The Physics of Wall Street by James Owen Weatherall

As I hinted obliquely a little while back, I don't have a terribly high opinion of Wall Street or Wall Street traders. Given that, I'm not the most obvious audience for a book titled The Physics of Wall Street, and truth be told, I wouldn't've picked it up on my own. The publisher sent me an advance copy back in August, though, and I had a plane trip coming up to go play golf with some guys who work in business, so I thought it might make a decent read and possible conversation topic, and took it along.

I was very pleasantly surprised to find that it contains much more physics than Wall Street. That is, it's a book about the physicists who have migrated into the world of finance and made important contributions, and only sort of incidentally about Wall Street itself. It also takes a very physicist-like approach to the general problem of applying scientific methods to finance, casting it in terms of making and refining models. Beginning around the start of the 20th Century with Louis Bachelier, whose work anticipated a lot of future developments but was largely overlooked at the time, Weatherall argues that there is a progression from crude but mathematically simple models through many different stages of increasing sophistication and complexity. Each new model, in Weatherall's telling, attempts to accommodate features that the previous model had approximated away in order to make the problem tractable, features which can lead the model to go spectacularly wrong when pushed beyond its natural realm.

Some of the names involved are familiar-- Benoit Mandelbrot makes it in under a somewhat expansive definition of "physicist," and there's a chapter on Fischer Black of the Black-Scholes model of options pricing. Others, like Bachelier and Maury Osborne, are more obscure (at least to me). Each gets a chapter providing colorful biographical details and an outline of the main mathematical innovations they brought to the world of finance-- including multiple graphs, but few or no equations-- and how this pushed the field forward. The individual profiles are very engagingly written, and the general explanations are good. The math involved is far enough outside my own field that I wouldn't be likely to catch subtle mistakes, but the explanations do a nice job of walking the line between including enough detail to be rigorous and chasing readers away with too much theory.

I was also amused to find someone I've met appearing in this book, namely Eric Weinstein, who I follow on Twitter (though he's been quiet of late), and who I met at a Perimeter Institute workshop a few years ago. He's one of a couple of contemporary physicists-turned-financiers cited as making potential progress, in Eric's case through applying gauge theory to economics as a way of accounting for changing preferences over time. This is, as you might guess, one of the most mathematically hairy parts of the book, but Weatherall does a good job of making it (seem to) make sense.

I haven't really changed my overall opinion of the world of finance as a result of reading this-- obviously, since the rant I linked at the start of this post came from well after my golf trip-- but it's an enjoyable read. The only real weakness is that the book necessarily gets a little vague toward the end, because the more modern algorithms it discusses are both fearsomely complex and proprietary, still being used to make large sums of money for some ex-physicists and their hedge fund clients. As is characteristic of political writing, the epilogue in which he attempts to suggest solutions to the parlous current state of the world economy is kind of short on useful details (my own brilliant plan to fix the crisis by firing everybody on Wall Street and replacing them with unemployed physics students and post-docs who will work for a tenth the salary is still viable...), but then it's an epilogue, so I can't complain too seriously.

Anyway, I'm somewhat surprised to find myself recommending a book with this title, but if you're interested in reading about the history of science and finance, this is a fun book. It does not, alas, provide any foolproof investing tips, and in fact spends a good bit of time arguing that there's no such thing. Or maybe not, depending on whether a few of the modern ex-physicists he cites are actually sitting on algorithmic gold rather than the lucky monkeys whose random picks happened to write Shakespeare. But you can read the book for yourself, and decide which you prefer...