Subprime: A Predictable Surprise

Michael Watkins on how the subprime crisis provides another textbook example of the theory of predictable surprises and how to prevent the disasters they bring

In 2004, my colleague Max Bazerman and I published Predictable Surprises: The Disasters You Should Have Seen Coming and How to Prevent Them. We defined "predictable surprises" as problems that (1) at least some people are aware of, (2) are getting worse over time, and (3) are likely to explode into a crisis eventually but are not prioritized by key decision-makers or have not elicited a response fast enough to prevent severe damage. We supported our thesis with detailed analyses of the September 11 attacks, the collapse of Enron, and the war in Iraq.

While embraced by many, our work on predictable surprises came under predictable attack. "Hindsight is 20:20" the critics said. In response, Max and I were able to point to specific instances where we had accurately predicted major problems: Max gave congressional testimony on conflicts of interest in the auditing of public companies, and I had written about the dire consequences of an invasion of Iraq. But some critics remained unmoved.

Hindsight: The Road to Ruin

Imagine my delight then, when the financial markets offered up a near-perfect natural experiment that supported our theory: the collapse of the market for subprime mortgage-backed securities.

The contributing causes of the subprime collapse certainly are clear in hindsight:

• To avoid damage to the economy following the collapse of the Internet bubble in 1999, the U.S. Federal Reserve lowered interest rates which, in turn, lowered mortgage rates. The rate for 30-year, fixed-rate mortgages in the U.S. declined from 8.25% in January, 2000, to a low of 5.25% in January, 2003, and remained at low levels not seen since the 1960s for several years.

• A resulting surge in investment in housing, speculative and otherwise, helped drive median housing prices in the U.S. from $170,000 in 2000 to $240,000 in 2005. The resulting "equity cushion" permitted homeowners to increase their borrowing and encouraged still more speculation, creating a feedback loop.

• Securitization of mortgages into collateralized debt obligations (CDOs) decoupled mortgage originators (brokers and others) from the credit risks of the loans they were writing. At the same time, U.S. investment law shielded sellers of these securities from the legal consequences of fraud by originators. This introduced corrosive conflicts of interest into the system.

• The lure of easy money, creativity in the design of subprime mortgages (e.g., various forms of adjustable-rate mortgages with low initial payments), the inability of borrowers to fully understand the consequences of what they signed up for, and laxness in regulatory oversight fed a dramatic surge in subprime lending. A Wall Street Journal analysis has shown that the number of subprime mortgages written in California alone increased from 273,000 in 2004 (11.8% of total mortgages) to 573,000 in 2006 (29.4% of total).

• Mortgage brokers wrote loans of increasingly poor quality, including many that were apparently fraudulently obtained. In another conflict of interest, appraisers, whose livelihoods depended on getting work from front-line lenders and brokers, colluded in fraud by making higher-than-justified assessments of home prices.

• The risk of the CDOs based on these mortgages rose dramatically. Rating agencies, for reasons not yet fully understood, systematically underestimated the risks associated with these instruments.

• The inevitable first wave of defaults and foreclosures eventually took hold, triggering the collapse. (Another, potentially larger wave of defaults triggered by upward resetting of rates on ARMs may be averted by government-led efforts to prevent interest rate rises.)

• Virtually all the major investment banks were trapped by the magnitude of their investments, but also by the herd mentality (the formal term is "a collective action problem") engendered by short-term, relative returns criteria that are used to evaluate their performance. In a July 7, 2007, interview with the Financial Times, former Citigroup (C) CEO Charles Prince III unknowingly uttered his epitaph: "When the music stops, in terms of liquidity, things will get complicated. But as long as the music is playing, you've got to get up and dance. We're still dancing."

• When the bubble burst, the banks, to varying degrees, were caught with their pants down. Citigroup's massive exposure cost the dancing CEO his job. Morgan Stanley (MS) anticipated the collapse but not the breadth of its impact on CDOs, and it placed the wrong bets. Total losses for investment banks have been estimated to be as high as $300 billion. Of the big investment banks, only Goldman Sachs (GS) really dodged the bullet.

Foresight: Clues Along the Way

All this is quite clear in retrospect. But what's the evidence that the subprime meltdown was a predictable surprise?

First, as early as 2003, some farsighted observers were raising concerns about whether the subprime market was built, if you will pardon the pun, on a house of cards. Prem Watsa, CEO of Fairfax Financial Holdings (FFH), commented in the company's 2003 annual report, for example, that, "We have been concerned for some time about the risks in asset-backed bonds, particularly bonds that are backed by home equity loans, automobile loans or credit card debt…. It seems to us that securitization…eliminates the incentive for the originator of the loan to be credit sensitive…. There is $1.0 trillion in asset-backed bonds outstanding as of December 31, 2003, in the U.S.… What happens if we hit an air pocket?" Well we certainly have the answer.

Second, we have the wonderful example of Goldman Sachs' recognition of the impending collapse and moves to limit its exposure. As the Times of London noted in a Nov. 22, 2007, article, "Late last year, as the housing market motored along, David Viniar, Goldman's chief financial officer, called a 'mortgage risk' meeting in his 30th-floor Manhattan office. After the meeting, the bank's senior employees concluded that the mortgage lending industry was making a growing number of loans that borrowers could not possibly repay. So the bank decided to reduce its holdings of mortgages and related securities and to insure against losses on its remaining portfolio."

That Goldman got it right when so many got it wrong is a testament to the ability of the firm to manage risk. This ability is rooted in the culture of the firm and its finely maintained balance of power between traders and risk managers. It also demonstrates beyond a shadow of doubt that the subprime collapse was a predictable surprise. How galling it must have been to the leaders at Citigroup, Morgan Stanley, UBS (UBS), and the others that Goldman didn't dance off the cliff with. They would have been able to argue that "no one could have seen it coming." One fund manager not employed by Goldman told me that a Goldman representative informed him that they were seeing unexplained anomalies in their models more than two years ago and began to prepare for the consequences.

Lest we laud Goldman too highly, however, let's ask whether they continued to sell products based on subprime mortgages even after they began to limit their own exposure or shorted subprime CDO products they had recently sold to clients. One Dutch investor recently expressed to me his fury about this, noting that proprietary trading and investing by Goldman (and many other investment banks) created conflicts of interest that will almost certainly explode in yet more predictable surprises.

Lessons to Be Learned

What should we take from all of this? First that many surprises really are predictable—it's not merely that they look that way in hindsight. Second, complexity, conflicts of interest, and collective action traps are at the root of most predictable surprises. So we shouldn't be surprised to see financial markets blow up in the future if:

1. We allow naive consumers to make highly complex financial decisions under the influence of sophisticated sellers. (Pension decisions associated with defined contribution plans is one such area that immediately comes to mind.)

2. We allow pernicious conflicts of interest to take root and thrive in our financial systems and we don't provide sufficient, coherent regulatory oversight. Proprietary trading and investment by investment banks certainly fits this bill, and I believe the arrangement will result in future predictable surprises.

3. We continue to ensnare the people who invest money on our behalf in a collective action trap by judging them on short-term, relative performance criteria, rather than long-term sustainable ones.

Absent a change in this, there is no reason to expect that investment managers won't continue to dance together off the edge of cliffs.

Join a debate about the subprime mortgage crisis.

Before it's here, it's on the Bloomberg Terminal.