Stress Test

  We would find out that was true the hard way.

  I DIDN’T enjoy public speaking, and I wasn’t good at it, so I wasn’t inclined to do much of it. But I couldn’t be invisible in my new job. I rarely discussed monetary policy or the broader economic outlook, continuing a New York Fed tradition of public deference to the chairman on those issues. But I did talk a lot about risks to the stability of the financial system, usually to financial audiences. I didn’t consider my speeches a particularly powerful way to influence behavior. I didn’t seek media coverage, and I didn’t get any. But I did try to convey what I was learning about the strengths and weaknesses of the system, and to outline my hierarchy of concerns. In my careful, qualified, occasionally tortured way, I tried to lean against the wind.

  “We need to maintain a degree of humility and caution about our capacity to anticipate the nature and dynamics of future stresses to the financial system,” I said in May 2004. I gave an entire speech that October about strengthening our regime for addressing systemic risk: “If you want peace or stability, it’s better to prepare for war or instability.” In February 2005, I said the financial community’s main challenge was a challenge of imagination: “How do we generate the will today to build a greater degree of insurance against a more uncertain future, particularly if the risk of adversity seems remote and the immediate future looks strong?”

  And so on. Even though the financial sector seemed healthy, I talked about systemic risks in almost every speech I delivered as New York Fed president.

  These speeches weren’t designed to be warnings. I didn’t try to be Chicken Little. I expressed my concerns in the on-the-one-hand-on-the-other-hand tone of a central banker, as if I were suggesting that the recent history of celestial stability did not necessarily rule out the possibility of some portion of the sky falling at some point in the future. In my first speech, “Change and Challenges Facing the U.S. Financial System,” I didn’t address the challenges until my sixteenth paragraph; the first fifteen were mostly about positive developments. I thought audiences might be more willing to think about our vulnerabilities if I acknowledged our strengths, but I probably made it easy for people who just wanted to hear good news to block out the caveats. And while I didn’t have Greenspan’s flair for eloquent fog, my speeches were never a model of clarity or hair-on-fire force. I was careful to express my concerns in understated, nuanced, deliberately dull language that wouldn’t move markets or depress confidence. I had once seen an offhand comment about the yen by Secretary Bentsen trigger a damaging run on the dollar, and I had no intention of making a similar mistake.

  Still, there was an undertone of anxiety to all my speeches. And the main source of my anxiety was the rapidly growing, heavily leveraged, lightly regulated nonbank financial system. Here’s some typical Fed-speak from my October 2004 talk about systemic risk: “There are a larger number of nonbank financial intermediaries operating outside the supervisory safety and soundness framework that are sufficiently large or integral to the financial system that their failure or anticipated failure could have major implications for the functioning of the markets in which they operate and their financial institution counterparties.”

  In other words, a big nonbank could cause big problems.

  I WASN’T confident that our rules would ensure that Fed-supervised banks had enough capital to survive a severe crisis. But I knew many nonbanks had much less capital, even though they didn’t have the safeguard of insured deposits and wouldn’t have access to Fed loans in an emergency. For example, by the end of 2007, capital levels at the five SEC-regulated Wall Street investment banks—Bear Stearns, Lehman Brothers, Merrill Lynch, Morgan Stanley, and Goldman Sachs—were just 3 percent of assets. At the mortgage giants Fannie Mae and Freddie Mac, they would drop to barely 1 percent of the assets they owned and guaranteed.

  The question was what to do about it. The Fed didn’t have the legal authority to force Bear Stearns, Lehman Brothers, or other investment banks to raise more capital. We couldn’t even generate stress scenarios bleak enough to force the banks we regulated to raise more capital. Perhaps our capital buffers were too thin, but they were already thick enough to drive trillions of dollars of assets—more than there were in the entire commercial banking system—outside our direct supervision. In a time of seemingly unbounded optimism about the trajectory of housing prices and the stock market, investors were willing to finance substantial increases in leverage, and that money tended to flow where the regulatory constraints were weakest.

