Stress Test

  “The right way to do this is not going around and using guarantees or injecting capital,” Hank told the Senate Banking Committee.

  In Hank’s view, even mentioning the possibility of direct capital investments would have been bad politics, raising the specter of nationalization. Government ownership stakes in private firms sounded un-American, and aid to the institutions that caused the crisis sounded corrupt. But within the Fed and the Treasury, there was a growing recognition that we couldn’t defuse the crisis until we recapitalized the financial system. Current capital levels just weren’t thick enough to absorb the magnitude of losses that could lie ahead.

  Part of our increasing enthusiasm for capital was a result of increasing skepticism about asset purchases. My own skepticism had crystallized during the Fed’s battles with JPMorgan over the Bear Stearns mortgage pool. Using TARP to buy assets would have been much more complicated and similarly slow; Treasury thought it would take at least forty-five days after passage before the program could begin. And it would have been incredibly challenging to figure out what the government should pay for assets the markets were no longer trading. If we set prices too low, we could have forced financial institutions to realize huge losses that could have threatened their solvency. If we set prices too high, the purchases not only would have looked like egregious giveaways to Wall Street, they would have been exorbitantly expensive for taxpayers. And Congress had given us only $700 billion to spend—a big number, but by no means an overwhelming number compared to the troubled assets in the banking system.

  That was the main problem with using asset purchases to stabilize overleveraged financial institutions. Imagine a bank with $1 trillion in mortgage assets and $25 billion in capital, a 40:1 leverage ratio. To get it to a much safer 20:1 leverage ratio, the government could buy $500 billion of its assets, which would drain most of TARP on one institution. Or it could inject $25 billion in additional capital, achieving the same ratio with one-twentieth of the cash. Ken Garbade had first flagged this disparity back in March, suggesting in an email that direct capital investments “might be a more efficient way to contain the crisis.” Asset purchases seemed like a way to burn through our cash without solving the problem, forcing us to go back to ask for more from a Congress focused on the next election.

  At the end of September, I began a series of consequential meetings with Garbade, Meg McConnell, and other New York Fed economists to try to figure out how to advise Treasury to design TARP. We all agreed capital investments would be essential to stretch every dollar as far as it could go. In a September 26 memo titled “Escalation Options,” Garbade and two colleagues noted that government capital had been part of the solution to most financial crises, including the Depression. In this crisis, the Fed had provided plenty of liquidity—for commercial banks, investment banks, foreign banks, and money market funds—but we hadn’t yet addressed the deeper solvency issues of undercapitalized institutions. They had already absorbed massive losses, and with even more massive losses potentially looming, most of them had no way to raise capital privately. Capital would be essential, even if it might not be enough by itself.

  “The missing policy, to date, has been bank capital injections,” the memo said.

  I strongly agreed. So did Ben. And by the time TARP came up for a vote, Hank had come around to this view, too. He still expected to buy assets down the road, but he now believed injecting capital would be faster, more powerful, and more efficient.

  I admired Hank’s willingness to change course so quickly. But on a September 30 conference call, the day after the first disappointing House vote, I told Hank capital alone wouldn’t save troubled firms.

  “What you really need is the authority to guarantee their liabilities,” I said.

  Hank noted that the likelihood of Congress agreeing to guarantee trillions of dollars in bank debts—in addition to the $700 billion in spending the House had just rejected as overreach—was zero. He was right. But those meetings with Garbade and the New York Fed economists had persuaded me we’d need to deploy broad guarantees alongside capital injections to quell the panic and stabilize the system.

  Capital was necessary to make banks stronger, but creditors were running away from banks regardless of their strength, and capital alone wouldn’t stop a run already in progress. We needed to stop the run on liquidity first, and that would require guarantees. Just as FDIC deposit insurance prevented runs on insured deposits, and Treasury’s money market guarantee was preventing runs on money market funds, the government needed to make it clear that until the crisis subsided, lenders would be able to provide credit to financial firms without fear of default. Across the Atlantic, we were seeing how the demonstrated willingness of the Europeans to inject capital had not prevented an escalating run on their banks—and the longer the runs continued, the more capital the banks would ultimately need. Ireland had moved first to guarantee its bank liabilities, and other countries were following suit, to avoid the spreading run on the European banking system.

  Without guarantees, creditors would continue to run from banks, so banks would continue to hoard cash, market liquidity would continue to erode, and the run would intensify. Without more capital, banks would face increasing pressure to sell assets at distressed prices, which would create more losses and bigger capital holes; they would also have to pull back their lending, exacerbating what already threatened to be a severe downturn. So we needed both capital and guarantees. Capital was mostly about averting insolvency, while guarantees were mostly about preserving liquidity. But solvency problems could sap confidence and cause liquidity problems, while liquidity problems could force fire sales and cause solvency problems. Together, capital and guarantees could make the financial system stable enough to support growth again, instead of encouraging flight and working against growth.

  Once TARP passed, I felt like we finally had some firepower to attack our capital problems. But I still had no idea how we’d get the authority to do guarantees. And while good news usually created lulls in the crisis where we could regroup and recalibrate our strategy, the passage of TARP did not quiet the markets at all.

