Meanwhile, the Fed’s new Commercial Paper Funding Facility had quickly proved its value. By the end of the day on October 29, two days after it opened for business, it had purchased $145 billion in three-month commercial paper. A week later, it held $242 billion and, at its peak in January 2009, $350 billion. The program halted the rapid shrinkage of this crucial funding market and helped return interest
rates on commercial paper to more normal levels.
Even with these important new tools and policy initiatives, the financial system was still reeling from the Lehman shock. Investors, whose fears a year earlier had centered on subprime mortgages, were shying away from funding virtually any type of private credit, such as credit card loans and auto loans. They had little reason to think that these other types of credit would suffer similar losses, and they never did. But, like subprime mortgages, these forms of credit also were routinely bundled into securities and sold to investors, leading to guilt by association. The sharp drop in investor demand for these asset-backed securities posed still another risk to the whole economy.
In response, we collaborated with the Treasury to develop yet another program. We invoked Section 13(3) again and announced the TALF (for Term Asset-Backed Securities Loan Facility) on November 25. But, given the complexities involved in setting up the facility, it would not extend its first loan until four months later. Under the TALF, we would lend for terms of up to five years to investors who would buy AAA-rated securities backed by credit card loans, student loans, auto loans, commercial mortgages, and loans guaranteed by the Small Business Administration. Our loans would be without recourse, meaning that the borrower could give us the asset-backed securities they had purchased in lieu of fully repaying the loan. That provided “downside protection” for the borrowers. But turning over the securities to us before maturity would make sense only if the return on the securities dropped below the cost of the loan.
We took measures to protect ourselves. Investors could borrow only a portion of what they paid for securities. Thus, they would incur the first losses, if any, and had “skin in the game.” Further, taking the step it had declined to take with the Fed’s commercial paper facility, the Treasury provided $20 billion out of the TARP as capital to support Fed lending of up to $200 billion. The TARP money would be next in line, after the private sector, to cover any losses. As it turned out, none of the securities financed by the TALF would ever be “put” back to the Fed. The program took no losses and returned profits to the taxpayer.
THE OUTGOING BUSH and incoming Obama administrations—with advice from the Fed and FDIC—also wrestled with home foreclosures. The problem was escalating as millions of homeowners lost their jobs and millions more found themselves “underwater,” with their homes worth less than the amount they owed on their mortgages. Once largely confined to subprime mortgages with teaser rates, foreclosures now were rising even for plain vanilla prime mortgages. The economic and social costs extended well beyond lenders’ losses and the pain suffered by displaced families. Empty, foreclosed homes blighted entire neighborhoods, pushing down the value of nearby homes and reducing local tax bases.
Hank Paulson’s voluntary Hope Now program, launched in October 2007, had made a creditable start at reducing foreclosures. However, without government funding, its scope was necessarily limited. The subsequent Hope for Homeowners plan, enacted in July 2008, provided for refinancing through the Federal Housing Administration, but Congress effectively sabotaged it by imposing onerous requirements
and fees that discouraged both homeowners and lenders from participating.
Meanwhile, the Fed had sponsored or cosponsored more than 100 foreclosure-prevention events across the country. The Federal Reserve Bank of Boston, for instance, helped organize a massive workshop in August 2008 in Gillette Stadium (home of the New England Patriots football team) that brought together more than 2,200 distressed borrowers with lenders, servicers, and counselors. We worked with the nonprofit NeighborWorks America to help communities minimize the blight created by foreclosures.
After the election, with a new administration coming in and with the TARP money available for foreclosure relief, it seemed a good time for new ideas. Sheila Bair had been pushing hard for more action to prevent foreclosures. After the FDIC took over IndyMac in July 2008, it began modifying mortgages that the lender had owned or serviced. It capped the mortgage payments of borrowers in trouble at 31 percent of their income. That was accomplished through a variety of strategies, including reducing the interest rate on their loans, forgiving part of the principal owed, and extending the maturity of the mortgage (say, from thirty to forty years).
