000,” says Rio Grande do Sul Governor Antônio Britto. “We have closed 10 state corporations. But our debts have grown from $1.3 billion in 1991 to $5.5 billion today. I’ve been telling the federal government that we’ve built an atomic bomb, made not of principal but of compound interest, that will blow up not only my government but all of Brazil. The federal government finally seems to understand the seriousness of this because it now faces the same problem itself.”
The dilemma between economic collapse and revival of chronic inflation arises from an expanded concept of looting as developed after the U.S. savings and loan banking crisis by two economists at the University of California at Berkeley, George Akerlof and Paul Romer, who argue:

Bankruptcy for profit will occur if poor accounting, lax regulation or low penalties for abuse give owners an incentive to pay themselves more than their firms are worth and then default on their debt obligations. Bankruptcy for profit occurs most commonly when a government guarantees a firm’s debt obligations. The most obvious such guarantee is deposit insurance, but governments also explicitly or implicitly guarantee the policies of insurance companies, the pension obligations of private firms, virtually all the obligations of large banks, student loans, mortgage finance of subsidized housing and the general obligations of large and influential firms....As a result, bankruptcy for profit can cause social losses that dwarf the transfers from creditors that the shareholders can induce. Because of this disparity between what the owners can capture and the losses that they create, we refer to bankruptcy for profit as looting....If net worth is inflated by an artificial accounting entry for goodwill, incentives for looting will be created.

The concept of bankruptcy for profit applies to Brazilian politics. What happened to Banespa shows how this kind of political looting works. This looting has been widely practiced in Brazil, in both the private and public sectors. In Brazil, the capital value of “good will” is a political concept that enhances the capacity of firms, especially public banks, municipalities, state governments and other franchised stakeholders, to extract resources from the political system, generating widespread parasitism and escalating claims against government that cannot be met within a clear framework of public accounting. The game can only be kept in motion through the deceptive accounting of chronic inflation. The obscenity of this escape lies in the continuing degradation of modern public infrastructure and institutions —schools, hospitals, public security, roads, water, sewage and electric power systems— that lay in waste because of previous waves of looting. At both the federal and local levels, state bankruptcy drains the will and capacity of government to absorb liquidations, among bothofficial and private financial institutions, which is the cost of moving from a regime of chronic inflation to market-oriented economic behavior. This may be a transitional problem, but the price of transition cannot be escaped and the price of delay continuously rises. At all levels of government and in its banking system, Brazil now faces cruel choices between liquidation and resurgence of inflation. Liquidation of failing public institutions is alien to Brazilian political culture. Closing state banks and firing their employees only has begun to be discussed recently. Awareness may be growing that monetary pressures bred by escalating domestic public debt, by bank rescues and by the tangled web of exaggerated government guarantees are undoing the hopes invested in economic stabilization.

Too Big to Fail

“Most countries faced with bank failures delay recognition of the failures and eventually assume responsibility to the depositors and socialize the loan losses," observes Allan Meltzer of the Carnegie Mellon University. "The main economic rationale for financial regulation and supervision is to prevent failure of the payments system. Some would broaden the statement of purpose to include the entire financial system. The emphasis in either case is on the system, not individual firms or institutions. The reason for this emphasis is that, in principle, well-designed regulation can reduce the risks that society must bear."
Like Crédit Lyonnais, Banespa was seen as being too big to be allowed to fail. Fear of financial and political turmoil, known as systemic risk, endows survival of big banks all over the world with an aura of high public interest, breeding large and growing claims on government resources. These claims are magnified by negligence in bank supervision and by political pressures on central banks. The "systemic risk" rationale focuses on the web of loans among banks on the interbank market in support of the argument that failure of one big bank would cause a chain reaction of panic among creditors, who in turn would be unable to meet their commitments, thus inflicting damage to the payments system. The historical cases cited most often in support of the "too big to fail" doctrine are the collapse in 1931 of Austria’s Creditanstalt, deepening the Great Depression, and the Federal takeover in 1984 of Chicago’s Continental Illinois, both of them big short-term international interbank borrowers. The difference between Brazil’s current banking crisis and historical "too big to fail" experiences is that the interbank exposure of failing Brazilian banks is almost entirely with government institutions. Over the past year, the number of Brazilian banks with access to private interbank lending has been halved because of increasing risk. For example, Banespa’s non-performing state loans were funded on the interbank market at shocking interest rates, reaching 16% monthly, before private banks withdrew their credit in September 1994 and the Bank of Brazil stepped in as an intermediary, guaranteeing Banespa’s overnight loans. Thus no risk to the payments system is posed if Banespa were to be liquidated.
