The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution by Gregory Zuckerman
My rating: 3 of 5 stars

The Philosophy of Financial Markets

There are essentially two ways, two visions, two philosophies, of conducting inquiry in the social sciences. In one, rational behaviour is defined by some plausible propositions; behavioural data are then analysed; and people are shown to often act irrationally. In the other philosophy, the observed patterns of human behaviour are used to define an implicit standard of rationality which may be hidden and even unconscious. These patterns (or ‘signals’) are then used to predict future states. The first is an example of the philosophy of Rationalism, which holds that laws precede and produce facts. The second is an example of Empiricism, which claims that facts precede and produce laws. The intellectual battle about which of these views is better is ancient and still hasn’t been resolved - not just in the social sciences, but also in all scientific inquiry.

Within the social sciences, financial economists are generally Rationalists. They create models of economic choice which they then use to judge the rationality (which they call efficiency) of markets, and sometimes to exploit what they find to be irrational behaviour by buying or selling to correct the situation (making the market more efficient makes money, a win/win for the individual and society, they believe). Financial chartists (or technical analysts if one prefers) are Empiricists. They look for patterns (‘structure’ in the jargon) in the movements of markets prices from which they attempt to predict future prices (chartists don’t apologise; they are in it for the money). Financial economists and chartists view each other as fools and hucksters. Economists point to the absence of chartists’ theory as proof of their irrationality. Chartists claim the lack of theory as a virtue and deride the economists ignorance of the real world. They don’t want to second-guess the market, only to understand its inherent rationality.

Historically, Rationalist financial economists had the upper hand in academic circles and among the big names in financial trading.* Beginning in the early 1950’s, its influence grew rapidly as it was taught to generations of MBA’s who spread it like an infection throughout the world. The bias toward rationalism was so pervasive that it was the primary cause of the 2007 financial crisis, which demonstrated just how irrational rationality could be. In the way of these things, fashions changed in the perennial attempt to beat the market. Empiricism was in; Rationalism was out. Old-fashioned chartism entered the realm of Artificial Intelligence and became respectable (hence the euphemism of ‘technical analysis’).

And the new chartism works. No one knows why it works. It just does, as Jimmy Simons and Robert Mercer discovered to their enormous personal benefit. Neither knows all that much about financial markets, but they know about data, raw information from a staggering array of sources, within which are hidden patterns like the traces of gold at Sutters Creek or like intelligible messages buried within the gobbledygook of an enemy code. Markets didn’t need a theory; they provide their own theory if one pays enough attention to the detail. And computer technology was just the tool that was needed to sift, sort, and correlate all the detailed data one might collect in the search for the El Dorado of financial trading.

Financial economics worked, while it worked, largely because big investors felt compelled by academic theory to act rationally. Fund managers, banks, and other fiduciaries had a duty to act rationally on behalf of their clients. In the absence of any plausible alternative, professional ethics demanded adoption of the theory. The theory, therefore, became a self-fulfilling prophecy - and the prophecies came true until the world discovered that its rationality was no more than a conventional fiction. By avoiding the intellectual arrogance of presuming it knows better than the market, the new chartists can claim to be grounded in reality not economic fantasy.

The problem of course is that no one knows why the various correlations, connections, and intersections of data work (when they do). Empiricists don’t usually look for reasons. And when they do, it is typically to rationalise the conclusions their algorithms have already produced.** The algorithms which manipulate the data may contain an implicit theory but that doesn’t really bother the Empiricist. Nor does the lack of reasons for the various correlations. Coincidence or cause, the empiricist isn’t worried. What he does worry about is someone stealing his proprietary algorithms. The Rationalist benefits by the widespread use of his theory; the Empiricist by the strict secrecy of his programmes.***

Therein lies the Empiricist’s Achilles Heel. There is literally no reason to believe his correlations are stable. There is no way to test or verify hypotheses. Technically, there are no hypotheses. And no one aside from the proprietor is checking the validity of the findings of inquiry (thus violating a fundamental principle of true scientific inquiry). Correlations may change randomly and without warning. The enemy code, if there is one, might be altered entirely from day to day. Investors who employ the chartist strategy will never know if they are, quite literally, entering uncharted territory. On the other hand, society is considerably safer in the hands of chartists, as long as they act independently of each other based on their own algorithms (something the old-fashioned chartists did not do). Some may win while others loose; but they’re unlikely to all win or lose together, thus provoking systematic misery. That, however, until enough big investors discover similar correlations and interpret them as signals rather than noise.

Ideas have life cycles just like ladies’ fashion and gentlemen’s fascination with machines. When ideas become widely adopted, they are more accurately described as fads. If you miss one, don’t worry; they be another along shortly. The publication of this book probably announces the entry of high-tech chartism into fad-dom. no doubt there will be more and more success stories reported which will generate more and more interest, and produce more and more demand for data-mining and other empirical techniques. The failures, of course, will go largely unreported. At least until one big enough occurs that is worthy of newsprint, airtime, or blog space. I am eagerly awaiting first reports.


*I am not entirely unbiased on this subject. My great uncle was Fischer Black who devised the options pricing model which is arguably a central concept of financial economics in theory and in practice.

** Simons’s mathematical background seems to make him unaware of this as a fatal flaw. Numbers, after all, have stable relationships with each other. Once discovered, these relationships never vary.

***This is not strictly true in that Goldman Sachs, for example, has an interest in keeping its proprietary pricing models confidential. However, it is essential that Goldman’s also can convince its clients that the proprietary model conforms to a responsible financial theory. The general acceptability of the theory is what matters. The rest is a matter of client faith... or gullibility.

View all my reviews