Stress Test Read online

  MEXICO HAD been hailed as a model for emerging markets, a fast-growing destination for foreign capital, the first new country invited to join the Organization for Economic Cooperation and Development in more than two decades. But during the country’s boom, the Mexican government became far too dependent on short-term borrowing—the kind of money that can easily run when confidence is shaken. Mexico also had a fixed exchange rate, pegging its peso to the dollar—a recipe for instability when confidence goes.

  Sure enough, when insurgents assassinated Mexico’s leading presidential candidate early in 1994, investors and bondholders and other creditors started to fear the country wasn’t as stable—and their money wasn’t as safe—as they had believed. Runnable capital began to run. Confidence in the peso began to evaporate. The government tried to buy pesos to prop up their value and defend the peg. But that just drained its dollar reserves and heightened fears that it would default on its debts, especially a pile of short-term bonds called tesobonos that were linked to the dollar. After taking office in December, Mexico’s new president, a Yale-trained economist named Ernesto Zedillo, faced reality and abandoned the unsustainable fixed exchange rate. But as the peso plummeted in value, so did confidence in the country.

  By the end of 1994, Mexico had clearly lost control of its finances. The government had only $6 billion left in reserves, with $30 billion worth of tesobonos coming due over the next year. And markets no longer considered it creditworthy, so it couldn’t raise money to pay its bills. I remember Jeff Shafer, an assistant Treasury secretary, suddenly announcing he had figured out Mexico’s problem: “It’s a sovereign liquidity crisis!” In other words, it was a run on the country, a national version of the rush to George Bailey’s bank in It’s a Wonderful Life. Mexico had a modest long-term debt burden and the power to tax, so in theory, it should have been able to pay what it owed over time. It did not have a fundamental solvency problem; it was by no means a hopeless case. But it had an immediate liquidity problem. Without cash on hand, it couldn’t meet its obligations. Like George Bailey, it needed help in a hurry.

  Default would have been a nightmare for ninety million Mexicans, a potential prelude to hyperinflation and mass unemployment. It also would have been a problem for us. Mexico was our third largest trading partner, and the Fed staff calculated that a messy crisis could affect hundreds of thousands of American jobs, reduce U.S. growth by an entire percentage point, and increase illegal immigration 30 percent. We also feared that investors unnerved by Mexico might abandon other emerging economies that seemed to have similar vulnerabilities. Brazil and Argentina were already experiencing this “Tequila Effect,” as their markets slumped in sympathy with Mexico’s. Finally, we knew that if Mexico cratered a year after the North American Free Trade Agreement eased its barriers to foreign capital, protectionists at home and abroad would claim a propaganda victory. It would build momentum for trade restrictions in Congress, while encouraging developing nations to wall themselves off from the world.

  Larry recognized that a typical International Monetary Fund loan, which would be limited to Mexico’s “quota” of $2.6 billion, would be woefully inadequate to stop the run. It would be up to the United States to fill the gap. He suggested that Colin Powell’s doctrine for U.S. military intervention—deploy overwhelming force, but only when American interests were at stake, and only with a clear exit strategy—should also apply to U.S. financial intervention. At an early meeting to discuss options, Greenspan suggested that $20 billion would be a “wall of money” large enough to overwhelm the tesobonos and reassure the markets, a figure we ultimately decided to double. The chairman was a free enterprise Republican, reluctant to meddle in markets, concerned that rescuing Mexico (and its bondholders) would embolden future Mexicos (and future creditors) to take similarly irresponsible risks. But we all agreed the potential moral hazard cost of a bailout paled in comparison to the actual cost of default. It was, as Greenspan said, “the least-worst option.”

  We called that plan Mexico One. But after the politics soured in Congress, we needed a Mexico Two. That’s when we turned to the Exchange Stabilization Fund, a pot of money that Treasury was authorized to use to reduce volatility in currency markets and promote financial stability. It had never been used on a scale like this, and I thought it would be imprudent to commit the bulk of our foreign exchange reserves to this cause. But Ted Truman, who ran the international part of the Fed in Washington, figured out a way for the Fed to help us make $20 billion available for loans to Mexico and still preserve some firepower for other contingencies. It wasn’t the $40 billion we had requested from Congress, but it was the only way we could act without legislation. And IMF director Michel Camdessus pledged $18 billion, by far the largest package in the fund’s history. So we were pretty close to Larry’s Powell Doctrine goal.