In June 1996 Standard & Poor’s reported: "Brazilian bank industry risk is high because government regulations are in some cases distortive and in some cases lax." One of the distortions leading to the current wave of Brazilian bank failures is the Central Bank’s exacting compulsory reserve requirements, which tighten bank liquidity and drive up interest rates, which in turn enable Brazil to attract foreign funds to support an overvalued currency that is one of the main props of the Real Plan. High interest rates, needed at first in most stabilization efforts, quickly outlive the usefulness and must be replaced by credible fiscal policies. Using three different indices, the Fundação Getúlio Vargas (FGV) calculates that the real effective value of Brazil’s currency increased between 29% and 47% since the Real Plan was launched, urging the government to "accelerate mini-devaluations of the real to avoid a devaluation shock later and to halt a fall in exporters’ profit margins that is "bad news for a country running a deficit in its external accounts and with urgent need to invest in order to export." This view follows a consensus among economists now cautioning against prolonged use of overvalued currencies as an anchor to stop chronic inflation. The Bank for International Settlements (BIS) in Basle, the central bank for central banks, warns: "Exchange rate-based stabilization programs, which have often resulted in sharp reduction in inflation, are frequently maintained for too long and the exchange rate becomes overvalued. The longer an unrealistic exchange rate is maintained, the greater will be the chance that finance flows to the wrong sectors and the higher will be the subsequent costs of dislocation. This will be especially true if a reversal of short-term capital flows forces a large and sudden adjustment. Indeed, several financial crises in Latin America have been triggered by very abrupt exchange rate changes," adding that poor economic performance also has been caused by "an uneven institutional fabric, bad banking practices, weak prudential oversight and the inevitable problems encountered in the transition to a more liberal system."
The initial success of the Real Plan, creating political and price stability in Brazil for the first time in many years, launched foreign investors into a euphoria only briefly dampened by the "tequila effect" hangover from the 1994-95 Mexico debacle. Punters buying Brazilian financial assets in early 1995 got a 30% annual real return in dollars, reduced to roughly 20% in 1996 as the Central Bank eased local interest rates. In the first five months of 1996, Brazil received gross long-term capital flows (at least 360 days) of $28 billion, double the 1995 investments in the same months. Direct foreign investment ($3.3 billion) grew at triple the 1995 pace, spurred by hopes of accelerated pri-vatization and increases in popular consumption that tempted international companies to seek a bigger share of a huge potential market.
Many blame Brazil’s current epidemic of loan defaults on real annual interest rates of up to 30%, the magnet that enabled Brazil to multiply its foreign exchange reserves tenfold since 1991 to $60 billion, fourth-largest in the world after Japan, China and Taiwan. Government officials proudly point to these huge reserves, four times the size of the monetary base, as a kind of Maginot Line, defending an overvalued exchange rate and the Real Plan itself. But admirers of this Maginot Line may not recall how fast Mexico’s reserves fell from a peak of $29 billion in February 1994 to only $4.4 billion in January 1995. What sank Mexico in 1993-94 was a big credit expansion in an overvalued currency. The same happened in the first 18 months of the Real Plan, but in 1996 the stock of private domestic credit has leveled while public debt has kept growing. Concern is spreading that events in the world economy outside Brazil’s control, such as a rise in U.S. interest rates or failure by Argentina to finance its 1996 deficits, may drain foreign funds from Brazil. "The resources will stop coming and will start to go," says Francisco Gros, twice Central Bank President who now works for Morgan Stanley on Wall Street.
Some broad purchasing power gains from lower inflation are being eroded in a consumer credit boom. Bankers estimate that 70% of personal cheks issued in Brazil now are predated, a conventional form of consumer credit invented during high inflation. Poor people also have been paying monthly interest rates of 14%-18% in 1995, falling to 4%-7% this year, to buy consumer durables ranging from electric fans to cooking stoves to used cars. For purchases at monthly interest of 7.5% in six installments, lenders would break even with a 29% default rate. If defaults are only 3%, the historical average for big São Paulo stores, profits reach 37%. The lending boom is financed by Brazilian and foreign banks. "Many banks are trying to buy finance companies because of the spread between foreign borrowing at low interest and domestic lending at high rates, which is very profitable because of the interest distortions imposed by the government’s economic policy," says Affonso Celso Pastore, a former Central Bank President. "Today whoever buys a finance company recovers his investment very fast. For this stabilization program to run its course, re-storing economic growth, we must have interest rates near international levels. Then the consumer credit market will be much bigger."
