"Investors should be skeptical of history-based models," Warren Buffett warned in Berkshire Hathaway's (NYSE:BRK-B) 2009 annual shareholder letter. "Constructed by a nerdy-sounding priesthood using esoteric terms such as beta, gamma, sigma and the like, these models tend to look impressive. Too often, though, investors forget to examine the assumptions behind the symbols."

"Our advice: Beware of geeks bearing formulas."

In this clip, Nobel Prize-winning economist Joseph Stiglitz discusses the problem with Wall Street's risk models in a recent interview I conducted in his office at Columbia Business School. Have a look (transcript follows):

Joseph Stiglitz: Wall Street got dangerous when it was managed by people who didn't understand the limits of those models and by the guys who are doing the models, many of them were not economists, but were math nerds, as you said. So that was a really dangerous combination. You have people who were the managers, who thought they were managing risk, and then the hired hands were just technicians. And the technicians were using their models, but didn't really understand also the nature of risk. They would put in a log-normal distribution or a beta distribution; one distribution or another without really understanding whether that was appropriate.

Didn't understand the risk of fat tail distributions. If they had managers who understood the risk, they would have said, What happens if the distribution has a fatter tail than you've assumed in assuming that log-normal distribution? That conversation didn't happen in the way that it should have happened.