The Chastening

by Paul Blustein

  Lopes scorned this approach as an exercise in macho central banking. Borrowing costs were already very high, he said, and if rate hikes proved necessary, it would make more sense to do them gradually. President Cardoso had supported rate increases in the past, he added, but the central bank needed to be able to explain the reasons to the president rather than tighten credit arbitrarily. In any case, Lopes continued, he couldn’t simply order a rate increase on Monday morning; such a step required a meeting and formal approval of the central bank’s monetary policy board. Okay, Camdessus countered, then how about calling the members of the committee on the phone immediately to obtain their assent? Lopes refused.

  Lopes departed in a rush to catch his plane back to Brasília on Sunday, January 17, leaving behind an unhappy group of IMF officials. Since Brazil was dropping its band-changing scheme for the real and replacing it with a float, the Fund was no longer threatening to cut off the country. But Lopes’s unwillingness to tighten money dramatically, the Fund believed, meant that the real would surely depreciate even further—and that, IMF economists feared, would decimate other Latin American currencies as well.

  A nightmarish two weeks ensued, during which the real followed a path similar to that blazed almost exactly one year earlier by the Indonesian rupiah.

  When the new floating policy was announced on Monday, January 19, the real dropped to 1.59 per dollar, 31 percent below its level before the devaluation, and it continued falling vertiginously, dipping below two reais per dollar by the end of January. Most disturbing, the dumping of reais continued apace even on days when Brazil’s Congress was approving politically controversial deficitreduction measures, and even on days when Lopes’s central bank board raised interest rates, which it did in a series of steps that brought the overnight level to 39 percent by the beginning of February. To the IMF’s way of thinking, of course, the problem was that Lopes was tightening credit too slowly and timidly. Whether this explains the market’s behavior or not remains a matter of dispute. But one fact was glaringly evident: With every downtick in the real, and every uptick in interest rates, Brazil’s predicament was becoming more insurmountable. A lower value for the real increased the government’s burden of servicing its debt, because the principal and interest payments on some of its bonds were linked to the dollar. And higher interest rates caused the debt-repayment cost to swell as well, by more than $1 billion a year for each extra 1 percent in rates.

  Rumors flew that the government was considering imposing exchange controls or a moratorium on debt payments—and capital continued to exit the country at about a half billion dollars a day. “Cardoso is under massive political pressure from his party, from already rebellious governors and from the public to switch policies quickly if this last gasp at conventional policy fails,” Medley Global Advisors, a New York firm providing information and advice to institutional investors, told its clients in its daily newsletter on January 24. “That would mean moving to Plan C—a new cabinet, perhaps capital controls and lower interest rates, perhaps a return to inflationary times.” On January 29, ordinary citizens became swept up in the panic as word spread that Cardoso would freeze bank accounts, as one of his predecessors had done a decade before. Queues of anxious depositors formed at banks. Brazil was at the precipice.

  The moon was full, imparting luminosity to the snow coating the Alps as Stan Fischer left Davos, Switzerland, at 3 A.M. on February 1, 1999. A long day lay ahead for the IMF deputy managing director, who was on his way to Brazil after attending the Global Economic Forum, an annual gathering of policymakers, tycoons, and assorted movers and shakers. Davos lies about 3 hours by car from Zurich’s airport; from there Fischer would fly to London, and from there to São Paulo, and from there to Brasília, where he would land late in the evening. He was going at the request of the Brazilian government but also at the urging of Ter-Minassian, who wanted to be sure the government would be convinced to take the steps necessary to pull the country out of its financial nosedive. “You should come,” she had told him, “to deliver the message at the highest level.”

  Electrifying news came shortly after Fischer’s arrival in Brasília: Lopes was going to lose his Job running the central bank after only nineteen days in the post, and President Cardoso was replacing him with Arminio Fraga, a top executive with George Soros’s hedge-fund firm. The decision was the result of a high-stakes maneuver by Malan, who had submitted a letter to President Cardoso suggesting that both he and Lopes should resign. The move forced Cardoso into deciding whether he was truly prepared to Jettison economic orthodoxy, and his answer was no. Instead of replacing both men, the president fired Lopes and asked Malan to stay on as finance minister.