The problems of transition have led to complicated flows of support between Brazil’s federal banks, the Treasury and weakened financial institutions, hard to measure because of opaqueness, secrecy, omissions and delays of published accounts. A piecing together of these rescue operations, both promised and already realized, suggests enormous displacement of resources. Official financial commitments for bailouts, mergers, recapitalization, debt swaps, advances toward purchases of state government assets, Central Bank losses and other forms of support of the banking system and the currency since the launching of the Real Plan may exceed $100 billion. This displacement of $100 billion is perplexing not only because much of it escapes conventional monetary and fiscal accounting and because it amounts to nearly twice foreign currency reserves and the whole banking system’s capital ($52 billion), over one-fifth of all its assets ($464 billion) and one-third of the broad money supply (M4). It also animates a culture of deranged economic transfers at the core of long-term processes of chronic inflation. This culture led the World Bank to report in 1983 that the volume of transfers "is such a unique feature of the Brazilian economy that one may refer to it as a transfer economy to distinguish it from the market and centrally planned economies." The Bank of Brazil used to record huge profits from its Conta Movimento, a massive rediscount facility of virtually interest-free funding from the Central Bank that by the early 1980s approached the size of the monetary base. In 1978 the flow of credit subsidy amounted to 54% of total federal revenue, with over one-third of all loans to the private sector highly subsidized. Intelligent, dedicated officials have struggled hard over the years to reduce and modify these distortions. Progress has been made in reducing subsidies and clarifying public accounts. Since passage of the 1988 Constitution, government lending was cut steeply. Yet new gimmicks are always appearing and their distortions can have more impact in a climate of lower inflation. The displacement of resources in Brazilian currency support and bailout of banks and state governments, now roughly 17% of GDP, falls within the scope of fiscal cost of recent bank rescues in other Latin American countries, ranging from 12-13% in Mexico, Argentina and Venezuela to 20% in Chile in the early 1980s. No economy can sustain such a huge drain on its resources on a continuing basis. The main Brazilian support operations, already disbursed or announced, come through a jumble of fiscal channels:

•$28 billion in Central Bank swaps with state banks of unmarketable state government debt paper for negotiable federal notes.
•$15 billion to recapitalize the Bank of Brazil, of which $8.7 billion already has been disbursed for a new stock issue. After declaring its huge loss for 1994, the Bank of Brazil’s financial position was weakened further by a political deal in Congress as the government, seeking passage of a gutted constitutional amendment for social security reform, agreed not to collect $7.6 billion in defaulted debts by big landowners in exchange for votes in Congress.
• $13 billion in published Central Bank losses accumulated from June 1994 to December 1995. Most of these losses came from mismatches between interest received on foreign exchange reserves and interest paid on internal public debt floated to prevent the inflow of foreign funds from swelling the domestic money supply. Also contributing to these losses is the mismatch between the market interest rates paid by the Central Bank on Treasury deposits and the low interest received by the Central Bank on its Treasury bond holdings.
• $13 billion in new federal loans to state governments to enable them to sell or liquidate state banks, or to recapitalize the banks if a state can pay half the cost of restructuring.
•$15 billion in emergency loans under PROER, the Central Bank’s Program of Incentives to the Restructuring of the Financial System, funded by compulsory reserve deposits of other banks, to restructure the balance sheets of failed private banks for merger with other banks.
•$13 billion in Treasury loans to state governments to pay loans from state banks, including $7.5 billion to São Paulo State to pay part of the state government’s debt to Banespa.
•$7.5 billion from BNDES (National Bank for Economic and Social Development) as advance against purchase of São Paulo airports and railroad for privatization later; the advance is to be used to repay Banespa loans to the state government. The reported $3.6 sale price to be paid by BNDES for Fepasa, the São Paulo railroad, is several times its appraised value.
•$12 billion in revolving overnight loans and the interbank market borrowed from private banks by Bank of Brazil and CEF relent to state banks, mainly Banespa, for daily funding of their balance sheets.
•Apart from these specifically announced commitments, a large share of the $12 billion in revolving overnight loans on the interbank market is borrowed from private banks by Bank of Brazil and CEF relent to troubled banks for daily funding of their balance sheets. Another large share of the $6 billion in revolving Central Bank rediscount window loans is funding troubled banks.

Recent financial crises in Scandinavia, Japan and U.S. savings and loan institutions show that the final cost of a banking system bailout are often greater than original estimates. While these bailouts are absent from most fiscal accounts, they can be seen in a broader perspective. Over the past year, bitter public protest arose in Japan against the use of $6.3 billion in taxpayers’ money to rescue seven failing jusen (mortgage lenders), to which politically-powerful farm cooperatives had lent heavily and which criminal gangs had looted. But this $6.3 billion jusen rescue was the only government bailout so far in Japan during its present banking crisis, involving just 0.14% of the banking system’s $4.5 trillion in outstanding loans. Of these, 10%-20% are now seen as unrecoverable thanks to collapse of real estate and stock market prices in the early 1990s. Brazil’s public opinion and politicians, meanwhile, have been more passive in dealing with bailouts involving 16 times more money than the jusen rescue in a banking system with only one-tenth of the assets of Japan’s, but with interest rates several times higher. Brazilians have been more passive because of enormous pressures to save bank and government jobs and to roll over unpayable public and private debts far into the future.