It seemed a worthwhile effort, but, as we debated options during the transition, it was also too early to judge its success. As of late 2008, only a few thousand modifications of IndyMac mortgages had been completed. We didn’t know whether the modifications would stick or ultimately result in another default. (A later FDIC assessment would find that two-thirds of seriously delinquent IndyMac loans would re-default within eighteen months of having been restructured.) Nevertheless, in mid-November, what became known as the FDIC’s IndyMac protocol was adopted in modified form by Fannie and Freddie. Sheila also pressed the administration to use TARP funds to offer lenders guarantees as an incentive to adopt the IndyMac guidelines. If a borrower with a modified mortgage subsequently defaulted, then the government would make up half the loss.
I didn’t always agree with Sheila’s views, but I had to admire her political talent. Ignoring the administration’s normal policy process, and by using the media and lobbying on Capitol Hill, she persuaded (mostly Democratic) legislators to endorse her plan—including House Speaker Nancy Pelosi and Senator Dodd. But Sheila came very close to characterizing anyone not fully in support of her plan as an opponent of foreclosure relief in general. I recall Hank being very exercised by a New York Times story criticizing the Treasury for failing to immediately act as Sheila advised and comparing it to the Federal Emergency Management Agency after Hurricane Katrina.
At the Fed, we agreed wholeheartedly with the goals of Sheila’s plan but questioned some of the specifics. Deputy research director David Wilcox and his team compared alternative strategies. Among several flaws, the FDIC’s guarantee plan seemed unnecessarily generous to lenders. In its original form, it provided a perverse incentive for lenders to modify the mortgages of the borrowers with the least capacity to remain current, since if the borrower re-defaulted the lender would pocket the government guarantee payment and could still foreclose on the home.
Fed economists proposed alternative plans, including variants of the FDIC’s IndyMac protocol, that they believed would achieve more sustainable modifications at lower cost to the government. We also suggested that the Fed and Treasury could create a new special-purpose vehicle to buy at-risk mortgages in bulk from lenders and investors. Under our plan, this new government entity would be capitalized by $50 billion from the TARP and could borrow from the Fed. The mortgages it purchased would then be modified by independent specialists, not by private-sector lenders and investors, and refinanced by the Federal Housing Administration.
Hank had his doubts about Sheila’s plan to use TARP funds to partially guarantee modified mortgages, for reasons similar to ours. But his time as Treasury secretary was short, so he focused his team on analyzing competing proposals without making recommendations. Larry Summers, who had been chosen to head Obama’s National Economic Council, weighed in with a memo to the president-elect on December 15, in consultation with other Obama advisers. Like the Fed, Larry favored fixing the Help for Homeowners program to make it more attractive to both lenders and borrowers. He also shared the Fed’s concerns about the FDIC’s plan to compensate lenders for re-defaults of modified mortgages. He instead supported incentives for lenders to reduce interest rates on troubled mortgages. Larry’s memo did not address the Fed’s idea of having a special-purpose vehicle buy troubled mortgages in bulk, but he let us know that he didn’t like the political optics of the government foreclosing on people, which would sometimes occur if the new government entity took responsibility for modifying mortgages. Final decisions would have to await the swearing-in of the new president and his new Treasury secretary.
MONETARY POLICYMAKING CONTINUED without regard to presidential transitions. Ongoing market turmoil had given us plenty to discuss at the FOMC meeting of October 28–29, just before the election. Buyers remained in full retreat. From the September FOMC meeting until the day before the October meeting, the Dow Jones index fell nearly 2,900 points, losing about a quarter of its value. The market’s volatility was mind-boggling; after its sharp declines, the Dow soared almost 900 points on the first day of our meeting, without any obvious good news to explain the jump.
Falling prices for houses and stocks, and tightening credit, in turn accelerated the decline in the economy. Household and business confidence—the “animal spirits” so important to economic growth— seemed to be in free fall. The University of Michigan’s well-known survey of households showed consumer sentiment at its lowest point in almost thirty years. Board economists forecast a recession that would last through the middle of 2009. Their guess about the timing would prove accurate, but, as with most outside forecasters, neither the staff nor most FOMC participants yet appreciated how extraordinarily deep the downturn would be. We know now that the U.S. economy shrank at a 2 percent annual rate in the third quarter of 2008, an astonishing 8.2 percent rate in the fourth quarter (the worst performance in fifty years), and a 5.4 percent rate in the first quarter of 2009. It was easily the deepest recession since the Depression. Inflation, meanwhile, was falling rapidly, reflecting a $30-per-barrel
drop in oil prices and overall weakness in the economy.