  THE WEEK of October 6 was the U.S. stock market’s worst week since 1933. The S&P 500 dropped 18 percent. The “fear index,” a measure of market volatility, hit an all-time high, while the spread reflecting stress in interbank lending also set a new record, four times its level before Lehman. Morgan Stanley returned to the brink, as rumors swirled that its deal with Mitsubishi might fall through; its stock price plunged while its credit default swaps soared. And AIG was burning through its $85 billion bridge loan so fast that we had to set up an additional $37.8 billion program. During the previous few weeks, we had committed more money to AIG than the federal government had spent on Social Security and Medicare.

  Asian and European markets were in disarray, too. A Financial Times headline summed up the situation: “New Panic Is Proof of Big League Crisis.” Ben and the ECB’s Jean-Claude Trichet, with my encouragement, worked with Mervyn King at the Bank of England and several other central bankers to launch the first-ever coordinated global interest rate cut. That was a remarkable act of cooperation for central banks that had always prized their independence and sovereignty, but it didn’t stop or even slow the collapse of the global markets.

  In the United States, private credit was now virtually unavailable. The financial system was increasingly unable to play its most basic role in the economy: facilitating the flow of money from those who wanted to lend to those who wanted to invest or consume. The run on money market funds after Lehman, followed by the broader run on the banking system that consumed WaMu and Wachovia, had spilled into the unsecured commercial paper market where businesses with strong credit histories could raise money to finance their day-today operations. My team at the New York Fed, along with some staff from the board in Washington, developed another creative lending program to try to prevent the collapse of commercial paper, but we ran into resistance elsewhere in Fed head
quarters.

  Our proposed Commercial Paper Funding Facility required the most expansive interpretation of Fed authority yet. The Fed can only lend against collateral, and the CPFF was designed for unsecured commercial paper, which by definition is not backed by collateral. It’s just corporate IOUs. My team proposed for the Fed to lend to a specialpurpose vehicle that would buy the unsecured paper, in return for a fee that would provide some protection against losses. The Fed would take considerable risk, more risk than some members of Ben’s team were comfortable with. But we didn’t think the more conservative alternatives they preferred would be as effective in reviving the market.

  The debate was slowing us down, and I feared we were on the verge of losing another critical part of the financial system. Companies such as Verizon, McDonald’s, and Caterpillar as well as the financial firms at the heart of the crisis all needed to issue commercial paper to pay their bills. Late one night, I called Ben at home and appealed to our common desire to avoid what he liked to call “Depression 2.0.”

  “You can authorize this. You just have to decide whether you want to fix the problem or not,” I told him. “There’s no point in trying some kind of limp solution.”

  I was tougher than usual on Ben, but I wasn’t telling him anything he didn’t know. As always, he was calm and patient, analytical but comfortable making quick decisions. He had a quiet force to him.

  “OK, I hear you,” he said. “I got it.”

  Our teams quickly worked out a framework we could all live with, and we decided to roll it out the morning of October 7. At the last minute, it was Bill Dudley from my shop who actually suggested putting the brakes on the announcement, citing some operational concerns. And this time, it was Washington that insisted on full speed ahead. None of us were certain the CPFF would work, but we had to keep acting, keep moving, keep doing stuff.

  “Look, we have no choice,” Don Kohn wrote at 6:25 a.m. “We have to make this work, however complex. I think we should announce this morning.”

  We did. And we sorted out the technical issues quickly enough that the CPFF would start lending money in late October, providing companies with nearly $150 billion in liquidity in its first week. But the financial system needed help sooner than that.

  WHAT IT really needed were comprehensive guarantees. I decided to call Sheila Bair to see if the FDIC could use its systemic risk authority to provide a solution.

  I probably wasn’t the best person to make the case to Sheila. We did not know each other well, and in our conflicts over WaMu and Wachovia I had been derisive of her arguments. While Hank and Ben had gently encouraged her to act in the interests of the system, selling her on the opportunity to be a firefighting hero, I had been more aggressive, dismissing her focus on limiting immediate risks to the FDIC insurance fund as parochial and shortsighted, warning that she was dragging us toward an epic disaster.

  But I thought it was vital for some government agency to stand behind the debts of the banks, so depositors and lenders wouldn’t have to assess which ones were good credit risks—because in the fog of a panic, nobody looks like a good credit risk. And the FDIC was the only plausible source of that authority. Congress had explicitly given it the ability to guarantee the debts of banks, even for banks that had already failed. I figured it was at least worth asking Sheila if she would guarantee the debts of banks before they failed—and not just one or two of them, but all of them.

  Sheila politely listened to my pitch. I again made the case that the more risk the government could commit to take, the less risk it would ultimately have to take. If we had unlimited capital to inject into the banks, we wouldn’t have needed guarantees, but we didn’t have unlimited capital, and there was no way to know how much capital would be enough to reassure the markets. By standing behind the obligations of the banking system, the FDIC could help prevent an all-out run on the system. The world couldn’t afford another Lehman—or another WaMu, though I didn’t mention that sore subject—and we had to take extraordinary measures to prevent it.