  Fischer holed up at his hotel on February 2, the morning after his arrival, waiting for the public announcement of the switch at the central bank. Going to the Finance Ministry before that, as he had originally planned, would give the impression that the Fund had orchestrated or demanded the personnel changes. Although suspicion naturally arose that the IMF was indeed behind Lopes’s sacking, Fund officials have always insisted they would never interfere so brazenly in a member country’s affairs. But they had made it clear in their conversations with the Brazilians that they didn’t like the central bank’s policies under Lopes, so they were naturally relieved to see him go, and they regarded Fraga’s nomination as a courageous masterstroke by Cardoso. The forty-two-year-old Fraga, a Princeton economics Ph.D. and former director of international affairs at the central bank, had worked for Soros Fund Management since 1993 as manager of its $1.4 billion emerging-markets fund. His appointment created an easy target for opposition politicians, who accused the administration of handing the country’s fortunes over to the very speculators responsible for the country’s woes. But the IMF viewed Fraga’s qualifications as almost ideal, since his Soros experience provided him with considerable insight into how to instill confidence in markets. After his nomination was unveiled on February 2, the real Jumped more than 8 percent.

  Once the Fraga announcement was out of the way, Fischer met with Malan and Fraga at the Finance Ministry around noon and the following day with Cardoso and Malan at the presidential residence. Brazil’s IMF program was in need of a major overhaul—or “strengthening,” in Fund parlance—for the depreciation of the real had made a hash of the first one. Strengthening the program had worked in Korea, where the accord struck on Christmas Eve 1997 saved the country from default. But the Indonesian case had shown that program strengthenings sometimes produce quite the opposite sorts of results.

  To induce the Brazilians to act, the Fund could dangle the $9 billion second tranche that was pending under its $41.5 billion international loan package. At the rate the country was losing dollars, that injection of hard currency would provide a most welcome boost. But Brazil would have to reciprocate. “You’ve only got one chance,” Fischer exhorted Cardoso. “There has got to be a stronger program.”

  The government would have to cut the budget deficit even more deeply than previously planned, Fischer said. The country’s debt dynamics were out of control, with bigger deficits and higher interest rates mutually exacerbating each other, creating an ever-rising debt with ever-rising servicing costs. One way to try changing the dynamic, of course, was for the central bank to cut interest rates and print the money to pay off the debt. But in the IMF’s view—and Malan’s and Fraga’s as well—that route led back to Brazil’s bad old hyperinflationary ways. Instead, Fischer said, the government would have to cut spending and raise taxes sufficiently to convince the markets of its resolve to contain the problem. Brazil had already committed to generate a budget surplus (not including interest payments) equal to 2.6 percent of GDP in 1999 and more in 2000 and 2001; Fischer wanted those targets to be raised substantially. If the government acted convincingly enough, he said, the debt dynamic might turn virtuous very quickly, because interest rates on much of its borrowing were tied to the overnight interest rate. But making the dynamic virtuous would require ano
ther bite on the bullet.

  The economy’s weakness was not conducive to such austerity. Joblessness in São Paulo had soared to around 20 percent, and nearly all forecasters anticipated a recession of somewhere between -2 percent growth and -5 percent growth for the year. But if the government didn’t move decisively, Fischer warned, all the stabilization that Cardoso had accomplished since 1994 would be in danger.

  Belt-tightening wasn’t the only thing Fischer had on his mind for Brazil. Another crucial component of a strengthened program, he told Cardoso, was that the private sector had to become “involved”—to wit, Brazil’s foreign bank creditors would have to reach some sort of arrangement to keep their money in the country, as Korea’s had done, and as the Europeans had been demanding.