Dismal markets and shrinking economies were now a global phenomenon, in emerging-market and advanced economies alike. Russia had suspended trading to try to stop the slide in its stock market, and Mexico had spent 15 percent of its foreign exchange reserves in an effort to arrest the fall of the peso.
Japan’s Nikkei stock index hit a twenty-six-year low on the first day of our late-October meeting. Given the American origins of the crisis, it was ironic that global investors, desperate for assets perceived to be safe, were snapping up dollar-denominated assets, particularly Treasury securities. The activity had pushed up the dollar’s value by a remarkable 9 percent since the last time we had met. With the U.S. economy declining rapidly, the rise in the dollar was hardly good news. It made U.S. exports more expensive and thus less competitive on world markets.
We expanded our currency swap lines yet again at the meeting by adding four carefully chosen emerging economies: Mexico, Brazil, South Korea, and Singapore. We chose these countries based on their importance to U.S. and global financial and economic stability, declining requests for swaps from several others. The additions brought the total number of our central bank swap lines to fourteen. Two weeks earlier, we had removed the limits on draws by the European Central Bank, the Bank of England, the Swiss National Bank, and the Bank of Japan, reflecting both the demand for dollars in Europe and Japan and our close relationships with those central banks.
The case for cutting interest rates further looked strong. I told the Committee that the steps taken so far to end the crisis had probably not yet reached their full effect. Nevertheless, I argued, we were facing what looked likely to be a deep and protracted recession, and it demanded bold action. The Committee approved, without dissent, a 1/2 percent cut in the federal funds rate, lowering it to 1 percent—the same level we had reached at the height of our concerns about deflation in 2003.
The prior evening, I had met with the Reserve Bank presidents at their usual pre-FOMC dinner. They were deeply worried about the political risks created by our interventions, particularly the bailout of AIG. And although the 13(3) lending programs were the responsibility of the Board, not the FOMC, some of the presidents felt that I hadn’t consulted them enough. Every president wanted assurances that they would be kept in the loop, if for no other reason than to have answers to the inevitable questions posed at their public appearances. Given the pace of recent events, I had had good reasons to depart from the careful consensus-building style that I preferred. But the presidents’ concerns were reasonable, and I offered to hold biweekly videoconferences to update them on any Board initiatives, as well as financial, economic, and legislative developments.
WE HAD HOPED that capital infusions under the TARP would end the dreaded weekend rescues of faltering financial giants. Unfortunately, worsening economic conditions and mounting losses continued to press on weaker institutions. AIG, the recipient of the Fed’s $85 billion bailout in September, was one of them.
We had already tacked on a loan of an additional $37.8 billion in early October, to help AIG finance its
holdings of private-label (not guaranteed by the government) mortgage-backed securities. But even that was not enough. The company’s losses had ballooned in the third quarter to more than $24 billion. To survive, the firm needed capital as well as some relief from the tough loan terms its board had accepted in the initial bailout in September.
Letting AIG fail was not an option, for the same reasons that we had intervened two months earlier, so we restructured the terms of the rescue and, in the process, increased its size (including Treasury funds) to more than $150 billion. Paulson at first wanted the Fed to provide all the funding for the new deal, but Geithner and I convinced him that the Fed couldn’t do it alone. AIG needed a major infusion of capital, which, with private markets effectively closed, could only come from the Treasury. On November 10 the Fed and the Treasury announced a restructured AIG bailout, which included the purchase of $40 billion of preferred stock by the Treasury. AIG did not qualify for the Capital Purchase Program, which was broad-based and aimed at strengthening relatively healthy, not troubled, firms. We required it to accept stricter conditions, including a higher dividend payment on the Treasury’s stock. As part of the restructuring, AIG repaid the supplementary loan the Fed had made in October, and we were able to reduce our original credit line to AIG from $85 billion to $60 billion. In exchange, we agreed to substantially lower the interest rate we charged and extend the time for repayment from three years to five.
