Learn about diversity at ASEE
ASEE would like to acknowledge the generous support of our premier corporate partners.




A global credit crisis exposes the pitfalls of financial engineering.

“It seems to me what our nation needs is more civil engineers and electrical engineers and fewer financial engineers.” —Former Federal Reserve Chairman Paul Volcker, October 2008

The term “financial engineering” is often loosely applied to all manner of Wall Street wizardry. But it’s actually an expert field, one that combines financial theory, engineering methods, math, and computer programming. Financial engineers – or “quants,” as they are also called -- are the original architects of quantitative finance, computer-driven trading that uses complicated math formulas to predict risk and to price options and the derivatives based on them. And it’s the quants, many of them graduates of top American engineering schools, who are now taking the heat for their apparent starring role in the still-unfolding meltdown of the world financial system.

That system broke down spectacularly in August 2007, when America’s subprime mortgage market crumbled. It was a massive event, and most big banks, funds, and investors didn’t see it coming. Instead, for years they eagerly bought billions of dollars’ worth of exotic financial instruments, like mortgage-backed CDOs, or collateralized debt obligations: bundles of mortgage debt sliced and diced into securities that often had AAA ratings. And, for a while, it made them very rich. Now, however, most of those “assets” are next to worthless, there’s a global credit crisis, and several major banks are history. And U.S. taxpayers had to rescue insurance giant American International Group, which lost billions in bets on credit defaults. The International Monetary Fund predicts the total cost of all that irrational exuberance will hit $1.4 trillion.

So, kill all the quants, right?

Not so fast, say many financial engineering academics, who claim the fault lies less with these specialists in computational and financial software than with a banking system that became too addicted to greed, then hijacked the work of the financial wizards, placing more faith in models than any right-thinking person should.


“I don’t think financial engineering is to blame. I think human overenthusiasm is to blame, and this time around, some part of it found a vehicle in relying on models,” says Columbia University professor Emanuel Derman, who knows this topic well, and from both sides. Derman runs the financial engineering master’s program at Columbia, but he used to head the Quantitative Risk Strategies group at investment bank Goldman Sachs and is the author of the 2004 bestseller My Life as a Quant.

“Quants,” Derman says, “have less power than people think. They build the models but often aren’t responsible for the inputs to the models or their use.”
That’s a point that’s also strongly made by Princeton University’s Ronnie Sircar, who says big banks often use quants as “fig leaves.”

“These are products that traders created. These were not designed by quants. No one with any math sense would have created these. Traders run the show, and quants serve at the pleasure of the traders,” says Sircar, a professor of operations research and financial engineering. He offers the analogy that blaming financial engineers for this crisis is like blaming Pentagon staffers for the botched planning of the Iraq War. Sircar reckons that plenty of officials worried about the consequences of the U.S.-led occupation but nevertheless had to accommodate leaders hellbent on war. David Spiegelhalter, a mathematician and risk expert at the University of Cambridge, also doubts that math-savvy quants would have placed so much faith in financial models. “They would understand the limitations of models,” he says. Yet the finance community was doomed by “an overbelief and overreliance on models and their assumptions.”

Financial models, as opposed to those used to predict, say, climate change, are particularly susceptible to the capricious nature of human activities, which are inherently unpredictable.

“Weather is not going to be changed by a lot of people all doing the same thing at the same time,” Spiegelhalter notes, but financial markets are. Adds Columbia’s Derman: “Financial models don’t work like physics models.”

“Traders run the show, and quants serve at the pleasure of the traders.”

For instance, much of quantitative finance is based around the Black-Scholes model. This Nobel Prize-winning construct essentially argues that securities prices behave not unlike Brownian motion, with the same kind of randomness and volatility as particles suspended in liquid. “Here, unfortunately, lie dragons,” Derman wrote in a recent paper, “On Models.” Or, as he tells Prism: “It assumes markets move smoothly and continuously with a constant volatility ... and that under those circumstances you can synthesize an option by trading in stock and riskless bonds. But markets don’t move that way.”

Traders and quants realized that lesson long ago and learned to accommodate the model by filtering in implied volatilities. But those manipulations amounted to the factoring in of opinions, and too regularly, those opinions proved overly optimistic. In his paper, Derman writes that “the gap between a successful financial model and the correct value is nearly indefinable . . . and so model success is temporary at best.”