  Foreign banks had cut nearly one-third of the $60 billion in short-term interbank and trade lines that they had extended to Brazil as of August 1998, and although that move didn’t constitute a run as devastating as the one that had triggered the Korean crisis, it had the potential to develop into one. So the Brazilians could no longer resist bailing in their banks as they had in autumn 1998. The G-7, including the United States, would not agree to pour more IMF money into the country without some assurances against the funds going to pay off foreign bankers, Fischer told his Brazilian interlocutors.

  More than a month passed before all the negotiations were finished and the loose ends tied up. But once things started coming together for Brazil, they came together quickly—and this time, the IMF was taken by surprise because of how positive the outcome was, instead of the reverse.

  Thursday, March 4 was Fraga’s first day as president of the central bank after being confirmed by the Senate, and he promptly drove short-term interest rates up from 39 percent to 45 percent in a display of determination to contain inflation and lure investors into reais. The Brazilian currency had fallen to an all-time low of 2.22 per dollar on March 2, but following Fraga’s move, the real appreciated to 2.06 per dollar, and it continued strengthening over the next several weeks, hitting 1.66 reais per dollar on April 13. For the IMF, this episode was sweet vindication of its faith in using high interest rates to stabilize a currency.

  On March 5, a Fund mission in Brasília completed negotiations on the terms under which the government would receive the next $9 billion from its international loan package. The Brazilians agreed to additional budget measures, as Fischer had urged, with the goal of increasing the government’s budget surplus targets by about 0.5 percent of GDP. They also pledged to keep monetary policy tight, the aim being to tamp down quickly the inflationary impact of the devaluation.

  One final element remained to be completed—bailing in the foreign banks. On March 8, Camdessus and Rubin suggested in separate statements that providing additional official funds to Brazil depended on a satisfactory arrangement in which the banks would undertake to maintain their credit lines. “The Brazilian authorities intend to seek, starting this week, voluntary commitments by their creditor banks.... This effort ... will be a key factor in the consideration of the program by the Executive Board of the IMF in late March or early April,” Camdessus stated.

  The New York Fed hosted a breakfast meeting of nine American banks and two Canadian banks on March 11. After considerable internal debate, U.S. officials had decided to take a relatively light touch in concerting Brazil’s creditor banks. Some within the High Command had wanted a Korea-style operation, with officials in each country prodding bank executives into signing individual commitments to keep their money in Brazil; these included the Germans, IMF officials, and one or two policymakers at the U.S. Treasury. But others, including New York Fed officials, argued that browbeating the banks might prove counterproductive and wasn’t necessary anyway. The banks regarded Brazil as a market with vast potential; inducing them to maintain their credit lines might require only a little encouragement and the tacit assurance that they were Joining in a collective effort.

  Terry Checki, executive vice president of the New York Fed, chaired the meeting, and Bill Rhodes of Citibank, who had served as coordinator for the banks during the Korean bail-in, was tapped to do the same for Brazil. As in the case of Long-Term Capital, New York Fed officials applied no explicit pressure on the banks—indeed, Malan, who was the featured speaker, emphasized that Brazil was asking for their voluntary cooperation—but the location of the meeting carried its own unspoken message. The light touch worked; the banks agreed to a Joint statement that “signaled their intention to maintain” their credit lines to Brazil, at least until August 31, 1999. In Washington, Fischer insisted that the statement be made public, so he could show the evidence of the bail-in to the IMF Executive Board before approval of the next loan tranche. Following similar meetings of bankers over the next couple of weeks at central banks in London, Paris, and other financial capitals, the Fund’s board issued its approval on March 30.

  With that new lease on life, the Brazilian economy proceeded to make fools of those in the economic forecasting profession. The consensus among economists in early 1999 had been that the collapse of the real would cause Brazil to undergo not only a deep recession but a bout of inflation of between 40 percent and 80 percent. Instead, the economy eked out growth of about 1 percent in 1999, and prices rose at less than 9 percent. (The IMF’s Mussa, who had made so many right calls during the crisis, was among those whose pessimism about Brazil proved way off the mark.)