The profits from inflation that kept banks and governments afloat, facilitating political arrangements, have nearly vanished since launching of the new currency in July 1994. The inflation tax transferred to banks from the population fell from $9.8 billion in 1993 to just $460 million in 1995. Brazil’s poorest families gained hugely in purchasing power as inflation fell, improving income distribution. Their gratitude gave Cardoso a landslide victory in the 1994 Presidential election because, as Finance Minister, he presided over drafting of the Real Plan in 1993-94. Incentives for ending chronic inflation remain high for the Brazilian people as a whole, less so for bankers and politicians as well as for bank and public employees, whose ranks grew over the past decade of high inflation and could lose jobs in a tighter fiscal regime.
Key parts of Brazil’s patronage system are Brazil’s state and federal banks, such as Banespa, making 55% of all loans and now forming a huge financial garbage dump, littered with wasted resources and unrecoverable assets, often created for political reasons. After reporting $12.1 billion in losses for 1995-96, purging huge accumulations of bad debts from its balance sheet, the Bank of Brazil’s new managers were praised for their courage, given the difficulties they faced in downsizing and restructuring the bank’s operations. As a result of these losses, the Bank of Brazil, which was the monetary authority before the 1960s and fiercely resisted creation of a new Central Bank in 1964, lost its place as the country’s biggest financial institution to the CEF, which has massive problems of its own. With an overhead of 115,000 employees and 3,000 domestic branches, the Bank of Brazil’s new reform administration, headed by a former Central Bank President, was blocked by political and labor union resistance in its plans to cut staffing levels and close money-losing branches. According to Mailson da Nobrega, a former Finance Minister who began his career as a Bank of Brazil officer: "The Bank of Brazil still has not recovered from the loss a decade ago of its Conta Movimento". The Bank of Brazil responded to its loss of free funding by trying to become a financial conglomerate, starting new leasing, brokerage, insurance and credit card businesses. But it was forced to adopt outdated computer technology from a failing government corporation, Cobra, and was burdened further by a bureaucratic culture and acquired rights of its highly-paid employees who resisted change. However, the institutional credibility of the Bank of Brazil is such that, despite these problems, it gained deposits in a "flight to quality" when other banks got into trouble.
At the core of a highly concentrated system, five government banks hold 45% of all assets and the Big Six private banks (Bradesco, Unibanco, Itaú, Bamerindus, Real and BCN) another 22%. Operating costs (8.3% of total assets) are among the highest in the world, compared with 3.2% of assets for banks in the United States, 2.3% in India and 1% in Germany and Japan. Earlier this year, Central Bank accounts showed that 16% of all bank loans to the private sector are overdue or unrecoverable, despite two-thirds growth of the loan portfolio since January 1994. While weaker banks keep rolling over bad loans, stronger ones shrink their balance sheets in real terms and make provisions for bad loans. "It’s a bad sign when banks expand their balance sheets under these conditions," said one Central Bank official. "It means that many banks are helplessly rolling over bad loans and making new ones to desperate firms with no prospect of paying today’s punishing interest rates."A new study by the FGV shows that, between the last full year of high inflation (1993) and the first full year of low inflation (1995), provisions for bad loans by the Big Six rose from $273 million to $3.7 billion, or 27% of their net worth. Personnel and administrative costs kept escalating and rents from the payments float collapsed from $5.1 billion in 1993 to only $344 million in 1995. This performance would have been much worse if Banespa and two of 1994’s private Big Six, Banco Nacional and Banco Econômico, were not excluded from the 1995 rankings after collapsing into the arms of the Central Bank, which absorbed their bad loans and poured another $14 billion into their carcasses to merge them into other banks amid scandals of fraud and diversion of assets. In addition, Bamerindus, Brazil’s sixth-largest bank, has been struggling to survive with Central Bank support despite a $1.9 billion liquidity shortfall.
