Advertiser Disclosure

NerdWallet Book Club: James Owen Weatherall, ‘The Physics of Wall Street: A Brief History of Predicting the Unpredictable’

Oct. 31, 2013
At NerdWallet, we strive to help you make financial decisions with confidence. To do this, many or all of the products featured here are from our partners. However, this doesn’t influence our evaluations. Our opinions are our own.

Critics often blame defective financial models for the financial crisis that started in 2008. As a result, the public has vilified the models’ creators. In The Physics of Wall Street: A Brief History of Predicting the Unpredictable, University of California, Irvine Professor James Owen Weatherall explores how physicists have forever changed finance. NerdWallet spoke with Weatherall about The Physics of Wall Street, his thoughts on policy-making and the future of financial modeling. 

You thought of the idea for this book in 2008. Although this book isn’t about the financial crisis, tell me about writing this book during a tumultuous time for finance.


You’re right that this isn’t a book about the 2007-08 financial crisis, but the crisis was certainly in the background, and it was part of my motivation for writing the book. In Fall 2008, when I started thinking about this, I was just finishing my doctoral dissertation in physics. Like many people at the time, I was following the news of the collapse of Lehman Brothers and the government bailout of AIG very closely. I was struck in particular by what seemed to be a consistent theme in the press coverage of the crisis. Over and over I read that somehow “quants” had played some role. “Quant”, I learned, is short for “quantitative trader” or “quantitative analysts.” These are people who use fairly sophisticated mathematical models to understand Wall Street. Many of them have backgrounds in fields such as physics, mathematics, or computer science, and the models they use were said to have their roots in these fields. And, it was suggested, these models had somehow failed in 2008.

There was a lot of moralizing and I-told-you-so’s at the time. Critics suggested it was crazy to think mathematics or physics could help understand a complex human enterprise like financial markets. But I thought there had to be more to the story than just this. Where did these models come from? What were they intended to do and why would a

nyone expect that they worked? It was really the crisis that prompted me to dive into trying to understand the history that I talk about in the book.

Many people think that quants were responsible for the financial crisis. What role should financial models play in finance?

It seems to me that financial models are essential to modern finance. Financial models are more or less necessary for banks and investors to trade financial products known as derivatives, including things like options and futures. And derivatives — though they get criticized all the time — are actually a big part of how our economy works. Derivatives help companies protect themselves against uncertainty, allowing them to use their capital more effectively. So models are just a fact of financial life these days.

The real question concerns how we should think about these models in order to use them as reliably as possible. Here, I think, the history makes a big difference. Lots of investment professionals who know about financial models learned about them in a finance textbook, where they are often presented as sets of equations that tell you how the price of some instrument depends on various factors, such as volatility or expiration date. What is suppressed in these treatments is the fact that very strong assumptions about market conditions often play an important role in deriving these equations. Financial models are approximations of a very complicated world. And those approximations can be helpful, if we use them carefully; but if we do not use them carefully, or if we do not pay close attention to the assumptions underlying our models, then we can get into trouble. One thing we can learn from the history of financial modeling is just what assumptions the people who first came up with these models were making.

This sort of mistake about assumptions played a big role in 2008. Basically, the model that many investors, banks, and even credit ratings agencies were using to price financial products known as CDOs made it seem as though these products were worth a lot more than they turned out to be worth. When the mismatch was noticed, a lot of money disappeared overnight, leaving some major players insolvent. (This is the very short story of the collapse of Bear Sterns, Lehman Brothers, and AIG’s financial products arm.) Afterwards, many people pointed fingers at the model that mispriced these products — and even at the quants who had designed the model. But it seems to me that what really went wrong, here, is that major institutions continued to use a certain model long after the assumptions behind it became very bad. It really should not have been a surprise that the model didn’t work very well. I think the best, most charitable interpretation of this failure to change models is that many of the people who use models every day do not think very hard about the assumptions those models make.

Critics of financial models often argue that humans act irrationally and financial models are thus flawed. What do you think?

I think this sort of criticism is short-sighted. For one, while it is true that a large class of models in finance and economics rely on the assumption of that humans act rationally, it is not as though one has to assume this in order to using mathematical models, and plenty of models do not make this assumption. But really, the more important issue concerns whether we can understand when investors fail to be rational, and in what ways. This is something that has been studied a great deal recently in a field known as behavioral economics. Researchers in this field want to understand how we really make decisions, and they have discovered many systematic ways in which we fail to act rationally. Sometimes critics of mathematical modeling in finance cite behavioral economics as a reason to think mathematics and physics are useless for understanding markets, but I think this gets things backwards. Really, behavioral economics has shown how the assumptions underlying some models will consistently fail — and thus, that we should avoid using those models. But it has also pointed the way in how to construct more effective models that better account for what we now understand about real investor decision-making.

Many financial regulators don’t understand the tools and instruments that they oversee. How should we deal with this problem?

Financial regulators, such as those who work at the Securities and Exchange Commission and the Commodity and Futures Trading Commission, tend to be trained as lawyers, not economists or mathematicians. This means that they are often poorly suited for understanding both the products traded in some markets and the strategies that many banks and hedge funds use to trade them. And this leads to problems. Regulators are often several steps behind leading-edge financial innovation, and so they cannot adequately respond to emerging systemic risks.

Some groups, including the SEC, have made some recent efforts to combat this by hiring more quants, and others, such as the Federal Reserve, have always employed economists in addition to lawyers. So perhaps we are already moving in the right direction. But in a sense, the problems run deeper than just whether there are regulators who understand how derivatives modeling works. As it is, market regulators get headlines and praise for enforcement action, not for policy-making. For this reason, groups such as the SEC spend far more of their resources on insider trading, fraud, and various sorts of malpractice than on understanding how markets are evolving and trying to set policies that will minimize new risks.

One example that makes this very clear is the so-called Flash Crash of May, 2010. That day, markets fell about 1,000 points in a minute or so, only to rebound just as quickly. The SEC took nearly five months to figure out what had happened, mostly because they did not even have access to the sort of fine-grained market data that many traders base their decisions on. The SEC has subsequently introduced a new computer system, called Midas, to track this data in real time. But investors had been tracking this data for over a decade, and it was only after the crash that the SEC started doing so as well. It seems to me that we need to re-conceive the role of regulators, so that they stay out in front of the newest developments on Wall Street, rather than coming in only after crises to figure out what went wrong.

How can more physicists be incorporated into economic research and policy making?

First, I should say that there are lots of physicists and mathematicians who are already working on economic research. I talk about only a tiny fraction of such people in the book, and those I do talk about are mostly concerned with finance. This is a bit misleading: there’s an entire field of study known as “econophysics,” consisting of physicists applying ideas from physics to a wide range of economic problems. Many of these people would have a great deal to contribute to discussions of economic policy. But right now, they do not have a seat at the table. So the first step is for regulators and more traditional economists — who often overlook heterodox approaches, including those from other fields — to recognize that there are many researchers who have important insights to offer and, most importantly, novel ways of approaching problems.

Why aren’t you working on Wall Street?

Why would I? I have my dream job!

Read More from NerdWallet: