by Adam Tooze
IV
With so many interests engaged, the expansion in US mortgage lending in the final burst of the boom was spectacular, not to say grotesque. Between 1999 and 2003, 70 percent of the new mortgages issued in the United States were still conventional GSE-conforming. With the end of the refinancing boom, that balance shifted. By 2006, 70 percent of new mortgages were subprime or other unconventional loans destined for securitization not by the GSE, but as private label MBS. In both 2005 and 2006, $1 trillion in unconventional mortgages were issued, compared with $100 billion in 2001. Fannie Mae and Freddie Mac were scrambling to keep up, purchasing $300 billion in nonagency securitized mortgages to hold in their own portfolios. The GSEs were driven. They were not the drivers. They were competing with upstarts like Countrywide, which in 2006 was responsible for originating 20 percent of all mortgages in the United States.51 They were competing with an ultrasophisticated investment bank like Lehman Brothers that had assembled an entire pipeline. In 2005 two thirds of the mortgages contained in Lehman’s issuance of $133 billion in MBS/CDO were sourced from its own subprime loan originators. A top Wall Street name was scraping the very bottom of the credit barrel.
The Rise and Fall of Subprime Lending in the United States, 1996-2008 (in $ billions)
Note: Percent securitized is defined as subprime securities issued divided by originations in a given year. In 2007 securities issued exceeded originations.
Source: Inside Mortgage Finance.
The message that this communicated down the food chain was simple: We want more mortgage debt to process, and the worse the quality, the better. By the magic of independent probabilities, the worse the quality of the debt that entered into the tranching and pooling process, the more dramatic the effect. Substantial portions of undocumented, low-rated, high-yield debt emerged as AAA. In any boom, irresponsible, near criminal or outright fraudulent behavior is to be expected. But the mortgage securitization mechanism systematically produced this race to the bottom in mortgage lending quality. It was the difference between the high yield of the underlying securities included in the collateral pool and the low interest that was paid to the investors who bought the AAA-rated asset-backed securities that generated the profit. From 2004, fully half the subprime mortgages being fed into the system had incomplete or zero documentation, and 30 percent were interest-only loans to people who had no prospect of making basic repayment.52
The ratings agencies would subsequently face penetrating questions about their complicity in this process. It did not help that they were paid by the banks for which they rated the bonds and that the big three ratings agencies competed with one another to offer the most streamlined and cheap route to AAA ratings. In their defense they would argue that they were operating tried-and-tested formulae that had the stamp of approval of the smartest economists in the land. But payment was by results. Fitch, which applied a risk assessment model that generated fewer of the coveted AAA-rated securities, found itself largely cut out of the subprime securitization business.53 As later congressional inquiries revealed, the ratings agency staff at Moody’s and S&P were clearly aware of the monster they were creating. As one ratings expert remarked to another in an e-mail in December 2006: “Let’s hope we are all wealthy and retired by the time this house of cards falters. :o).”54
Certainly some people were getting wealthy. The profits of the 1980s and late 1990s had been good in investment banking. Now everyone was making money. In the early 2000s 35 percent of all profits in the US economy were earned by the financial sector. At the very top it was dizzying. Though in the course of the 1990s they had converted to public companies selling shares to investors, the Wall Street firms continued to operate effectively as partnerships. The rule was that half of net revenue after interest costs was reserved for staff payments, the other half being paid to shareholders. The 2006 business year generated $60 billion in bonuses for the finance crowd in New York, and 2007 topped that with $66 billion in bonuses.55 For senior staff at the investment banks, that translated into payments in the tens of millions of dollars each. Richard Fuld, who drove Lehman’s dramatic growth as CEO from 1994, earned $484.8 million in salary and bonus between 2000 and 2008. That was staggering enough, but to understand the psychology of those operating the system, one has to appreciate that even the top investment bankers knew that they were not the real kings of the money game. Their remuneration paled by comparison with that of the hedge fund managers with whom they dealt in the prime brokerage, repo and ABCP markets. At the hedge funds and private equity groups, individuals could earn hundreds of millions, or even billions, of dollars per annum. In 2007 the six top hedge fund managers earned at least a billion dollars each in compensation.
Nor was the greed confined to the top. No doubt many mortgage borrowers were victimized by a process that systematically misled them and had every interest to do so. But once a real estate market shifts from a state of equilibrium to one of boom, everyone becomes a speculator willy-nilly. As capital appreciation came to be expected, the meaning of home ownership changed. Home owners, whether they liked it or not, were taking a speculative position. At the bottom, those who got on the housing ladder by taking out adjustable-rate, low-credit-score mortgages were speculating that their properties would rise in value so much that their equity would be sufficient to refinance on better terms. Those further up the ladder engaged in a fiesta of real estate speculation. In 2006 fully a third of new mortgages issued in the United States were for second, third or even fourth properties. In what became known as the “bubble states”—Florida, Arizona, California—the percentage was as high as 45 percent.56 Obviously, these were not the fortunes being made on Wall Street or on the Gold Coast of Connecticut, but real estate speculation had become a mass sport.