Such debate has focused on the $15 billion in rescue loans of PROER to support merger of failed banks, mainly Nacional and Economico, into stronger institutions. As in Argentina in the early 1980s, banks with stronger loan portfolios thus were forced to finance the bailout of troubled banks by holding high reserves so that the Central Bank could make emergency loans. Of 271 Brazilian banks, 58 suffered Central Bank intervention or liquidation or changed ownership in the past two years. Of these, five state banks, including Banespa, and 26 private banks came under Central Bank management. Another 22 changed owners without Central Bank PROER loans, which go to the new owners to clean up the balance sheets of failing banks during mergers. Under a formula developed in previous banking crises in Chile, Japan and the United States savings and loan industry, bad assets of the failing bank are taken over by the government as a "bad bank", like the U.S. Resolution Trust Corporation, while the failing bank’s viable assets, deposits and branch network are taken over by the new owner, who can return additional loans to the "bad bank" if they produce no income. They demanded collateral with face value of 120% for PROER loans, but then accepted as guarantees value-impaired government obligations, known as moedas podres [rotten money], of which $75 billion are outstanding, bought at 40%-60% of face value on secondary markets. This financial engineering enabled the "bad bank" of the Nacional, which failed with a $5.4 billion hole in its balance sheet, to claim a profit after revaluing inflation-eroded housing mortgages, purchased at a 58% discount for $3.2 billion, at their face value of $7.8 billion, according to lawyers for the former owners. It also gave windfall profits to banks that had written off these government debts that never were paid or whose value was eroded by inflation. Financial results of the biggest banks for the first half of 1996 would have been worse without sale of some $5 billion in moedas podres sold to "bad banks" under Central Bank management, led by the $2 billion sold by Bradesco, and from risk-free interbank lending to fund Banespa, with the Bank of Brazil acting as intermediary. PROER gives the Central Bank new powers to intervene in banks and to hold auditors responsible for failing to report financial irregularities. Time will tell whether the Central Bank can improve bank supervision with these new powers.

The Central Bank as Scarecrow

Roberto Campos, Planning Minister in the 1960s, called the Central Bank of Brazil "The Monster That I Created. " The Central Bank now is struggling to rise above its historic role as a monetary accomplice to high inflation. In his brilliant memoir of public life over the past half-century, Lanterna na Popa [Lantern on the Stern] (1994), Campos argued that the Central Bank "strayed from its intended role as intransigent defender of the currency to finance Treasury deficits and to aggravate uncertainty in the financial market with a torrent of regulatory instructions and norms." According to Campos, the Central Bank’s "normative diarrhea" produced 4,692 regulatory changes from March 1985 to November 1992.
Managing a banking crisis on the scale of Brazil’s present difficulties would be a daunting task for much stronger central banks, such as the Federal Reserve, the Bank of England or the Bundesbank. That Brazil’s banking crisis so far has not led to a general financial panic or collapse is due to the courage and creativity of a small group of Central Bank officials who made mistakes but who managed to keep the system intact. The price of temporary salvation has been waste, injustice and problems for the future, but things could have been worse. The crisis of the big private banks may be reaching an end, but more failures of smaller ones are expected while the problems of the federal and state government banks remain unsolved and will demand firm and continuous policy over successive Presidential administrations.

The institutional problems of the Central Bank are key issues in efforts to stabilize the Brazilian economy. They are only beginning to be discussed in public debate. They are (1) political weakness; (2) excessive functions; (3) crippling distortions in its personnel structure, and (4) laxity in bank supervision.
1. Political weakness. In the rough and tumble of failed stabilization plans since 1985, the 14 Central Bank Presidents at the helm of Brazil’s financial system have stood like scarecrows in a stripped field whom the crows learn to ignore. They take political orders and always can be overruled by Brazil’s Presidents making deals with state governors and leaders of Congress. Under protest, Central Bank Presidents hand out money under threats of a general financial crisis and of the loss of jobs protected by leading politicians.
2. Excessive functions. The Central Bank generally has been able to meet short-term goals of monetary policy. When the government wants inflation in order to pay its bills, the Central Bank produces inflation. When the government wants tight money, the Central Bank produces tight money, even with suicidal interest rates. Its failure as a guardian of price stability and the integrity of the financial system is due both to political weakness and too many responsibilities in too many areas. Until the late 1980s, 32 development programs, such as lending to agriculture and small business, were operated by the Central Bank, financed almost entirely by printing money. Even today, its traditional Central Bank functions of managing currency emissions, buying and selling public debt and controlling foreign exchange operations compete with vast and neglected responsibilities for supervising 4,875 banks, funds, brokers and other financial firms, public and private.
