Stress Test

  On March 30, the President announced that we would provide additional TARP loans to keep GM and Chrysler afloat, but with aggressive deadlines for them to propose viable restructuring plans or else face government-facilitated bankruptcies. “Year after year, decade after decade, we’ve seen problems papered over and tough choices kicked down the road, even as foreign competitors outpaced us,” he said. “Well, we’ve reached the end of that road.” By the end of the spring, both firms would be in bankruptcy, and Team Auto would be restructuring them.

  Once again, we hit the political sweet spot where the right, the left, and much of the middle disapproved of our actions. Most of the country saw the GM and Chrysler rescues as new big-government bailouts for mismanaged firms. But the industry and its Democratic defenders saw the stringent conditions as a betrayal, especially compared to our approach to the financial industry. Bankruptcy would mean haircuts for creditors—as well as hardships for autoworkers, retired autoworkers, and auto dealers—that we hadn’t imposed during our bank rescues.

  “The banks have received ten times more money than the auto industry. And yet they seem to be receiving very different treatment,” Elizabeth Warren observed when I testified before her oversight panel. She challenged me directly: “Do you think the banks are better managed than the auto companies were?”

  Not necessarily, but GM and Chrysler weren’t viable in their current forms. The stress test would help determine the extent to which that was true for the banks. More important, the financial industry is very different from the auto industry and every other industry. A bank’s most important asset is its reputation for creditworthiness; its entire business depends on customers and creditors trusting it with their money. Banks are also much more reliant on short-term financing, and they are much more vulnerable to failure if that financing dries up.

  As we had seen with the Lehman collapse and WaMu haircuts in the fall of 2008, banks are also inherently more vulnerable to contagion. The automakers did not rely heavily on runnable short-term financing, and lenders could more easily tell which of them were in trouble. Using the bankruptcy process to reduce their debt obligations would send a message about the dangers only of lending to failing car companies, not car companies in general. By contrast, confidence is the lifeblood of the entire financial industry, and in a crisis, investors have a much harder time distinguishing the strong from the weak among banks than they do in other industries. That’s why haircuts for bank creditors or disorderly failures of systemic banks during a panic can trigger runs on the entire financial system, especially when the system is relying on short-term funding that can flee in an instant. Financial failures are different from auto failures or any other failures, which is why Congress equipped the FDIC with special bankruptcy-like powers to deal with failing banks.

  We thought that was pretty solid logic. But we were under no illusions that the public would embrace it. We just had to focus on what approach was most likely to work, and hope the public would judge us on the results rather than the optics.

  AT THE start of April, President Obama and I went to London for his first G-20 conference, a high-profile test of the international community’s ability to work together to attack the crisis. During the Depression, nations had turned inward, erecting new trade barriers in a damaging race to the bottom, embracing austerity while global demand withered. We were determined not to repeat those mistakes. Our fortunes were closely tied up with the rest of the world, and it would be tough to turn the U.S. economy around if the global economy continued to contract.

  Some G-20 nations, particularly Germany and France, wanted the meetings to focus primarily on long-term international regulatory reforms that could help mitigate the next crisis. But this crisis was still burning. Larry and I thought our main imperative should be building consensus for a coordinated short-term program of economic stimulus and financial support that could alleviate the crisis and boost global demand. The President decided we would push for both.

  A powerful stimulus effort would require a serious mobilization of resources for emerging economies. At an early stage in our internal discussions, Mark Sobel, a veteran Treasury civil servant who held my old international job, and the Fed crisis maven Ted Truman, whom I had recruited to Treasury to help oversee our international efforts, proposed that we should push to expand the IMF emergency fund that we helped create after the Mexican peso crisis. Sobel suggested we try to increase its financing from $50 billion to $300 billion, to make sure it had enough firepower to support countries in trouble.

  “Let’s do five hundred billion,” I said. The magnitude of the collapse had been huge, and there was no point in undershooting. Just like that, we decided to propose $500 billion.

  At the meeting of finance officials in Horsham, I had pushed to get the major G-20 countries to commit to a substantial fiscal stimulus program; my staff suggested a target of 2 percent of GDP. But I hadn’t made much headway. The post-summit New York Times headline was “No Clear Accord on Stimulus by Top Industrial Nations.” In Europe, support for stimulus was already fading. The Germans talked up the moral virtues of fiscal discipline for the world. There was talk that the French, who were also promoting a premature shift to austerity, might actually walk out of the London summit. Many foreign governments blamed U.S. profligacy for the crisis, and didn’t think we were in a position to tell them how to restore growth.

  The world desperately needed more stimulus, and Larry wanted us to push much harder to get the rest of the G-20 on board with a specific quantitative commitment. I was worried that the summit would end in division, giving the President an early failure on the international stage. So I told Larry’s deputy, David Lipton, to tone down the confrontational rhetoric in the press and to stop raising expectations we might not be able to meet. “We need a ‘Kumbaya’ moment,” I said. I had known many of the other finance ministers for years, better than I knew many of my new colleagues in the administration, and I knew they didn’t like lectures from the White House. Given the damage our crisis had caused and the extent of the failures in our financial system, we had to approach the negotiations with a measure of humility, and try to rebuild some credibility internationally.