Funding of Outstanding US Private-Label ABS and Corporate Bonds in 2007 Q2 (in $ billions)
Private-label ABS
Corporate bonds
Amount
%
Amount
%
Total outstanding
5,213
100%
5,591
100%
Short-term funding
ABCP
1,173
23%
Direct holdings
MMF
243
5%
179
3%
Securities lenders
502
10%
369
7%
Repo
MMF
31
1%
42
1%
Securities lenders
165
3%
121
2%
Total short term
2,113
41%
711
13%
Source: Arvind Krishnamurthy, Stefan Nagel and Dmitry Orlov, “Sizing Up Repo,” Journal of Finance 69, no. 6 (2014): 2381–2417, table II.
All told, if we focus on that section of the market that was truly a product of the bubble, by the summer of 2007, $5.213 trillion in private-label asset-backed securities had been issued—that is, MBS generated from unconventional mortgages and credit card, student loan and auto debt. Of this total, the most dangerous mortgage component, subprime mortgage MBS, amounted to $1.3 trillion. Though this was “only” 12 percent of the total American mortgage market, the $1.3 trillion had been produced in a single surge since 2003. Of the total sum of $5.13 trillion, more than $3 trillion had been placed with long-term investors and $700 billion were placed directly with investment funds or investment banks. But $1.173 trillion were held by banks that funded them off balance sheet by issuing ABCP. As a result, ABCP had become the largest short-term money market instrument for investors looking to park cash for less than three months. The market for ABCP was larger even than for the short-term Treasury bills issued by the US government to manage its cash flow. If there was a channel through which the crisis in real estate could ramify outward to unleash the global financial crisis, this was it—ABCP, the place where private label MBS met wholesale funding.
V
Every year in August the elite of the central banking and monetary economics world gathers at a resort in Jackson Hole, Wyoming. In August 2005 the theme of the conference was not the crisis brewing in the US housing market but a celebration in honor of the outgoing Fed chairman, Alan Greenspan. Most of the presentations were appropriately upbeat. But one rang an off note. It was given by Raghuram G. Rajan, an Indian by birth but a fully paid-up member of the American economics elite, professor at the Chicago Booth business school and chief economist at the IMF. His paper bore the heretical title “Has Financial Development Made the World Riskier?”57 Rajan worried that the dramatic expansion of modern financial intermediation was building up a dangerous new appetite for risk. At Greenspan’s farewell party, the message was not welcome. Rajan was slapped down by Larry Summers. Wielding his full authority as former Treasury secretary, Summers introduced himself as “someone who has learned a great deal about this subject from Alan Greenspan . . . and . . . who finds the basic, slightly Luddite premise of this paper to be largely misguided.”58 To highlight risks in a complex, modern financial system, as Rajan was tactlessly doing, was to invite “restriction” and other “misguided policy impulses.” It would be like giving up air travel for fear of crashes.
This response from Summers—his warning that even to discuss risks within the system was to incite dangerous political reactions and tantamount to resisting technological progress—was emblematic of the attitudes that had driven a forty-year deregulatory push. The truly decisive early moves went back to the reemergence of a global capital market in the 1960s, the collapse of Bretton Woods, the deregulation of interest rates and capital flows in the early 1980s.59 It was those moves that unleashed monetary instability and precipitated Volcker’s interest rate shock. It was that turmoil which forced innovation in the housing market and gave rise to the hyperactive new breed of Wall Street investment banks. The competitive flows of capital thus unleashed drove all that followed. Rubin and Summers added a personal touch with the 1999 Financial Services Modernization Act, which released the final restraints on the fusion of commercial and investment banking. Within months of departing the Treasury, Rubin had returned to banking at Citigroup. Summers took slightly longer to join Wall Street, but barely a year after the clash with Rajan at Jackson Hole he had joined hedge fund D. E. Shaw as a part-time managing director.
Summers’s reaction to Rajan is all the more telling because the signs of stress in the world economy were so obvious. At the level of macroeconomic policy, Summers himself was willing to sound the alarm with his talk of a balance of financial terror. The commonplace recommendation to tighten fiscal policy might have helped. But this pointed the finger away from where the stress really was, in the financial system. Indeed, from the point of view of financial stability it might have been desirable if more of the AAA securities in circulation had been genuine US government debt rather than the products of financial engineering. In the final analysis, it was convenient to make the case at the level of macroeconomic aggregates. It was particularly easy to demand that a Republican president change his course. It was far less comfortable to question the house price boom and the giant Wall Street edifice erected on top of it.
The boosters far outnumbered the Cassandras and not just at Jackson Hole. The mortgage industry lobby did its job. David Lereah, the chief economist for the National Association of Realtors, chipped in with a book titled Why the Real Estate Boom Will Not Bust.60,61 Conservative pundits such as Larry Kudlow, economics editor of the National Review, railed against “all the bubbleheads who expect housing-price crashes in Las Vegas or Naples, Florida, to bring down the consumer, the rest of the economy, and the entire stock market.”62 Kudlow need not have worried. There was little sense of any urgency on the part of the authorities about limiting the boom.