3. Personnel structure. The Central Bank employs 6,200 people, 4,000 of them working in Brasília and the rest in branch offices in other large cities. According to Ibrahim Eris, who as Central Bank President in 1990-91 froze 40% of the financial system’s assets in one of the more spectacular stabilization efforts of recent years, "We had too many employees. A few were doing too much work. Too many were not doing enough. Our research department was weak. Pay was low. A director could earn four times as much in the private sector as at the Central Bank." No competitive examinations to recruit new staff were held from 1979 to 1993. Now the Central Bank’s pool of qualified people is shrinking as its aging staff retires in ever-greater numbers. However, many young recruits also are leaving the bank because entry pay is very low. During Eris’s Presidency, the number of bank inspectors was cut by 15%. Now only 632 inspectors in Brasília and the branch offices are directly involved in supervision of a complex and volatile financial system. Moreover, the time spent on different kinds of supervision is badly distributed. Only 0.5% of inspectors’ time is devoted to the CEF, holding 17% of the banking system’s assets. The huge problems of the Bank of Brazil absorb only 9% of their time, barely half the scrutiny given consórcios [privately owned pools for installment-buying of consumer durables, mainly cars], accounting for only 0.3% of the financial system’s assets.
4. Bank supervision. "Bank supervision is subject to as much political pressure as monetary policy," Central Bank President Gustavo Loyola recently said. "If the Central Bank loses its supervision responsibilities, oversight of banks might become even worse if an even weaker regulatory agency is created in the present institutional environment." Yet the scarecrow role of the Central Bank lost whatever credibility it previously enjoyed in the scandals of the Econômico and Nacional.
A "secret" report of the Federal Accounts Tribunal (TCU) on operations of Nacional and Econômico, leaked to the press, found that "the Central Bank for several years observed recurring irregular practices without taking effective steps to safeguard the investing public, capital markets, public banks and the Treasury from false accounting and massaged balance sheets." Internal documents examined by the TCU show that Central Bank inspectors repeatedly had reported to their superiors in Brasília that both banks were close to failure as far back as 1987. Instead of taking corrective measures, Central Bank officials chose to overlook or cover up these irregularities.
At the Econômico, "not once was anything done to regularize credits [in arrears] and accounting for unduly registered income," quoting Central Bank inspectors who in 1989 reported: "This picture is not new and is being painted with ever-darker colors over the years in the absence of corrective measures." According to the TCU: "More timely action by the Central Bank possibly would have avoided the extreme of lending the enormous sum of $6 billion to the Econômico after the bank failed." Central Bank officials taking over the Banco Econômico in August 1995 found in the personal safe of its boss and principal stockholder, Angelo Calmon de Sá —former Minister of Industry and President of the Bank of Brazil— a handwritten spread sheet listing $2.4 million in campaign contributions to 25 candidates in state and federal elections, $1.1 million of which went to Senator and former Governor Antônio Carlos Magalhães, the political boss of Bahia.
Banco Nacional was ranked sixth in size and second in profitability among the big banks in 1994. It listed profits of $140 million on assets of $11 billion, before Veja magazine revealed that the failing bank added $6.7 billion in fake assets to its loan book, the biggest bank fraud in world financial history. Red-faced Central Bank officials still are trying to explain how a big bank continuously could invent phony loans since 1986, spread among 652 false accounts in hundreds of branches. A former executive of KPMG-Peat Marwick, the accounting firm that audited Nacional’s accounts over the past two decades, explained that "most big frauds are done by top management. Computer access to these accounts were blocked by special codes. Brazilian banking laws were adopted in 1964, before computers were widely used. Central Bank supervision concentrates on whether the bank does its regulatory paperwork correctly, not on what is really happening inside the bank." After 107 inspections of the Nacional since 1987, Central Bank auditors reported in 1992: "It is ascertained [Pedro: Constáta-se] that, despite the country’s grave crisis, the Nacional manages to main the posture of a traditional commercial bank." In 1995, months after the Central Bank gave the Nacional a $5 billion emergency loan, its technicians reported: "The Banco Nacional is not a source of risk to the market."
In recent years, intense debate has taken place, both in Brazil and elsewhere, over whether central banking should be independent of government, so that the goal of monetary stability can be pursued without political interference, and whether bank supervision should be a central bank function or be given to an independent agency. However, the experience of many countries shows is these structural formalities are less important, in terms of growth and inflation, than the political will to achieve fiscal and institutional stability. Given continuing instability, it is hard to quarrel with the argument of Gustavo Loyola, the current Central Bank President, that "Central Bank independence only can be achieved when government accounts are organized. Otherwise any independent action by us would be crushed by a fiscal crisis. The Central Bank now is financing the Treasury with its own capital, which is one reason why we face the problem of decapitalization."
