Physics of Wall Street

Yesterday I attended a book talk given by a young professor here in the Philosophy department, James Weatherall, promoting his a recent work, The Physics of Wall Street. James told a rousing good story concerning many famous characters. Each chapter of the book focuses on a different character (personality) that fields, from a mathematical discipline such as physics to finance.

Weatherall himself was fascinated by the media coverage of the collapse in 2008. In particular he noticed that physicists, beyond all other professions, was especially vilified, and decided to investigate why that might be. This dovetailed nicely with Jim’s background in the Philosophy of Science, as it made for a strong case study on the social impacts of applying scientific models.

A strong part of the story behind the use of such models comes from DuPont’s success with Nylon. By assembling a team of engineers, marketers, and scientists, both applied and theoretical, into a team of forced and close collaboration, DuPont took nylon from a research interest into a product (stockings) in only 2 years. The unified and explicit goal of this team, enabled all parts to work together in parallel. Later, during the war, the Army took this team, nearly whole, and used its organizational structure as the model for their Manhattan Project, in order to rapidly develop the ability to mass produce plutonium for weapons. Thus, people like Thorpe, were raised in a new culture driven and organized to apply research as quickly as possible.

The physicist MFM Osborne hit a low point in his career, and turned his attention to finance. Upon the purchase of a Wall Street Journal, he observed page after page of numbers, a situation with which he was quite familiar. Using tools from physics to analyze and discover patterns in these numbers, he plotted some histograms and noticed familiar statistical patterns. In 1959, Osborne wrote a paper “Brownian Motion in the Stock Market” which built on Einsteins work explaining the probability distributions of floating pollen observed by Robert Brown. In this paper, Osborne determined that it’s not possible to make money on Wall Street, because the expectation of a stock price is 0. The argument, based on an observed balanced gaussian distribution in the price movement sequence, convinced Thorpe when he read the paper as a student. Interestingly, Osborne then spent the rest of his career trying to find where the assumptions in the random walk model break down.

As a student, Ed Thorpe had been part of a cohort of students thinking all the time about get rich quick schemes. At that time (1950s) Las Vegas was just getting it’s start, and attracted Ed’s attention. Despite everyone’s persuasion, he thought that he could beat roulette. Eventually, he came upon a paper publish by the US. Army, which, through tables of brute force calculation of all possible hand, showed that, although there was no way to win, there was a strategy that allowed you to lose the slowest. Thorpe, took this knowledge straight to Vegas, and played a few rounds. He certainly did lose slower than the other players, but noticed that one of the assumptions in that paper was flawed. Because cards are not placed back in the deck, each round is not independent. He took this observation and, together with Shannon at MIT, used the computer to create a system of card counting, which was eventually published in the book “Beat the Dealer, A Winning Strategy for the Game of Twenty-One”. Thorpe and Shannon also worked together on a wearable personal computer, that was used to beat roulette.

Eventually Thorpe also came across the work of John Kelly, who had answered the question: “How much do you bet when odds are in your favor?” Putting that together with the blackjack odds calculations, allowed him to give a talk, strategically titled “Fortune’s Formula” which gain him press, notoriety, and contact with professional gambler, Manny Kimmel. After playing for several months, Manny was convinced that Thorpe’s system worked, and agreed to bankroll a trip to Vegas. They did spectacularly well, turning 10k$ into 21k$ over a single weekend.

Paul Cootner, also at MIT, had been developing Osborne’s random walk model, and published this work in a famous book, “The Random Character of Stock Market Prices”. In it, Thorpe realized, is a model which allows for the pricing of options. The key insight being, that while nobody can predict what the value of a stock will be in the future, we can all agree on what the value of an option are at it’s expiration date. Putting the two together — Kelly’s betting criterion and option probability distributions — Thorpe was able to build a successful stock betting strategy (avg. 20% return over 28yrs).

Although James told a thoroughly good tale, my purpose for attending the talk, was not to learn about this captivating story (because I knew the gist of of it already, having read Poundstone’s book “Fortune’s Formula”), but to ask James whether he’d perchance interviewed any local hedge funds during research and interviews. He had not, but was able to point out that Ed Thorpe is still working on his stock gambling and lives in Newport, and he doesn’t mind giving the occasional talk on campus (last one was May 2012). Also, PIMCO is also based in Newport.