  Several of our European allies, especially the Germans, were furious with the IMF’s commitment—partly because of moral hazard fears, partly because they felt inadequately consulted. Members of Congress were also furious about our use of the ESF—again, partly because of substantive concerns about putting taxpayer money at risk to save Wall Street speculators and a reckless neighbor, partly because we had just authorized the largest U.S. aid package since the Marshall Plan without their approval. And we were still negotiating terms with the Mexican government, so Mexico Two was not a done deal.

  We thought the deal had to include some tough conditions, including credible government commitments to get its finances under control and to raise interest rates high enough to keep private money from fleeing the country. Ultimately, the rescue wouldn’t work unless Mexico’s leaders proved they were worthy of investor confidence, and as Larry liked to say, we couldn’t want reform more than they did. But we didn’t want to impose conditions so punitive that they would weigh down the Mexican economy and depress confidence even further—or force Mexican leaders to resist to prove they weren’t helpless supplicants to the United States.

  It was a delicate tightrope, and investors—uncertain about the Mexican government’s appetite for reform, and rattled by the political opposition in the United States—were skeptical that we’d make it across. Capitol Hill leaders were pushing to block us from using the ESF, or at least tie our hands with restrictions governing everything from Mexican labor standards to Florida tomato exports. Internally, even after we signed the deal in late February, Rubin kept playing devil’s advocate, asking us to persuade him we weren’t dumping tax dollars into a lost cause. News of those discussions seeped out of the Treasury building, fueling rumors that we were reconsidering the loans, which further unnerved the markets.

  Rubin liked to say that nothing in life was certain, and none of us felt highly confident of success. Every blip of bad news—a rebel advance in Chiapas, a drop in the peso’s value—made us fear for the program. But we went ahead with our rescue plan. And President Zedillo and his team of technocrats kept their promises to raise interest rates while imposing tax hikes and budget cuts, which helped persuade investors they were committed to getting Mexico’s finances under control. After a few months of lurches, markets stopped running.

  By the end of 1995, capital was trickling back into the country. By 1997, Mexico’s economic output had returned to pre-crisis levels, and its government had repaid all its loans early, netting U.S. taxpayers $1.4 billion in interest. Markets in South America, Asia, and eastern Europe that had suffered from Mexico comparisons all rallied after Mexico stabilized. The rescue worked. By that time, though, Mexico had fallen out of the news, and neither the success of the rescue nor the fact that we recovered our investments plus a profit got much attention—certainly not enough to offset the political hit that Clinton and Rubin took around the initial decision.

  Mexico was a bracing lesson in the terrible politics of crisis response. I had never worked on something so controversial before, and it was searing to watch the abuse showered on Rubin and Greenspan—then near the peak of their public credibility—for
taking a risk that seemed so compelling to me. It was instructive to contrast Clinton, whose support for the Mexican rescue never wavered even though his aides warned it could make him a one-term president, with congressional leaders who were with us until the public was against us. Senate Banking Committee Chairman Al D’Amato, a Republican from New York, actually urged us to expand the Mexico package before becoming one of its most vehement critics. That kind of congressional opportunism made it harder to restore confidence in Mexico, because it damaged confidence in our ability to keep our commitments.

  In fact, once the crisis was over, Senator D’Amato pushed legislation through Congress that temporarily restricted our ability to use the ESF to fight future crises. And we knew there would be future crises. Globalization had unleashed enormous sums of “hot money” that could instantaneously flow across borders, while the aspects of human psychology that had helped produce financial booms and crises for centuries remained unchanged.

  WE DIDN’T think the global community was ready to navigate this perilous new world of mobile capital. My colleagues at Treasury and the Fed believed the IMF needed more money for when the next Mexico exploded, plus the capability to deploy the money quickly and forcefully enough to contain a run. It had to be able to provide countries in crisis with sufficient cash to overwhelm the crisis as well as tough conditions to restore confidence in the markets.