  A big part of the problem was America’s balkanized regulatory system. It was riddled with gaps and turf battles. It was full of real and perceived sources of capture. And nobody was accountable for the stability of the entire system.

  Even the traditional banking sector was a byzantine mess, with responsibilities for supervising thousands of commercial banks divided among the Fed, the Federal Deposit Insurance Corporation (FDIC), and the Office of the Comptroller of the Currency (OCC), as well as state banking regulators working from fifty different sets of rules. The Office of Thrift Supervision (OTS) regulated “thrifts,” which were essentially banks focused on mortgage markets; their deposit-taking subsidiaries were subject to capital requirements and eligible for the Fed’s discount window, but the riskier affiliates and parent companies generally operated without adult supervision. There were also geographic divisions within the Fed. The New York Fed oversaw Citigroup and JPMorgan Chase, while the Richmond Fed supervised Bank of America and Wachovia, and the San Francisco Fed handled Wells Fargo.

  Often, multiple agencies oversaw a single institution. For example, the Fed supervised the “holding companies” of Citigroup and JPMorgan Chase, the umbrella entities sitting atop hundreds of bank and nonbank subsidiaries, but not their actual commercial or investment banks. We were supposed to keep tabs on the institutions as a whole, but by law we had to defer to the primary supervisor, in those cases the OCC for their commercial banks and the Securities and Exchange Commission for their investment banks. The Fed also shared responsibility for the U.S. affiliates of foreign banks with state regulators, as well as home-country supervisors in London, Zurich, Frankfurt, and around the world. These divisions of labor evoked the parable of the blind men and the elephant, with nobody accountable for seeing the full picture of a corporation, much less the interconnections of the entire system.

  This glut of watchdogs with overlapping jurisdictions encouraged regulatory arbitrage. Banks often reorganized as thrifts to get the notoriously weak OTS as their supervisor, or shopped for another regulator they thought would give them favorable treatment. The OTS and OCC were both funded by fees they collected from their member banks, which gave them an incentive to try to woo banks into their orbit by offering lighter enforcement. At the time, Republicans controlled Congress as well as the White House, and the prevailing mood in Washington was averse to regulation; the FDIC’s 2003 annual report featured a photo of bank representatives helping regulators slash “red tape” with a chain saw.

  But if the scrutiny of commercial banking was problematic, the rest of the financial system was under even less scrutiny, with none of the constraints of bank regulations and virtually no safety and soundness oversight. Investment banks were monitored by the SEC, which was much better equipped to focus on investor protection issues than on the financial health of the investment banks, and imposed few constraints on leverage. Money market funds, which functioned much like traditional banks, offering deposit-like instruments to their customers, were also under the SEC umbrella, but without the capital requirements or constraints on risk taking faced by banks. Hedge funds were growing fast without any meaningful oversight or leverage limits. The same was true of large finance companies such as GE Capital and the General Motors Acceptance Corporation (GMAC). Fannie and Freddie had their own compliant Washington regulator, which allowed them to build up a huge portfolio of mortgage assets without requiring them to hold much capital. The CFTC was supposed to monitor future
s traders. State insurance departments were largely responsible for insurance companies, even though AIG and other insurance firms had evolved into multinational financial conglomerates.

  Again, none of those firms were subject to strong supervision or capital rules, and none of them would have access to the Fed’s safety net in case of a run. Unless we wanted to be like the proverbial drunk searching for his keys under a streetlight because it was dark everywhere else, we couldn’t just worry about risk within the banks we supervised. We had seen during the LTCM crisis in 1998 how quickly the potential failure of a financial firm could infect the overall financial system—and there was now much more risk outside the banking system in institutions with much more leverage.

