Page 2 of 2
« Previous 1 2
It is possible that the recent credit collapse could have been averted, or at least anticipated, had banks taken an ecological approach to assessing risk. The collapse is widely ascribed to a market-wide overexposure to credit default swaps. (A CDS is essentially a wager that a third party will not default on its loan repayments.) And while an individual bank’s exposure to CDS-based risk may have seemed manageable, according to a risk manager charged with assessing it for his own institution, the most cursory analysis of the banking system’s exposure to such risk would have revealed an unsustainable ecosystem: The amount of money pledged between institutions in CDS contracts alone was $62 trillion, or more than twice as large as the combined GDPs of the US and the euro area, and roughly equal to the GDP of the entire world. When homeowners across the US found themselves unable to make payments on mortgages whose interest rates suddenly jumped, large numbers of CDS contracts started to be called in, and banks were unable to raise enough money to honor their myriad obligations.
Sugihara finds an analogous, instructive failure in fisheries management. When the US National Marine Fisheries Service seeks to maintain a sustainable population of, say, halibut, it assesses only halibut, not the state of the halibut’s predators and prey. In short, it ignores the food chain, the very system that makes life possible. Managing a species as though it lived in a vacuum, according to Sugihara, “doesn’t make any sense at all.” Acts of Congress have required that managers consider the larger ecosystem, but, Sugihara says, “nobody knows what that means or how to do it.”
The failure in fisheries management is in part institutional, he says, the result of an entrenched and stagnant culture within government. A more serious hurdle is a technical one, and it applies as much to ecology as to finance. Ecosystems and markets are difficult to model because their behavior is highly nonlinear — the state of the system depends on the complex and dynamic interaction of variables with one another. Effects in nonlinear systems, Sugihara says, are multiplicative rather than additive; two variables can be at a maximum value, but if a third variable is even slightly below or above a certain threshold, the two maximal variables are, in effect, multiplied by zero, and the system remains unchanged.
Sugihara cites the reproductive patterns of the damselfish, on which he and a student published a 1999 paper, to illustrate nonlinear dynamics. Upon plotting the larval population through time, they noted that it followed a boom-bust curve: The population remained at zero for a period, then spiked dramatically, then fell to zero again. What they discovered was that there were three variables that had to coincide at specific values for damselfish to reproduce successfully. First, damselfish spawn only during a full moon. Second, for the larvae to survive, they must be blown back to the protection of a nearby reef, so following the full moon, there must be a moderate wind in the right direction.
Finally, three days after spawning, water turbulence must be moderate, neither too strong nor too weak. “They’re born with a yolk sac,” says Sugihara, “that gets absorbed and disappears on the third day. If there’s no turbulence, there’s no contact with food, and they starve.” But strong turbulence can damage the larvae’s yolk sacs, killing them. “That’s the perfect storm,” he says, “this simultaneous occurrence of all these things. If one’s missing, it’s like multiplying by zero.”
A nonlinear system can exist in one configuration for a long period of time during which it is well understood, but it can switch gears suddenly, due to the interdependence of a multitude of shifting variables, and seem to become a totally different system governed by new laws. Superficially, everything has changed, but on the deepest level, nothing has.
With nonlinear systems, it’s a question, says Sugihara, of looking at the right scale — the population spikes of damselfish larvae are nonlinear and appear random, but the moon waxes and wanes at the same rate forever. Identifying the right variables and the correct scale in financial markets, though, is quite a bit more daunting than marking the phases of the moon on a calendar. “It could be that it’s not just the flow of money,” says Sugihara. “It could be that there are networks of personal relationships that may be a bigger factor, or the network that causes the flow of rumors.” It’s naive, he says, to think that Fedwire is the only relevant network.
Economics, then, is inextricable from society, and though it is often called the dismal science, it is driven by human desire, which, despite several centuries’ worth of work, have hardly become transparent to objective analysis. But better models can be built. Following the 1929 stock market crash, the financial system was overhauled, and firebreaks were put in place to insure deposits and prevent mass stock sell-offs. The recent crash will inspire similar reform, and it encourages Sugihara and Levin that Timothy Geithner, President-elect Obama’s choice for treasury secretary, was at the 2006 meeting on risk that featured their work. Perhaps Greenspan should have seen the recent credit collapse coming, and perhaps not; but, as Sugihara points out, even if he had, it might not have mattered. “People don’t respond unless they feel a lot of pain,” says Sugihara. “We now have had this painful experience. Hopefully some good will come out of it.”
Originally published February 3, 2009
Page 2 of 2
« Previous 1 2