  So we toned down the rhetoric and worked to build consensus. The results far exceeded expectations, and the agreements we announced in London helped begin to restore stability in the global economy. There was no specific target for fiscal stimulus—that was never going to happen—but we got a general commitment “to deliver the scale of sustained fiscal effort necessary to restore growth.” More concretely, the G-20 agreed to deploy $1.1 trillion in new international financing to fight the crisis, including the full $500 billion for the IMF emergency fund. We committed not to introduce new measures of trade protection. And we agreed on the broad outlines of future financial reforms, including higher capital standards. This time, the media takeaway from the summit was about coordination and cooperation.

  “World Leaders Agree on Global Response,” the Wall Street Journal headline said.

  London was an important success for the President, and it felt particularly satisfying to me. Despite our own economic weakness, and lingering resentments over U.S. foreign policy in the Bush years, we had set the agenda. After the summit ended, a beaming President Sarkozy walked up to President Obama and me, pumping his fists and chanting “Geithner! Geithner!”—as if I were his favorite soccer star.

  It was a goofy gesture, and the President looked at me with raised eyebrows as if to say: What’s up with this? But it was nice to feel less embattled for a moment.

  I RETURNED home to our crisis, bruising congressional hearings, and meetings, lots of meetings, endless meetings. Meetings are life in Washington. Often they’re just for show, a way to suggest motion or commitment to an issue. Sometimes their main purpose is to make people feel included. But occasionally they’re the real thing, a forum for actual policymaking. I got into the habit of walking into crowded meetings in Larry’s office and jokin
g: “Is this a real meeting or a fake meeting?” In other words, are we talking about a policy that requires a decision, or just talking? When it was a real meeting, I’d usually suggest that we skip the throat-clearing and fast-forward to the end of the PowerPoint deck so we could get to the debate about options. I wore my impatience too openly.

  Still, talking could serve a cathartic purpose. That April, as we waited for the stress test results, I held a series of follow-ups to the President’s Roosevelt Room meeting at Treasury, to give our internal critics an in-house outlet to continue to vent about the risks in our plan and propose ways to make it better. Larry and his colleague Jeremy Stein did push several sensible ideas that we would incorporate into the stress test, including tougher loss assumptions, a higher capital target, and more transparency, and the war-gaming of our overall strategy did help flesh out its strengths and weaknesses. But when they proposed more radical departures in strategy, I mostly just tried to wear them out, explaining the problems with their alternatives. Sometimes, after Larry would propose some new approach, I would look at him and say: “Larry, I heard what you like about that. Why don’t you make the case against?” And he was really good at that, often better than I was at explaining why his shiny new idea had its own crippling faults. Gene Sperling said later that no alternative plan presented in those meetings could withstand thirty minutes of debate.

  I wasn’t always on my best behavior. I got irritated when the critics offered anxieties without alternatives. When they did propose solutions, I could be pretty dismissive as I sucked the hope out of them. I remember before one late-afternoon meeting with this group of internal critics and advisers, I scheduled a one-on-one dinner with Larry at 8 p.m., so he would have an incentive to end the meeting; he always thought he had a better chance of getting me to see things his way if he could get me alone. I told the Treasury team before the meeting not to bring any food, in hopes that everyone would get hungry and lose interest. But 8 p.m. came and went. Around 9:30 p.m., my senior adviser, Sara Aviel, brought in some orange matzo boxes that had been left over from someone’s Passover seder.

  “Sara, no!” I shouted, only half in jest. “Don’t feed them! It will only encourage them!”

  One frequent theme in these sessions was the idea that my former Fed colleagues and I were so scarred by Lehman Brothers that we were afraid to do anything that might unsettle the financial system. Larry and Jeremy Stein would say I suffered from “Lehman Syndrome,” a kind of financial post-traumatic stress disorder. Sperling sometimes quipped that whenever anyone mentioned the words “systemic risk,” everyone’s IQ instantly dropped fifty points.

  I pleaded guilty to Lehman Syndrome. I thought that if you weren’t traumatized by the fall of Lehman and terrified by the thought of another Lehman, you weren’t paying attention. I was sometimes uncharitable about the “chicken hawks” of the crisis, the financial equivalent of ardent Iraq War supporters who had never fought in war and had the luxury of distance from the battlefield. Letting systemic firms fail, nationalizing them, haircutting their creditors—it all sounded alluring when you hadn’t lived through the consequences. I preferred to err on the side of avoiding another calamity. One meeting about AIG devolved into psychoanalysis, with various colleagues offering theories of how the crisis had warped me.

  “OK,” I said after twenty minutes of speculation. “We’re done with that.”