  But the Fed had no legal authority to impose limitations on risk outside the banks we supervised. We could use private and even public pressure to try to persuade nonbank institutions and their supervisors to address problems, but I wasn’t impressed with the power of those tools. I asked my colleagues at the New York Fed to analyze the capital buffers at Fannie Mae and Freddie Mac, and they concluded quite reasonably that the buffers were thin. I noted in a speech that Fannie and Freddie’s “regulatory framework, capital regime, and … internal risk management framework need to be upgraded to a standard more commensurate with their risk profile and the risks they present to the system.” But it was Congress that set and defended their low capital requirements, and Fannie and Freddie had a lot of friends in Congress. They ended up with $75 worth of leverage for every $1 in capital, the corporate equivalent of a $4,000 down payment on a $300,000 home.

  We were even more concerned about the big investment banks, which New York Fed Vice President Mike Silva described to me in an email as “the 800-pound gorilla” of our financial stability efforts. It was obvious that the SEC was poorly equipped to judge how safe they were, or what would happen if one of them ran into trouble, but we couldn’t do much more than politely raise awareness and urge vigilance. “Since [our] relationship with the SEC is chilly at best, a very subtle approach is probably the only one that has any chance of actually achieving results,” Silva wrote.

  Since simply expressing concern was not a strategy, I tried to figure out a way to get more traction across the entire financial system. In late 2004, I asked Jerry Corrigan, the former New York Fed president, to lead a new Counterparty Risk Management Policy Group of industry officials. I thought of Corrigan as a kind of John Madden of finance—big, gruff, old-school, a well-respected student of the game, with a similarly animated rhythm to his commentary. Corrigan had co-chaired the original counterparty risk group after LTCM imploded in 1998, and had produced recommendations to help firms prepare for the failure of a major financial player. The new group examined the state of risk management across investment banks, foreign banks, hedge funds, and commercial banks. Incidentally, despite his anti-regulatory instincts, Greenspan was fully supportive of this work. When I asked for his blessing before launching, his response was: “That’s what you’re supposed to do.”

  The Corrigan group provided a venue for risk managers and risk takers from all the relevant parts of our financial system to sit around a table and talk about vulnerabilities. In 2005, it issued a 273-page report that included dozens of sensible recommendations to reduce risk. It was leaning against the wind, too, promoting caution at an incautious moment. We encouraged banks to adopt its recommendations, and shared them across the regulatory community.

  Over the next two years, we used these findings to initiate a new round of horizontal risk reviews, applied not just to banks but to investment banks and foreign banks that were large players in our markets. We brought foreign agencies and our colleagues at the SEC into the process, hoping to improve coordination and help the firms in our various jurisdictions prepare for a severe recession or financial crisis. We were trying to create a kind of race to the top, pushing state-of-the-art stress testing and more conservative risk management throughout the system, encouraging peer pressure that might offset the competitive pressure to take on as much risk as possible. We also tried to rein in nonbanks a bit through our supervision of the banks that lent them money, which we called the “indirect channel” of supervision. We had some success limiting leverage by unregulated hedge funds by pressuring the banks that funded them, but we never launched an all-out effort to starve the most dangerous parts of the nonbank sector of financial oxygen.

  Our efforts weren’t much of a countervailing force against the vast sums of money looking for opportunity in an expanding global economy. Credit was exploding worldwide, while default rates were low and trending lower, making borrowers look unusually safe. Institutions outside the Fed system that could take on more leverage found it easy to do so; private markets enthusiastically financed the massive migration of risk and capital toward nonbanks. In retrospect, these rosy assumptions about the future look delusional, but the triumph of hope over fear was highly profitable until it wasn’t.

  We tried to push back against the frenzy, but our impact was modest. We sought improvements in risk management, but we couldn’t compel change beyond the banks we regulated, and we missed major weaknesses in some of those banks, too. One of the best examples of our accomplishments, but also of the limits of our accomplishments, was in the realm of derivatives.

  BY 2005, the market in over-the-counter derivatives had quadrupled since Brooksley Born had started warning about them seven years earlier. Investors were gobbling up an array of new products, from “credit default swaps” to “synthetic CDOs.” But there was still no single authority over derivatives, or over the banks and securities firms that manufactured and sold them. All this worried me.

