Blowing Bubbles

Friday, June 27, 2008

The turmoil in the housing finance market raises fascinating questions about economic behavior and investment strategy. Here are some brief observations about the mortgage tumult:

  • The element of surprise.  I have been preoccupied in recent years with intelligence failures, about which I have written several books. The subprime mortgage “crisis” follows a classic pattern that should help us understand the inevitability of such failures. Consider spectacular examples such as Pearl Harbor, the Tet Offensive, the Egyptian-Syrian surprise attack on Israel in October 1973, and 9/11—such failures to anticipate threats typically are not due to a lack of essential information or an absence of warning signs or signals. Rather, there is a lack of precise information concerning time and place, without which an effective response is impossible except at a prohibitive cost. Alarms over risky mortgage practices had been sounded for years, and ignored for years. A home buyer, or an investment bank buying mortgages, who had heeded those warnings when they were first made, or indeed for many years afterward, would have left a good deal of money on the table. No one knew the day or hour of the reckoning.
  • The bubbles.  There were two bubbles: one in housing and another in investment. The bubble phenomenon is intimately related to the phenomenon of surprise. A bubble begins when prices start rising at a rate that seems inexplicable in relation to demand. In conditions of uncertainty, there is a tendency to base expectations on a simple extrapolation: that is, if prices are rising, they are expected to continue to rise —for a time, but no one knows how long. There is a reluctance to employ restraint, to act as if they will not continue rising, for by doing so one is forgoing profits. As a bubble expands, the rational response is to reduce risk, without forgoing profit, by getting in and out of the market as quickly as possible. Meanwhile, the increased trading may continue to expand the bubble.
  • Ignorant or calculating?  It has been argued that the people who took out subprime mortgages with adjustable interest rates did not understand the risks they were assuming, and that the banks that bought mortgage-backed securities did not understand the risks they were assuming. I am skeptical. Suppose you have a low income, you would like to own a house, and a mortgage broker offers to arrange a mortgage that will cover 100 percent of the price of the house. What do you have to lose by accepting such a deal? Because you haven ’t put up any capital, you have no capital to lose if the house is lost because you cannot make your monthly mortgage payments. As for the banks, they rode the bubble for too long; but, to repeat, had they got out too early, they would have left a lot of money on the table.
  • Asymmetry of risk.  Bubbles are more likely to occur when the downside risk is less than the upside risk. My example of the asset-less home buyer is one illustration. For another, recall that the savings and loan crisis of the 1980s was exacerbated, or perhaps even created, by the fact that federal deposit insurance was not experience-rated: savings and loan associations paid the same rates regardless of the riskiness of the loans that they made with the depositors ’ money. Since the potential loss to depositors was truncated, financial institutions had an incentive to take excessive risks. CEOs of banks, as of other large, publicly owned firms, also face asymmetrical risk. If the bank is profitable, the CEO ’s compensation soars. If his investment gambles fail, he may be fired, but he will be consoled by receiving tens or even hundreds of millions of dollars in deferred income, stock options, or severance pay. This may have influenced the decisions of many banks to try to ride the bubble to the end.
  • Psychology at work.  Economists have become increasingly sensitive to the findings of cognitive psychology, which teaches that emotions and cognitive quirks influence us all and lead to behavior that often deviates from simple models of rational choice, important as those models are. Among the psychological tendencies relevant to an understanding of bubbles are these: optimism bias and a related belief in luck (there is no such thing —some people are lucky, but that is a product of randomness, not a quality that people possess in different proportions); herd behavior; excessive discounting of future costs; and difficulty in thinking sensibly about probabilities.
  • What is to be done?  In my opinion, nothing. There have always been bubbles; there will always be bubbles for the reasons I have been discussing. The Federal Reserve Board, though ably led and staffed, missed the mortgage bubble just as it missed the tech-stock bubble that popped in 2000. The proposals now on the table for resolving the subprime mortgage “crisis” or preventing future such fiascos include requiring that more information be given to prospective borrowers and freezing mortgage interest adjustments or taking other such measures to reduce foreclosures. I have argued that information is not the problem. Moreover, bailing out the borrowers, which is to say truncating downside risk, will set the stage for a future housing bubble. Nor does it excuse draconian measures to say that we must avoid a recession at all costs. A major depression, such as we last experienced in the 1930s, imposes immense social costs in the form of lost output; a recession involving some temporary unemployment may impose lower social costs than the governmental interventions designed to head it off.