To cap future risks to AIG’s stability the Board, by invoking Section 13(3) once again, allowed the New York Fed to create and finance two new legal entities, to be called Maiden Lane II and Maiden Lane III. The New York Fed would lend $22.5 billion to Maiden Lane II, which in turn would acquire the private-label residential mortgage securities that had inflicted so many losses on AIG. Maiden Lane III, with a New York Fed loan of $30 billion, would buy from AIG’s counterparties the collateralized debt obligations insured by AIG’s Financial Products division. By acquiring the CDOs, we effectively closed out the insurance policies that had brought AIG to the brink in September. Tim Geithner likened the removal of troubled assets from AIG to a tourniquet, intended to stop the hemorrhaging. By that analogy, our loans and Treasury’s capital were transfusions. I hoped they would be enough to save the patient— again, not out of any care about AIG, but for the sake of the broader system. In particular, we wanted to avoid further credit rating downgrades of AIG, which would automatically lead to massive new demands for collateral and cash.
To protect the Fed, we hired outside asset managers to value the securities we were acquiring. AIG provided $1 billion to absorb the first of any losses in Maiden Lane II, as well as an additional $5 billion to absorb the first losses in Maiden Lane III. The two new vehicles would appear on the Fed’s balance sheet with the market values of the securities they held, updated quarterly. On Sunday, the day before the November 10 announcement, Paulson and I had sat in his office and called congressional leaders, explaining the new package and why we had little choice but to implement it. As was often the case with Congress, pushback was minimal at the time. It would come later.
Stuffed with new capital and relieved of many of its troubled assets, AIG seemed stable, at least for
the moment. But the restructuring also reignited anti-bailout outrage (some real, some intended for the TV cameras) that we had faced after the original rescue. I understood the anger, especially since the bailouts had not averted what now looked to be a serious recession. But I had no doubt that keeping AIG afloat was essential to prevent the crisis from metastasizing yet further. Moreover, helping AIG to remain viable was the best—really, only—way that we could get back the money the taxpayers had put into the company.
Our Maiden Lane III purchases of the CDOs that AIG had insured raised a new issue, which perhaps I should have anticipated but didn’t. In buying the insured securities, we had essentially allowed AIG’s counterparties—most of them large financial institutions, some of them foreign—to receive the full benefit of the insurance. As these facts sunk in, we were pilloried by Congress and the media for conducting “backdoor bailouts.” Why had we not insisted that those counterparties, which included companies like Goldman Sachs, bear some losses?
I was so intensely focused on controlling the panic that initially this criticism befuddled me. The New York Fed had broached the idea of voluntary payment reductions to some of the counterparties, without success (not surprisingly). As we would point out repeatedly, we had no legal means to force reductions. AIG’s counterparties had contracts to receive the insurance payments, which were no less valid than the claims of AIG’s other creditors or the customers who held more traditional kinds of insurance policies with the company. Applying supervisory pressure to the counterparties to accept reduced payments, as many critics said we should have done, would have been a clear abuse of our authority. And many of the counterparties were foreign institutions outside our jurisdiction, whose national regulators in some cases backed their refusals to take reductions.
We also faced criticism for not immediately disclosing the identities of the CDO counterparties. Although we had legitimate reasons, our decision in that case was tone-deaf. We had initially focused only on the legalities, including provisions of the Uniform Trade Secrets Act, which did not permit us to make such unilateral disclosures. And we worried about the subsequent willingness of counterparties to do business with AIG. After being pounded by Congress and the media for weeks, however, we asked AIG to disclose the names of the counterparties, which it promptly did.
I did my best to explain our actions and defend our response. Michelle Smith had by this time abandoned the Fed’s traditionally conservative communications strategy and was bringing me proposals for engaging more intensively with the media and the public, including an invitation to speak and take questions from reporters at the National Press Club (I would appear there in February). I cooperated with John Cassidy of The New Yorker magazine on a lengthy piece that offered one of the first inside looks at how the Fed had battled the crisis. And I increased the pace of my public speaking, picking high-profile venues around the country to describe the many steps that we were taking to fight the crisis and how they fit together. I also testified regularly to congressional committees and stayed in frequent contact with legislators in off-the-record meetings or on the phone. I focused on party leaders and members of our
oversight committees, but, except in cases of unresolvable schedule conflicts, I always accepted requests to meet or call from any member of Congress.