Briefly, following Enron, there was a push for greater regulation. There was talk about requiring the parent sponsors of the off balance sheet SIVs to put more capital behind them. The threat alone was enough to bring growth in the ABCP industry to a halt. Moody’s warned investors that banks might soon face the end of one of their easiest funding sources. But in July 2004, as subprime was really hitting its stride, the regulators agreed to provide a permanent exemption that effectively allowed assets held in SIVs to be backed by only 10 percent of the capital that would have been required if the assets were held on the balance sheets of the banks themselves. This was particularly attractive for big commercial banks, like Citigroup and Bank of America, that were subject to relatively tight capital regulation, putting them at a huge disadvantage to the lightly regulated investment banks. It was following that regulatory shift that the ABCP market exploded from $650 billion to in excess of $1 trillion.63 By the summer of 2007 Citigroup alone was guaranteeing $92.7 billion in ABCP, enough to wipe out its entire Tier 1 capital.
More than the grand gestures of deregulation, like the 1999 act, it was this kind of apparently small-scale regulatory change that unfettered the growth of shadow banking. The same was true for repo. Traditionally, repo had been limited by the fact that the categories of assets that were exempt from the automatic stay in case of bankruptcy included only US government and agency securities, bank certificates of deposits and bankers’ acceptances. If those classes of security were offered as collateral in repo, in cases of bankruptcy they could be seized without delay and any losses made good. In 2005 the Bankruptcy Abuse Prevention and Consumer Protection Act gave creditors much stronger protection against defaulting borrowers, which ironically increased their willingness to lend. But it also expanded the repo collateral provided with special protection to include mortgage loans and mortgage-related securities. Not surprisingly, in the wake of the act there was a surge in bilateral repo secured on nonstandard assets.64
Could the Fed have contained the bubble through tougher interest rate policy? Greenspan’s cuts of the early 2000s had triggered the lending surge. Indeed, it had clearly been Greenspan’s intention to unleash a refinancing boom to help the recovery from the dot-com bust and the shock of 9/11. But what the Fed did not appreciate was the structural change in the mortgage machine the refinancing boom would trigger. Certainly by 2004 it was clear that it was time to raise rates. In seventeen tiny steps the Fed inched rates from 1 percent in June 2004 to 5.25 percent in June 2006. It was fine-tuning, not shock and awe. The mortgage boom continued undeterred, as did global demand for American safe assets and the expansion of the shadow banking sector. By the spring of 2006, to the alarm of many commentators, the result was that the yield curve was inverted. Long-term rates were below the short-term interest rates set by the Fed. This was usually a signal for trouble. It meant that the normal bank-funding model of borrowing short to lend long no longer made any sense.
In due course, the inversion of the yield c
urve might by itself have produced a recession. But it wasn’t Greenspan or Bernanke who killed the mortgage boom. It killed itself. By 2005 at the latest it was clear that low-quality mortgage debt was a ticking bomb. Many of the subprime mortgages were on balloon rates that would rapidly increase after a period of two or three years. In 2007 the typical adjustable-rate mortgage in the United States favored by low-income borrowers was resetting from an annual rate of 7–8 percent to 10–10.5 percent.65 As traders such as Greg Lippmann at Deutsche Bank realized, between August 2006 and August 2009, $738 billion in mortgages would experience “payment shock.”66 As the escalated interest payments hit, a wave of defaults was more or less inevitable. Once that began it was only a matter of time before house prices stopped increasing and the market turned. At that point, millions of speculative real estate investments would go bad. Families would lose their homes. Thousands of MBS would suffer default and whoever held insurance would get rich. Nowhere in Lippmann’s extensive document arguing for Deutsche Bank to short MBS was there any mention of Fed tightening. The subprime mortgage machine had a self-tightening timer built in. Unless house prices continued to rise at record rates, it would activate mercilessly and stop the boom in its tracks.
It was the first round of that tightening that was beginning to make itself felt in the most stressed communities across the United States already in 2006. Default rates were rising. It would not be long before the AAA rating granted to the lowest-quality CDO would be in doubt. To take advantage a growing band of contrarian investors began to build the “big short” positions that would make them famous. Those making the play included Lippmann at Deutsche Bank, J.P. Morgan and Goldman Sachs as well as a cluster of hedge funds. To build the position they bought CDS, derivatives designed to provide protection against default. Anticipating shipwreck, the holders of the big short were making advanced bookings in the lifeboat. They could either hold their insurance until the bonds failed and their payouts were due or they could sell their positions at a huge profit to lenders who were desperate for default protection. The question was one of timing and the problem was funding. Going long in CDS when majority opinion was still driving the market up was an expensive and nerve-racking proposition. You were on the other side of the last surge in ABCP and repo deals. At Citigroup in the summer of 2007, CEO Chuck Prince was still telling journalists that “as long as the music is playing, you’ve got to get up and dance. We’re still dancing.”67 The question was what would happen when the music stopped.