  But while the United States was the most powerful force at the IMF, we weren’t a controlling force. The U.S. founded the IMF, and it was still based in Washington, but we provided only about a quarter of its total funding. As Camdessus once observed, we could block things, but we couldn’t make things happen unless we persuaded our partners that they made sense. Even before Mexico was resolved, we began a major international effort to create a stronger global architecture for dealing with future financial crises.

  Part of our work was promoting preventive medicine to help countries avoid becoming the next Mexico. We helped forge consensus in international communiqués for what we saw as best practices, urging countries to avoid excessive short-term debt in foreign currencies, to let their own currencies float freely, and to make sure their banks had plenty of capital to cushion against sudden losses. But those recommendations were purely voluntary. We couldn’t force sovereign nations to follow them, and our ideas didn’t get a lot of traction in those days.

  Alongside our crisis prevention efforts, we also worked to improve our crisis response options for the next Mexico. I came up with the idea of a new $50 billion IMF reserve fund, which seemed like a lot of money at the time. We wanted to make sure that in future crises, the world wouldn’t be dependent on the United States as the dominant funding source, especially now that the D’Amato restrictions limited our ability to offer bilateral loans. We also proposed an interesting design for the new fund. Instead of raising the money exclusively from the traditional group of advanced economies, we proposed that emerging markets should help finance it and help govern it. This was partly to reflect the new global balance of power; the rising Asian and South American economies deserved a more influential presence alongside rich establishment nations at the IMF. But it was partly to dilute the power of more conservative European countries; we didn’t want their occasional parochialism and moral hazard fundamentalism to paralyze future crisis responses.

  I flew around the world to make the case and negotiate the arcane details, often with Ted Truman from the Fed. I liked the challenge of the substance and the diplomacy—long flights to long meetings in windowless rooms, sometimes back and forth across the Atlantic without spending the night. I once did five countries in Asia in five days. I built relationships with a new generation of finance ministry and central bank officials. It was interesting, sometimes even exciting.

  Many foreign officials were initially skeptical of our proposal for a new crisis fund, thinking we were just trying to find a way to deploy other people’s money to finance our own interests. Some of my European counterparts called me “the smiling hegemon.” And it’s true that I tried to maximize U.S. influence. I remember asking Truman and another Fed economist, Lew Alexander, if we could somehow structure the fund to give the United States the power to force action as well as veto action. They laughed and said they didn’t know any math that would give us all the power for just 25 percent of the funding. We settled for a veto. And our counterparts eventually realized the fund made sense for everyone.

  By 1997, we had the framework of a deal. But before Congress would authorize this new arsenal for attacking the next crisis, the next crisis had arrived.

  THAILAND HAD ignored the IMF’s warnings about the dangers of fixed exchange rates and short-term borrowing in foreign currency. So had many of its fellow “Asian tigers.” That didn’t matter until their economies stopped growing rapidly. Then it mattered a lot.

  Throughout the nineties, Thailand’s banks enjoyed easy access to dollars and yen, which they used to finance an investment boom, much of it in real estate. But a lot of the investment was not productive. When the bubble popped in 1997, the banks were overloaded with nonperforming long-term loans, and their creditors cut back their short-term access to dollars and yen. Confidence in the Thai baht flagged. Instead of letting it adjust, Thai leaders followed Mexico’s bad example, draining their foreign exchange reserves to defend an indefensible peg to the dollar, hoping the storm would blow over. It didn’t. In July 1997, they gave up and devalued the baht. Panic ensued. Thailand was less connected to the United States than Mexico, but we knew there was a chance its crisis could drag down other Asian economies. And Asia was an increasingly vital part of the world economy.