AIG WAS NOT the only corporate patient in intensive care. Within a few weeks, Citigroup also looked to be on the brink. Since its founding in 1812 as the City Bank of New York, Citi has remained a powerful but controversial institution, involved in nearly every financial panic the country has experienced. It almost collapsed after facing crippling losses on loans to Latin American countries in the 1970s and 1980s, and, again in the 1990s, when its U.S. commercial real estate loans soured. It survived thanks to a $590 million capital infusion from Saudi Arabian prince Al-Waleed bin Talal in 1991. After its historic 1998 merger with Travelers Group, Citi became the nation’s largest bank holding company, but, by the end of September 2008, it had slipped to second behind JPMorgan Chase. The company had retreated somewhat from the “financial supermarket” vision of Sandy Weill, but it remained enormous, with total global assets hovering above $2 trillion.
Citi was sprawling and complex, with many lines of business and operations in dozens of countries. Its managers struggled to forge a coherent strategy. Our supervisors had been particularly concerned about the bank’s ability to identify and measure its risks on a company-wide basis. Its management weaknesses and risky investments left it exceptionally vulnerable. Its problems had been made worse after some of its structured investment vehicles lost their outside funding and were brought onto its balance sheet. Sheila Bair had sharply criticized Citi’s management, including CEO Vikram Pandit, and its board of directors, headed by Richard Parsons. How much of Citi’s troubles could be blamed on Pandit is debatable; he had become CEO only the previous December. But Sheila was right that Citi was a weak organization and that the Fed and the Office of the Comptroller of the Currency had not done enough to fix it.
Citi had gotten a lift from $25 billion in capital that Pandit had accepted at the Columbus Day meeting at Treasury, but its condition remained precarious. Market concerns about the company were again escalating as the overall economy deteriorated. Its bank subsidiary, Citibank, relied heavily on $500 billion in foreign deposits, which were not insured by the FDIC, and on wholesale funding—both potentially subject to runs. Indeed, increased access to U.S.-based, FDIC-insured deposits had been a principal reason behind Citi’s pursuit of Wachovia. In a pattern that we had seen many times, Citi’s less stable funding sources began to pull away.
On Thursday, November 20, I entertained Board members in my dining room at the Fed for our annual Thanksgiving lunch. It was a pleasant occasion, but calls about Citi had already begun that morning and would continue through the weekend. Rumors were spreading that Citi was looking to sell itself, although it was not clear what institution could make such a huge acquisition.
A further complication arose on Friday, when President-elect Obama announced his intention to nominate Tim Geithner to be his Treasury secretary. Although Tim had not previously been part of Obama’s inner circle, he had been asked during the campaign to brief the candidate about the crisis.
Obama had obviously been impressed. Following the announcement, Tim immediately recused himself from his crisis-fighting duties at the New York Fed, as well as from monetary policy decisions. But he would remain in close touch with us in his new capacity as adviser to the president-elect.
This time without Tim, we again debated how to stabilize a giant financial company in danger of collapse. At the G-7 meeting in Washington in mid-October, we had committed publicly to avoiding the failure of any more systemically critical institutions. Citi easily met that criterion, but Sheila initially said she thought that Citi could be allowed to fail. I suspect she was being provocative, and in any case she ultimately joined our effort to prevent the company’s collapse. I did agree with Sheila that Citi was being saved from the consequences of its own poor decisions. But, as we couldn’t say often enough, we weren’t doing this for Citi, its executives, its creditors, or anyone on Wall Street, but in the interest of overall economic and financial stability.