  To her immense credit, Sheila replied: “I think we might be able to do that.”

  Sheila and I would continue to have our differences, including on this very issue, but her willingness to use the FDIC’s guarantee authority was one of the turning points of the crisis. She had already invoked the systemic risk exception after watching Wachovia disintegrate so abruptly in the wake of WaMu, and now she seemed open to considering a more sweeping use of that emergency authority. She had gotten her glimpse into the abyss, her taste of the burden of fear that Hank, Ben, and I had carried for more than a year.

  As soon as I hung up the phone, I called Hank and Ben to report that Sheila might be willing and able to provide some form of broader guarantees. Hank invited Sheila to his office on Wednesday morning, October 8. Ben was there, too; I called in from New York. Hank began the meeting by telling Sheila it would be wonderful if she could apply the principles of the initial Wachovia plan in a broader way—in essence, if she could guarantee the liabilities of the banking system. She got defensive, reminding us there was only $35 billion left in her deposit insurance fund. Hank replied that if we couldn’t guarantee the system, the run dynamic would feed on itself, more major banks would fail, and there would soon be nothing left in her fund. I made my usual case for overwhelming force being safer and cheaper than tentative half-measures.

  It was clear Sheila had reservations, and she was too adept a bureaucratic warrior to let us box her in. But it was also clear she was willing to provide some kind of guarantees. And with TARP, we’d be able to provide capital as well.

  Now we just needed to nail down the details. We had spent the past three frenetic weekends trying to save dying firms with systemic implications. We would spend the next weekend, Columbus Day weekend, trying to make sure we wouldn’t have to do that anymore. We had been chasing an escalating crisis for more than a year, always behind the curve, patching together ad hoc fixes to try to hold the system together. Finally, we had a chance to design a more definitive solution.

  I SHOWED up early Saturday at the south entrance of the Treasury building, the place where my government career began almost exactly twenty years earlier. The Secret Service didn’t have me cleared in the system, so I had to wait outside for a half hour until they could verify that I had a meeting with Hank. My New York Fed colleagues and I sat down on the Treasury steps—in front of a statue of the first Treasury secretary, Alexander Hamilton, holding a three-cornered hat—and joked about all the reasons it might make sense to keep me out of the building.

  It was a sunny, crisp, beautiful fall day in Washington, and even though the world was still disintegrating, I felt as light as I had in months. We finally had the outlines of a plan and the authority to execute it. We were going to deploy federal resources in ways Hamilton never imagined, but given his advocacy for executive power and a strong financial system, I had to believe he would have approved. It was Hamilton who had insisted on assuming the debts of the states and paying the new nation’s creditors at face value to establish the U.S. government’s reputation as a solid credit risk, even though haircutting foreigners and speculators who had bought Revolutionary War debt for pennies on the dollar would have been much better politics. He was America’s original Mr. Bailout. As the inscription on his statue said: “He touched the dead corpse of the public credit and it sprang upon its feet.”

  One of our main tasks that weekend would be to persuade the FDIC to use the power of that public credit to provide guarantees that would help revive private credit. Ever since Sheila had acknowledged her authority to backstop the banking system, she had been backpedaling, proposing a variety of limitations and conditions that would have undermined the power of the guarantees. She wanted to absorb only 90 percent of the losses from bank debts, which potentially meant 10 percent haircuts. She also wanted to charge near-punitive fees for the guarantees. And she wanted to limit them to the new debts of banks with FDIC-insured deposits, excluding existi
ng bank debts as well as all debts of bank holding companies and their broker-dealer subsidiaries.

  Some Fed and Treasury officials shared the FDIC’s unease about the scope and generosity of the guarantees. There was concern in some corners of the White House, too. But Sheila took the lead in trying to minimize the FDIC’s risk, and we spent most of the weekend in tense negotiations. I pushed for the broadest possible guarantees, with fees modest enough that institutions would actually want to use them, and with no haircuts whatsoever. Once again, I was the bad cop, while Hank and Ben were conciliators. At one point, Sheila got so mad and dug in that Hank invited her into his office and essentially begged her to help us avoid Armageddon, promising to make sure she would get the credit she was due.

  Ultimately, Sheila agreed to 100 percent guarantees—imposing haircuts would have been ludicrous—with fees that weren’t prohibitive. But her efforts to limit the scope of the guarantees raised a thornier problem, a boundary problem. Once a government sets a dividing line between what’s guaranteed and what isn’t, it can spark a run from the debts and firms just beyond the line to the debts and firms on the safe side. We needed the guarantees to cover the obligations of bank holding companies, not just their FDIC-insured bank subsidiaries, or else the broader institutions and their nonbank affiliates would have come crashing down. The escalating guarantees in Europe were already creating the possibility that U.S. banks would be the only major banks operating without the full protection of their government behind them. Ireland’s guarantees had prompted investors to shift cash from British banks to Irish banks, prompting the United Kingdom to offer its own guarantees; this arms race put additional pressure on us to follow suit.