  Subjecting our plan to criticism—you could call it stress testing the stress test—helped make the President somewhat more confident in our approach. On April 27, for instance, he hosted a White House dinner for some of America’s leading economists, including critics such as Paul Krugman, Jeff Sachs, and Joseph Stiglitz. The President began with a pointed reminder that our strategy was my baby. “I’d like to hear your views,” he told the group. “You already know what Tim thinks.” The President invited each of them to speak, and most of what he heard was the familiar complaints that our plan was limp and unambitious, a recipe for zombie banks, too generous to the financial system. When the President asked each of them what they would suggest we do instead, he was exposed to an outside version of what he had heard inside the White House: more concerns and critiques, a few ideas, but nothing that sounded achievable, nothing that didn’t bring problems of its own.

  Krugman was typically thoughtful, and in some ways I found his critique the most gratifying. He said that while he wasn’t certain, he’d be inclined to do more nationalization, especially Citi and Bank of America. But he acknowledged that if we did that we would have to guarantee the rest of the system to guard against runs, just as I had been saying internally. “That could be expensive,” he conceded. I knew the President read Krugman’s columns—including some that first floated Sweden as a model—and I was happy to hear him validate my warnings. Early in the meeting, the President had mentioned how Larry and I often invoked the Hippocratic oath: First, do no harm. Now the economist widely regarded as the intellectual leader of the nationalization brigade was acknowledging that preemptive nationalization could do significant harm without a dramatic expansion of government guarantees.

  Krugman later downplayed how expensive it would have been to seize Citi and BofA, since their market capitalizations had dwindled. But we didn’t have the authority to do preemptive nationalization and comprehensive guarantees. And even if we had been able to take over two megabanks that held a quarter of U.S. banking assets, we would have faced the unappealing Hawk One/Hawk Two choice of either liquidating quickly, which would have created huge fire-sale losses for taxpayers, or trying to run the sprawling institutions for the long haul, which would have created all kinds of different problems.

  During this uncomfortable limbo period before the stress test was done, we came under pressure to do all sorts of things to demonstrate initiative. Several of the President’s advisers in the West Wing suggested that I fire Bank of America CEO Ken Lewis, who had engineered the troubled acquisitions of Countrywide and Merrill Lynch, and was pretty clearly on his way out anyway.

  “Tim, I’m trying to look out for you,” Gene Sperling told me. “If he’s going anyway, why don’t you push him out?”

  I understood the impulse to chop off a banker’s head and mount it on a stake, but while Bank of America had taken lots of risk and had received extraordinary assistance, we didn’t control the firm. If the firm failed to raise private capital after the stress test and we ended up having to take a majority interest, then its management would be replaced, as had occurred at Fannie, Freddie, AIG, and GM. But if we arbitrarily decided that Lewis had to go before the stress test revealed his firm’s plight, we’d undermine our own plan.

  “I know you’re looking out for me,” I told Gene. “I’m just not going to do things for message reasons.”

  We just had to wait.

  WAITING WASN’T easy. The markets were bouncing all over the place. From Inauguration Day through April, the S&P 500 index of financial stocks averaged a 5 percent daily shift; in normal times, a 1 percent swing is a big day. In White House staff meetings, Larry used to talk about how we never knew if the stock market was going to crash the next day and send us straight to the 1930s. The stress test would reveal whether the banks faced manageable difficulties, which could end the harrowing volatility, or giant capital holes, which would require drastic new measures.

  My former colleagues at the New York Fed were doing most of the analytical work on the stress test, but they kept a close hold on it that spring. They were trying to estimate future revenues as well as losses, an entirely new exercise for bank supervisors, made even more complex by this unusual moment of uncertainty. So their numbers were in flux. They didn’t want us to design policies based on preliminary data that could shift dramatically, which was sensible but frustrating.

  “I’m the fucking Treasury secretary and I can’t see these numbers?” I groused during one conference call.

  As we tried to pry results out of the Fed, we continued to debate various questions abou
t the design of the stress test. One involved transparency. Fed supervisors were compiling loss estimates for all nineteen firms, broken down by asset classes such as residential mortgages, commercial real estate, business loans, credit cards, and so on. They were also compiling revenue estimates, which they would combine with the loss estimates to calculate each bank’s capital gap. I thought the Fed ought to make these results as transparent to the public as possible, not just for the system overall but for each individual firm. I thought the best way to reduce uncertainty would be to allow investors to look under the hoods of the banks, so they could decide how stressful the stress test really was.

  But from the outset, Fed officials were reluctant to publicize bank-specific numbers. They feared that publicizing their guesses about future losses and capital positions could undermine confidence, prompting investors to run from banks that looked weak. They were also concerned that releasing detailed results compiled with proprietary bank data would hamper their access to that data in the future. “It violates confidentiality of supervisory information and will irreparably damage the credibility of the supervisory process,” they wrote in an early memo.

  Larry and I believed that we needed the fullest plausible disclosure, that even bad news would be better than no news. Investors were already assuming the worst about many banks. And I worried that voluntary disclosures by stronger banks alone could be even more stigmatizing to weaker banks than mandatory disclosures for all. I made the case to Ben and he ultimately agreed, overruling the traditional conservatism of the Fed staff. The stress test would end up revealing each bank’s capital gaps, projected losses, and even losses by asset class. There was a bit less transparency for revenue forecasts, but in general the Fed adopted the anti-Japan approach, coming as close as U.S. bank supervisors ever had to flinging open the system’s books.