Financial models are also at the mercy of history. “Our ability to predict risk is limited to what we have seen,” says H. Nejat Seyhun, interim director of the financial engineering program at the University of Michigan. And that’s a weakness.

“Rumsfeld was right: There are unknown unknowables. You can’t really know what you don’t know, so you can’t factor it into models,” Spiegelhalter says, referring to a famous quote from former U.S. Defense Secretary Donald Rumsfeld. Quants refer to these outlier events, these rare occurrences, as “fat tails.” But they’ve recently been popularized as “black swans” after the bestselling book of the same name, The Black Swan by Nassim Nicholas Taleb, chair of the department of finance and risk engineering at New York University’s Polytechnic Institute. The book’s title draws from the assumption Europeans held for centuries that black swans were a fiction — until explorers discovered these creatures in Australia. Taleb argues that analyzing history is a flawed means to predict risk because there will always be black swans that show up to shatter hardened perceptions. Indeed, he says, rare events rule in finance, and yet models always underestimate them.


STOCK EXCHANGE BUILDINGFor instance, because the United States had not had a housing market crash in more than 50 years and because the country’s population was growing at 1 percent a year, most models factored in the likelihood of a housing market collapse as almost negligible, perhaps no more than 5 percent. “They just got carried away and forgot what goes up must come down,” Seyhun says, but adds, nevertheless, that that wasn’t an unreasonable assumption, based on historical data. Spiegelhalter agrees but points out that what should have been stressed to buyers, and wasn’t, was that if that rare event did occur, the outcome would be catastrophic.

But that kind of caveat emptor marketing was pretty much nonexistent in the go-go years before the subprimes blew a gasket, a collapse that cascaded throughout the housing market. The overoptimism was very likely the unfortunate result of sellers and buyers both not fully understanding the limitations of financial models.

There was also a herd mentality at work: When all your competitors are making buckets of money on mortgage-backed securities, it takes a brave banker to bail out. “There is a saying that Wall Street is always looking sideways,” Sircar says. Still, he adds, it seems certain that models were manipulated by traders to ensure rosy outcomes.

The IMF thinks so, too. It claims that traders — knowing that some products were essentially toxic laden — also manipulated big rating agencies like Standard & Poor’s to hide that fact. They would, for example, take a package of CDOs and run it through the same analytical software used by the agencies. If the score was too low, they’d repackage things until the results were more favorable. So it was no surprise that when the agencies finally reviewed and rated those products, most were certified AAA, even though they contained thousands of mortgages that were clearly junk grade. Sheri Markose, director of the Center for Computational Finance and Economic Agents at the University of Essex in the United Kingdom, says the rating agencies were used to game the system. “It was just a big con game; it was nothing but a Ponzi scheme.”


So, if financial modeling can never fully measure risk or precisely anticipate rare events and will always have a limited range of applicability, is it a useless exercise? Not at all, these experts say. Ongoing research, bolstered by ever more powerful computers, should result in future models that are more robust. “We must account for extreme events, because they do happen,” Markose says, adding that it’s possible to design statistical models that show that the likelihood of an extraordinary event is more than negligible. In the United States, there are now dozens of schools that have master’s programs in financial engineering, beyond those at NYU, Columbia, Michigan, and Princeton. And Seyhun and Sircar insist that the demand for their graduates will increase, because the finance industry will have to rely on the skills of quants to reassure clients that it’s taking steps to better manage risk and costs.

But more industry regulation is needed, too. “I am a free-marketer, but mutual interest needs to be served; you need to have a level playing field and more transparency,” Markose says. That said, she admits that whatever new regulations are put in place, traders will look for ways to circumvent them.“It’s amazing what people will do to escape regulations.”

Indeed, though he knows that predictions involving human nature are fraught at best, Derman is sadly certain of one thing: “This is going to happen again.”


Thomas K. Grose is Prism’s chief correspondent, based in the United Kingdom.




© Copyright 2009
American Society for Engineering Education
1818 N Street, N.W., Suite 600
Washington, DC 20036-2479
Telephone: (202) 331-3500