  The IMF deserves considerable credit for helping Brazil recover in spring 1999 from the collapse of the real. The Fund’s prescriptions were apt, and its support—combined with the bank bail-in—produced the desired effect on confidence. A particular source of pride for Fund officials was the speed with which Brazil was able to lower interest rates following the initial upward spike in early March that stabilized the currency. In a textbook example of IMF economics working according to plan, the Fraga-run central bank cut the overnight borrowing rate to 34 percent in mid-April, 11 points below the March peak, and to 23.5 percent on May 19.

  But the Fund’s credibility suffered a grievous blow when its first Brazil program failed to achieve its declared aim of sustaining the country’s currency peg. Later in 1999, IMF and Treasury officials publicly declared that the international community should be chary of bailing out countries with fixed or crawling exchange-rate systems like the one Brazil had. Though they left open the possibility of exceptions, they said that rescues should be confined generally to countries with pure floats or rock-solid pegs like a currency board.

  This does not mean that Fischer, Summers, and their colleagues acknowledge having erred by approving the late-1998 program for Brazil. On the contrary, they believe it may have helped save the global economy. Their reasoning is that since the devaluation of the real was averted until January 1999, global markets were spared the trauma in autumn 1998, when the impact might have been too great to withstand. “You didn’t want a devaluation then [in the autumn],” said Fischer. “The world was highly unstable. You didn’t want to throw another firecracker into the mix.”

  Sensible though this Justification may be in hindsight, the IMF and the Treasury can take little credit for concocting a masterfully timed strategy. No one recalls hearing the architects of the Brazilian rescue suggesting in autumn 1998 that the point was to defer devaluation for a few months. Perhaps the best way of summarizing the outcome is that the Treasury and the IMF stumbled on the right solution.

  Stan Fischer could tell by his personal schedule that a change for the better was afoot in April 1999. Weekend strategy meetings at the IMF had suddenly become a thing of the past. He was able to finish his taxes more than two weeks before the April 15 deadline; the previous year, he had barely filed on time because of the crisis in Indonesia. His optional reading material at home was finally getting some attention.

  Practically every day, the world’s most economically troubled countries were exhibiting new signs of health. Hong Kong’s stock market hit an eighteen-month high on April 16, and earlier that week
, South Korea announced it would make early repayments on its IMF loans, as the Korean stock market soared above its precrisis level. Brazil Joined a host of nations, including Colombia, Hungary, and Argentina, in announcing plans to raise billions of dollars on the international bond market.

  The Brazilian crisis had not produced the worldwide contagion that Fischer and others in the High Command had feared. (It would later be blamed for contributing to Argentina’s 2001-2002 crisis, but that is another story.) Nobody could be certain why the contagion was so limited. The most logical explanation was that hedge funds and other international money managers had been forced, at least temporarily, to stop borrowing to the hilt as they had done formerly, which meant that losses in one market no longer required them to sell their holdings in others.

  In the same unpredictable manner with which it first materialized and spread, the global financial crisis was ending, leaving perplexing questions about how to stop the next one.

  13

  COOLING OFF

  The table in the Treasury secretary’s private dining room was set with a breakfast of fresh fruit and toasted bagels as a small group of reporters and editors from The Washington Post filed in to interview Paul O’Neill, the new secretary, in early February 2001, a couple of weeks after the inauguration of President George W. Bush. Eager to get a close-up look at the feisty O’Neill, who had previously headed Alcoa Inc., the Journalists opened by asking him how he might handle international financial crises differently from his Clinton administration predecessors. With the tart tongue for which he would become renowned, O’Neill conveyed his conviction that the crises of the late 1990s could have been anticipated and managed much better than they were. Deriding the 1998 Russian rescue as particularly ill-advised, he said: “What happened in Russia was foretold. It was not a surprise. You ever try to do business in Russia? You ever try to write an enforceable contract? It doesn’t take a genius to figure out it’s not a great place to put capital.” When the Indonesian crisis was mentioned, he demanded, “Was that a surprise?” In response to the assertion that Indonesia’s collapse had indeed come as a terrible shock, at least to the IMF and World Bank, he shot back: “To the world? They didn’t have internal structural problems?”