Quants Down But Not Out

In the aftermath of Wall Street’s epic collapse, few players have come in for more opprobrium than "the quants," those supposedly brainiac financial whizzes who designed ever-more complex investment products at the heart of the sub-prime mortgage debacle and subsequent financial chaos. These products, which included complex securities and derivatives tied to the value of mortgage debt and the risk of credit default, were the result of years of research and modeling undertaken by a group of math and physics PhD's (many from MIT and the University of Chicago) who came to Wall Street in growing numbers over the last two decades. They were lured not only by big money, but also the challenge of uncovering grand theories of finance economics akin to the elegant precepts found in physics and math. Originally called "rocket scientists" when they appeared on the scene over 20 years ago, the quants aimed to purge finance of messy human emotion and replace it with the dispassionate logic of mathematical equations and computer programs.

Now, amid the wreckage of Wall Street, the quants, once practically worshiped as the smartest minds on Wall Street, have become viewed in the public eye as poster children for hubristic science run amok in your 401K. One of their most prominent critics is a former quant himself, Nassim Nicholas Taleb, the best-selling author of The Black Swan: The Impact of the Highly Improbable.

"Quants are dangerous to society," Taleb told me recently. For over a decade, Taleb has railed against one of the fundamental concepts of quantitative finance, called Value at Risk (VaR), which is used to measure the chance that a portfolio of assets will lose its value over a short period of time. Taleb argues that since VaR cannot possibly forecast the occurrence of highly unpredictable events, the model is dangerous because it provides a false sense of security and justifies heaping massive amounts of leverage on top of investments that are much riskier than they appear. VaR is all the more pernicious, Taleb said, because it works well most of the time under "normal" market conditions. But when it doesn’t work, the results can be catastrophic. "It’s like saying that airbags work really well except when you have an accident."

Earlier this month, two of the field’s leading figures, Emanuel Derman, a Columbia professor and former Goldman Sachs managing director, and Paul Wilmott, whose online hub Wilmott.com is a water cooler for the quant community, issued what they called "The Financial Modeler’s Manifesto," a new code of conduct for quants, in the wake of the financial collapse. "We do need models and mathematics – you cannot think about finance and economics without them – but one must never forget that models are not the world," Derman and Wilmott wrote. "You must start with models and then overlay them with common sense and experience."

Interviews with quants, both in finance and academia, paint of picture of chastened field engaging in a round of self-examination. To be sure, mathematics and computer programs will be part of the future of finance, but the financial meltdown has led many quants to revisit old questions about the unpredictable role of humans in finance and economics. For example: How do you model dishonesty? Unscrupulous mortgage brokers peddling home loans to people they knew couldn’t afford them are at the root of the sub-prime meltdown. How can models account for that? How about ineptitude? The federal government’s response to the crisis has been characterized by a very small group of people making enormously consequential – yet sometimes seemingly arbitrary – decisions.

Dale W.R. Rosenthal, a University of Illinois finance professor who worked on Wall Street at Morgan Stanley and Long-Term Capital Management, (the storied hedge fund stocked with quant "geniuses" that blew up spectacularly in 1998), remembers standing on the floor of the Chicago Mercantile Exchange last November watching a speech by Treasury Secretary Henry Paulson on the monitors. "I was watching the grain markets close when Paulson said the government should restrict any instrument with the potential for systemic risk," Rosenthal recalled, by email. "That is a foolishly broad statement and I watched the markets tank as Paulson spoke. I remember thinking, 'Just shut up, you're tanking the markets!'  But he didn't. It shows a real lack of understanding about markets that he thought he could eliminate all risk by fiat."

When the SEC issues a snap ban on short-selling financial stocks, or the Treasury Department decides to allow an enormous investment bank to fail, while throwing a lifeline to other major financial institutions, there is simply no way for quantitative models to "understand" such rare and unpredictable events.

"Finance will never be physics," one real-estate derivatives specialist at one of the few remaining major Wall Street banks told me. (Like many quants I spoke to, he insisted on anonymity.) "Physics can be boiled down to laws which are universal in nature and will always hold. In finance, that won’t work. Financial engineering got out ahead of itself because people thought they had found out the laws of everything."

Read Part II: Rise of the Machines
Despite meltdown, bright future for AI in finance.