As before, we found ourselves in “before-Asia-opens” mode, with the inevitable conference calls, spreadsheets, and collective mood swings as possible solutions were proposed and shot down. The availability of the TARP reduced the pressure somewhat. On the other hand, not only were we negotiating with Citi, but the Treasury, Fed, and FDIC also were negotiating about each agency’s share in the rescue. The leaders and staff of each of the agencies felt tapped out—politically, financially, and often physically —and under pressure. It didn’t help that Citi was maddeningly slow in responding to our requests for information, further reducing our confidence in the company’s management. At one point Sheila emailed: “Can’t get the info we need. The place is in disarray. How can we guarantee anything if citi can’t even identify the assets?” Despite these tensions, the stakes involved kept us on track. I’m sure that everybody involved in the negotiations, even Sheila, knew that, in the end, letting Citi go was not an option. We would all have to make concessions to find a workable solution.
We announced the package to stabilize Citigroup late Sunday evening, November 23. It included an additional $20 billion of TARP capital, in the form of preferred stock that paid the government an 8 percent dividend rather than the 5 percent required by the Treasury’s Capital Purchase Program. In addition, we agreed to provide a “ring fence” (a backstop guarantee) for a $306 billion portfolio of Citi’s troubled assets, which included residential and commercial mortgage-backed securities. Citi would bear the first $37 billion of losses from this portfolio, including $8 billion of existing reserves set aside against those losses. The government would bear 90 percent of any additional losses, with Treasury (through the TARP) shouldering the first $5 billion of government’s share, and the FDIC the next $10 billion. In exchange for this guarantee, the Treasury and the FDIC would receive preferred stock in Citi. In the unlikely scenario that the losses became so severe as to exhaust the Citi, Treasury, and FDIC commitments, the Fed would make a backup loan equal to 90 percent of the remaining assets, taking all of the assets as collateral. As part of the deal, Citi agreed to virtually eliminate its stock dividend and to adopt mortgage modification procedures advocated by Sheila and the FDIC to reduce unnecessary foreclosures. The FDIC’s participation in this arrangement required the agencies to invoke, again, the
systemic risk exception to the requirement that FDIC interventions be conducted at least cost. The market liked the deal, at least initially. Citi’s stock price soared nearly 60 percent.
Why did we ring-fence Citi assets, rather than just add more capital? The ring fence, modeled on the deal that the FDIC had struck when Citi was trying to acquire Wachovia, was intended to protect Citi from a worst-case scenario. Eliminating the relatively small risk that Citi could suffer extreme losses reassured investors while requiring a relatively smaller commitment of dwindling TARP funds.
Of the $350 billion that made up the first tranche of the TARP, Paulson had already committed $250 billion to the Capital Purchase Program, $40 billion to AIG, and now $20 billion in new capital for Citi. He was facing demands to help homeowners and auto companies, and it was surely important to keep some money available for emergencies. The obvious solution was to ask Congress to release the second $350 billion of the TARP, but, concerned about the politics and whether Congress would go along, the Bush administration had not yet made the request.
IN THE WAKE of the Citi episode, on Tuesday, November 25, the Fed made an announcement that foreshadowed the next phase of our response to the crisis. We said that we planned to buy up to $500 billion in mortgage-backed securities guaranteed by Fannie, Freddie, and the Government National Mortgage Association, or Ginnie Mae. (Ginnie Mae is wholly owned by the government, unlike Fannie and Freddie, which had been owned by private shareholders before their takeover by the Treasury.) We also announced plans to buy up to $100 billion of the debt issued by Fannie, Freddie, and other government-sponsored enterprises to finance their own portfolios. We were motivated by our concerns about housing. Uncertainty had driven buyers from the MBS market. Investors didn’t know how long the government would support Fannie and Freddie or how much worse housing would get. Moreover, some financial institutions, short of liquid assets and capital, were actively dumping MBS on the market— pushing up mortgage rates. Our purchase program would provide increased demand for MBS while signaling the government’s commitment to the companies. Even though we would not actually purchase any MBS until January, the announcement itself had a powerful effect on investor confidence. The spread between yields on MBS issued by Fannie and Freddie and yields on longer-term Treasury securities fell by 0.65 percentage points within a few minutes of the press release, a large move. Rates on thirty-year mortgages dropped from around 6 percent at the end of November to around 5 percent at the end of December.