The Central Bank’s balance sheet is roughly $130 billion, against $80 billion each for CEF and the Bank of Brazil, the two biggest public banks, and $29 billion for Bradesco, the largest private bank. But the quality of the Central Bank’s assets are deteriorating rapidly. Its balance sheet is now loaded with bad loans absorbed to refloat failing private banks in exchange for marketable federal notes. Even excluding these bad assets, its net worth is now negative (-$2.5 billion) after losses of $13 billion from July 1994 to December 1995, thanks to mismatch of assets and liabilities.
Central Bank officials voice concern over mounting losses and accelerating decapitalization. Chile’s Central Bank also ran big losses from its rescue operations in the banking crisis of the early 1980s, making it harder to reduce interest rates and inflation, but was saved from greater difficulties by a tight fiscal policy and higher savings generated by private pension funds. Brazil’s political system refuses so far to provide this safety net.

Political weakness and excessive
function make the Central Bank
a servant of bankers and politicians.

What does Central Bank decapitalization mean? If decapitalized, a Central Bank will lack the credibility needed to carry out monetary policy and, specifically, will lack enough marketable assets to absorb excess money from the economy in order to contain inflation and to provide emergency liquidity to the financial system. To rescue banks and to avoid inflationary swelling of the money supply by the inflow of foreign exchange reserves, Brazil’s Central Bank borrowed local currency equal to roughly $53 billion on domestic financial markets since the Real Plan was launched, in addition to another $46 billion raised by the federal Treasury to finance its deficits. In other words, federal authorities have leveraged more than a year of their tax revenues ($90 billion) in short-term borrowings, accelerating in early 1996. At real interest rates of 15%-20% yearly, these borrowings become a powerful engine of monetary expansion to service maturing debt, a Ponzi game driven by ballooning interest. The game is made more dicey by mismatch between low interest earned on Brazil’s hard currency reserves and high interest paid on domestic public debt contracted to sterilize capital inflows. One critical difference with Mexico is that Brazil, in effect, has not one but three central banks. The Central Bank of Brazil is an outgrowth of the Bank of Brazil, which continues to carry out some central bank functions. The third central bank is the Caixa Econômica, which besides its banking activities runs six national lotteries and 11 special forced savings and social welfare funds with little transparency or supervision, all together moving volumes of money equal to two times the federal budget. The current banking crisis has bred division of labor between the three central banks. The Central Bank mainly rescues and merges private financial institutions and the Bank of Brazil and the CEF bail out state banks and governments. The three central banks are more like government agencies, disguised as banks, to distribute financial transfers. The weight of transfers in the Brazilian economy has become more focused under provisions of the 1988 Constitution, which Cardoso and Covas played leading roles in shaping as Senators from São Paulo. Transfer payments to states and municipalities have nearly doubled, although government lending was cut steeply. More than 1,000 new municipalities have been created to make localities eligible for these earmarked transfers, even though local taxes raise less than 10% of revenues of many local governments. Provisions for job stability and salary escalation, written into the new Constitution, doubled federal personnel expenses since 1988, driving efforts by Cardoso and Covas to rewrite the Constitution as President and Governor. Intelligent, dedicated officials have struggled hard over the years to reduce and modify these distortions. Progress has been made in reducing subsidies and clarifying public accounts. Yet new gimmicks are always appearing and their distortions have greater impact in a climate of lower inflation. In view of these institutional difficulties, there seems to be a need to return to simpler forms of central banking in which the Central Bank becomes an independent, specialized agency dedicated to managing the money supply and foreign exchange. Bank supervision should be assigned to another independent agency, as it is in many other countries. Institutional safeguards should be designed to protect the independence and integrity of these agencies. This is not an easy job, but there seems to be no other alternative.

The Real Plan Needs Deepening
The gain to society from the present banking crisis is to reveal the full measure of these institutional difficulties, bred by decades of chronic inflation, giving us a deeper understanding of the time and courage needed to achieve the goal of political and economic stability. A resolute effort to deal with these problems has its own rewards. It strengthens a government’s credibility. It also can buy time and develop a momentum of its own. Deals are being cut with state governments for new federal loans in exchange for shares in state banks and corporations that could be sold or liquidated later. In these deals there always is a danger that state governors again will default on their debts, leading to another cycle of renegotiation by their successors, and that employees of public corporations will be able to block privatization by mounting political mobilizations. More failures of this kind would be an enormous setback for Brazil. Credibility is gained by persistence. This test of an anti-inflation consensus is still to be met.
