Slashdot Mirror
Incorporating Human Behavior Into Wall Street Mathematical Models
After watching the stock market struggle for the past year, financial experts from Wall Street and academia are putting more effort into bringing behavioral modeling into their complex financial calculations. "The risk models proved myopic, they say, because they were too simple-minded. They focused mainly on figures like the expected returns and the default risk of financial instruments. What they didn't sufficiently take into account was human behavior, specifically the potential for widespread panic." Analysts are looking at research from other fields to supplement the hard mathematics of risk assessment. "Financial markets, like online communities, are social networks. Researchers are looking at whether the mechanisms and models being developed to explore collective behavior on the Web can be applied to financial markets." Another avenue they're exploring is how we react to the spread of disease. Jon M. Kleinberg, a computer scientist at Cornell, said, "The hope is to take this understanding of contagion and use it as a perspective on how rapid changes of behavior can spread through complex networks at work in financial markets."
While I dislike how suddenly the financial markets have gotten back into these windfall risky investments, there's little push to stop it, so I guess taking into account the kind of behavior that, you know, actual people would do, is better than nothing.
Most 'risk analyses' done by these things almost go as far as to assume everyone involved acts as Economic Man - the theory that everyone will always act in such a way as to best improve their position, in a 100% rational way. This is a pipe dream put up in economic theory and doesn't always work. If you assume everyone involved acts that way, then some possible outcomes - like the ones we saw in the past year - can't be the slightest bit possible, therefore the models that were being run at the time disregarded them. Of course, the models were wrong - because people don't act that way.
Consider what is sometimes called the Ultimatum Game - everyone's heard of it. Person A has a pile of money to divide between themselves and Person B. They split it, and Person B can either accept the division, in which case each gets their share, or reject it, in which case neither player gets one red cent and the money is lost.
Economic Man theory would say Person A should give the smallest possible amount (let's say 1%) to Person B, and keep 99%, or whatever the maximum share is, and that Person B should then readily accept, because they're better of taking something rather than nothing. In reality, when this "game" is tested, it doesn't work that way - if Person A doesn't offer enough to B (say, 20%), Person B tends to reject it, whether out of spite, or a sense of fairness. The responses change depending on how much money is involved, and culture (different countries and regions have different thresholds) and everyone seems to have their own threshold of course - but very few Person B's say "OK, I'll take one penny and Person A can have $99.99" even if that's what Economic Man would do.
Likewise, Economic Man doesn't see that much of a difference between, say, 10% chance of loss, or a 5% chance of losing double that amount and a 2 1/2% chance of losing quadruple - while real people tend to disregard a small chance of large losses, but be quite averse to a reasonable chance of smaller losses - they'd probably go for the last option, even if percentage wise the "odds" are the same.
Most of these financial models, in essence, assume people are Vulcans, when they're not - they're people, and no amount of economics saying "You should act like Economic Man!" is going to change that.
If they're going to continue using these models, a push to start getting them better is at least some progress.