  In some ways, derivatives were just another form of risk taking, like simple bank loans or any other financial transactions. They could be useful tools for companies that needed to hedge risks, helping importers and exporters protect themselves against changes in exchange rates, or helping banks protect themselves against the bankruptcy of a counterparty. On balance, they seemed likely to add to the stability of the financial system, dispersing, diversifying, and balancing risks across individual firms and across broader markets. After all, there had been financial crises for centuries before derivatives even existed. Simple banking had proven to be pretty dangerous, and the hedging benefits of derivatives could reduce some of that risk.

  But derivatives did create new dangers. If you were making a loan, and you were confident you could hedge some of the credit risk of that loan, you might be tempted to make a larger and riskier loan. And the instruments themselves often had leverage embedded in them, so investors could be exposed to greater losses than they realized. Firms weren’t required by law to post any collateral (or “margin”) to make derivatives trades, and the market wasn’t requiring them to post much, either. This meant fewer shock absorbers for the system if those trades went bad. That’s why Warren Buffett had called derivatives “financial weapons of mass destruction.”

  One of my concerns was that much of the derivatives market was untested by crisis. Nobody knew how the new products, particularly credit derivatives, would perform in a bad state of the world. They had been launched in a period of stability, and I feared their short history of relatively favorable performance could breed overconfidence and instability. Firms that considered themselves well hedged might discover that their insurance wasn’t as solid as they thought. There were millions of contracts among tens of thousands of individual counterparties, and little capacity to monitor trends and vulnerabilities in the overall market. Idiosyncratic corporate difficulties that wouldn’t have had larger implications before the explosion of credit derivatives—such as a downgrade of General Motors bonds, or the bankruptcy of the auto parts supplier Delphi—were already taking on what the Wall Street Journal called “a hold-your-breath air” as markets sorted out billions of dollars’ worth of side bets in real time.

  In general, derivatives made it tougher for firms to figure out what was going on—what expos
ures they had, which counterparties might be in trouble—and anything that added to uncertainty during a crisis could accelerate a panic. I said in my speeches that while derivatives might help prevent financial crises from erupting as frequently, they might also make the crises that do erupt more damaging and harder to contain.

  To make matters worse, the infrastructure behind the derivatives markets was a mess. My friend and former colleague Lee Sachs, a Bear Stearns veteran from the 1980s and early ’90s who had worked at Treasury and was now back in the financial industry, explained to me shortly after I arrived in New York that the plumbing of the derivatives business was hopelessly obsolete. Greenspan later called it a twenty-first-century industry reliant on nineteenth-century practices. Complex orders were scribbled down in pen and faxed to unattended machines. Trades remained unconfirmed for months, and dealers often reassigned them without notifying counterparties. So investors didn’t know who was on the other side of their trades—again, a significant source of uncertainty. In a crisis, nobody would have any idea who owed what to whom, or whether whoever owed it would be able to pay—a recipe for a panicky sell-off and a deeper crisis. As Corrigan told me, it would be like trying to untangle a vat of cooked spaghetti.

  This was not a problem the market could solve on its own. Financial infrastructure is a public good much like transportation infrastructure, benefiting businesses that are nevertheless unlikely to build it on their own. It wouldn’t make sense for one firm to upgrade its back-office operations, address its backlogs of unconfirmed trades, and invest in electronic trading unless all the major derivatives dealers did the same. We decided to try to induce all of them to do it together. The fourteen top dealers handled 95 percent of the derivatives market, and even though the New York Fed supervised only two of them, Citi and JPMorgan, we invited all of them to one of our more imposing conference rooms in September 2005. We invited their regulators, too, including British and German authorities. Lloyd Blankfein of Goldman Sachs dubbed the group “the Fourteen Families,” as if they were warring Mafia bosses meeting in a back room to discuss their common interests.