The Real Plan has not failed. What it needs now is a coherent and courageous deepening. The immediate issue is the survival of state capitalism with its costly and entrenched privileges. The challenge gains immediacy from the electoral cycle, the impact of compound interest, weakening trade figures, inflationary pressures on an overvalued currency and President Fernando Henrique Cardoso’s confused priorities, especially his attempt to secure reelection. So far he has won respect and affection because of his decency of character, because of his eloquence in favor of market-oriented modernization and social justice and because he has given Brazil an honest government which is a hopeful change from the scandal-ridden gangs of the recent past. Cardoso repeatedly has reminded Brazilians that he is no savior and that the changes needed to consolidate democracy and stability can only be achieved over several years by successive administrations. In seeking these changes, however, he so far has relied too much on sweet talk and personal charm and has been timid in dealing with the politicians from whom he must win concessions to consolidate the regime of low inflation needed to sustain economic growth, to rebuild public institutions and revive public investment, which Brazil needs to strengthen its capacity to operate a complex society on a continental scale.

The first phase of the Real Plan was relatively easy because decisions were made by a small group of technicians who freed prices and imports, supported an overvalued new currency with big reserves and high interest rates and promised fiscal reform. We now face the second phase of the Real Plan. Hard choices must be made by thousands of people and backed by the whole society and promises must be kept to stop fiscal leakage and sustain credibility. Brazilians still need convincing that these goals must and can be achieved. But Cardoso’s efforts to win support from the political community to achieve fiscal balance and abolish chronic inflation have been undermined by five big mistakes made in his first 18 months in office:
1. Even though he served as Finance Minister for a year before launching the Real Plan, he failed to use the three months between his election and inauguration to prepare specific proposals to Congress for fiscal balance and consolidation. When these proposals finally were presented, months later, they were weaker than needed and the political momentum gained from his landslide election victory was receding.
2. Further political initiative was lost by intensive Presidential travel abroad during Cardoso’s first 15 months in office, distracting his own and public attention from his legislative program and leaving more room for maneuver to its opponents in Congress.
3. Instead of creating new opportunities for federalization of state banks and continuous rollover of state debts, Cardoso should have consolidated the fiscal and financial framework of Brazilian federalism by offering to help to state governors only if they first privatize or liquidate state banks. The World Bank is lending money to regional governments in Argentina and Brazil for these purposes.
4. Cardoso’s premature maneuvers to change the Constitution so he can seek reelection in 1998, which started shortly after his inauguration in January 1995, have generated an extra focus of disturbance in Brazilian politics and harmed chances for reform in his first term. Presidents Carlos Menem of Argentina and Alberto Fujimori of Peru won constitutional changes of this kind in recent years only after carrying out painful stabilization programs that continue today, each based on an anti-inflation consensus much stronger than Brazil’s. Cardoso has not passed this test yet and has far to go.
5. In democracies throughout the world, national governments tend to lose mid-term elections, especially municipal elections in which local issues and personalities predominate. Instead of standing aside from Brazil’s October 1996 municipal elections and focusing on national priorities, Cardoso belatedly plunged into the campaign for mayor of São Paulo, already contested by three strong candidates, to stake the prestige of his government and his party in the candidacy of his Planning Minister, José Serra, who for several months had disclaimed interest in running. Serra now stands fourth in the polls with only 9% of intended votes. The outcome of this election is likely to weaken Cardoso politically for the rest of his four-year term and may lead to a loss of support from other parties.
The Real Plan, if it is to progress and ultimately succeed, cannot remain the electoral property of a single politician. The opportunity and the responsibility for achieving political and economic stability must be shared. Brazil’s politics is being further confused and agitated by the prospect of Cardoso bargaining away the fiscal austerity demanded by the Real Plan to gain the support of state governors and Congress to change the 1988 Constitution to make him eligible for reelection. This kind of bargaining would evoke memories of the concessions made by President José Sarney (1985-90) during the Constituent Assembly to secure a fifth year for his term of office. Instead, Cardoso could bargain away his fading reelection option for firm and broad support for a deepening of the Real Plan that would guarantee him a revered place in Brazilian history and heighten prospects for an acceleration of economic growth and political development. In effect, Cardoso could say: "I will seek reelection only if Congress fails to pass measures to secure the viability and continuity of the Real Plan. If Congress passes these measures and other candidates agree to a binding consensus on future reforms, my candidacy for a second term will not be necessary."
In the recent experience of the few democracies allowing Presidential reelection, the second term is almost always a disappointment. This is as true in the United States in the second terms of Eisenhower, Reagan and Nixon as in Venezuela after the return to power of Carlos Andrés Perez and Rafael Caldera. A more lasting triumph would be a broader consolidation of responsible government which the Brazilian people have sought, and only partially achieved, in the elections of 1989, 1994 and 1996. Responsible government will come to stay only if people keep demanding it and if they do not lose hope.


	THE INSTITUTE \| WHO IS WHO? \| PUBLICATIONS \| RESEARCHERS \| GUEST BOOK \| PORTUGUÊS