  Thailand was also less prepared for financial shocks than Mexico, which had experienced repeated crises in the past, and had attracted a lot of financial talent to serve in government. I remember calling a senior Thai finance ministry official from my parents’ house on the Cape early in the crisis to ask what was going on. He didn’t seem to know much, and didn’t seem to want to share what he did know. The Thais were so reticent it was hard to get a sense of what they were thinking, and none of us really understood their country. My colleagues sometimes assumed I’d have a feel for the place after spending my high school years there; Larry teasingly called me “Mr. Asia.” But my time in Thailand gave me no relevant insight into the country’s crisis—empathy, maybe, but no additional knowledge that could help us help the Thais.

  Our inclination at Treasury—and the IMF’s inclination, too—was to try to replicate what had worked in Mexico. We hoped to put a lot of “money in the window,” enough to look big compared to the liabilities that could run. We would loan the money at a fairly expensive rate, to make sure it would be repaid as soon as possible once the crisis passed. And we would attach other conditions designed to prevent Thailand from repeating mistakes that had led to the crisis, with the goal of restoring investor confidence in the country.

  But in the Thai crisis, unlike the Mexico crisis, the United States couldn’t take the financial lead, so we wouldn’t be calling the shots. Senator D’Amato’s restrictions on the ESF blocked us from providing the large long-term loans we thought were needed. That left the IMF as the only large-scale source of finance. At a meeting of finance officials in Tokyo that August, after Japan pledged to lend $4 billion to the Thais, I had to explain that the United States could not make a direct commitment, even though our economy was in stronger financial shape than Japan’s. “How does it feel to be a superpower?” I said to my Japanese counterpart. It was a deeply uncomfortable situation for me—and, I thought, for the United States.

  Despite resistance on its board, the IMF leadership committed about $4 billion and cobbled together another $13 billion worth of other commitments. But the package didn’t look as generous as Mexico’s, and the Thais felt betrayed that none of it came directly from us. Other Asian countries were offended, too, and the Japanese tried to exploit our perceived weakness, quietly floating the idea of an Asian Monetary Fund tha
t would supplant the IMF’s role in Asia. We thought this was a bad idea for the global financial system and for Asia; a regional fund model would be more susceptible to being overwhelmed by a regional crisis. I warned Bob and Larry that we were suffering huge damage to our credibility in Asia. Even with one hundred thousand troops stationed on the continent, our influence was waning.

  We risked making the problem worse with a fight about transparency. The Thais were publicly claiming they still had $20 billion in foreign exchange reserves, but we knew the real number was closer to zero; the Thai central bank had sold its dollars in the forward markets to conceal the depth of its problems. Chairman Greenspan felt strongly that as a condition of any IMF assistance, Thailand should have to reveal the truth. Bob and Larry agreed. I expressed doubts. I thought full disclosure could shatter confidence and accelerate the run.

  I was wrong. Allowing the Thai government to withhold information might have avoided some near-term pain, but it would have risked a lot of damage to their credibility and the IMF’s when the truth came out. In a financial crisis, uncertainty is the enemy of confidence. The markets didn’t trust the Thai numbers anyway, and the absence of reliable information already encouraged investors to assume the worst. At the time, though, it looked like my fears were coming true, like America was messing up an intervention it wasn’t even funding. After the IMF announced the loan in late August—and revealed Thailand’s lack of reserves—the baht resumed its swan dive and capital continued to flee.

  That fall, I was promoted to assistant Treasury secretary for international affairs, my first political appointment after a decade as a civil servant. At Larry’s suggestion, I had switched my party registration from Republican to unaffiliated, to make it easier to get the White House on board. I had voted for President Clinton twice, anyway. But I was too busy worrying about Asia to savor my promotion; Rubin had to administer my oath of office in a hotel in Hong Kong. There was a lot of excitement in those days, but what I mostly remember was a constant feeling of dread. The crisis was spreading to Indonesia, Malaysia, and Korea, as pressure built on their fixed exchange rates. Markets in Brazil, Argentina, and Mexico were falling in sympathy with Asia’s, as investors tried to get ahead of the spreading contagion. The United States was enjoying strong growth, wages were rising, and the federal budget was on the verge of its first surpluses in decades, but the Asian crisis still hit home in late October, when the New York Stock Exchange had to suspend trading after a sudden 7 percent drop in the Dow.