Parsing the Efficient Markets Hypothesis

Eugene Fama, creator of the Efficient Markets Hypothesis, is mystified by the claim that the financial world took him seriously, and that that sort of market fundamentalism caused the financial crisis:

The premise of the Fox book ["The Myth of the Rational Market"] is that our current economic problems are largely due to blind acceptance of the efficient markets hypothesis (EMH)…

The book is fun reading, but its main premise is fantasy. Most investing is done by active managers who don’t believe markets are efficient. For example, despite my taunts of the last 45 years about the poor performance of active managers, about 80% of mutual fund wealth is actively managed. Hedge funds, private equity, and other alternative asset classes, which have attracted big fund inflows in recent years, are built on the proposition that markets are inefficient. The recent problems of commercial and investment banks trace mostly to their trading desks and their proprietary portfolios, and these are always built on the assumption that markets are inefficient. Indeed, if banks and investment banks took market efficiency more seriously, they might have avoided lots of their recent problems. Finally, MBA students who aspire to high paying positions in the financial industry have a tough time finding a job if they accept the EMH.

I continue to believe the EMH is a solid view of the world for almost all practical purposes. But it’s pretty clear I’m in the minority. If the EMH took over the investment world, I missed it.

Mike Konczal comments:

But he’s right. Most market participants don’t think markets are so efficient that they can’t get some alpha out of it. So who does believe in market efficiency? Is there a group of people who believe it significantly more than Fama believes people that participate in markets believe it? Yes: Our regulators and our government. ...

Justice Holmes once famously dissented that it’s a form of judicial activism to base our courts on “an economic theory which a large part of the country does not entertain.” It seems like the same should be said for our government and our regulatory bodies, especially as they try and figure out how to fix the mess that is the financial markets. And it’s worth noting that the founder of this economic theory, The Efficient Markets Hypothesis, doesn’t even believe that people actually in the financial markets entertain it.

The notion that regulators were more fundamentalist than market participants is perhaps shocking but seems accurate to me. (It should be noted, as Konczal does, that a minority of regulators wanted to regulate OTC derivatives, but it's not quite clear what that would have looked like in practice. Summers and Rubin are ridiculed now for thinking that banks wouldn't screw up the way they have, but is it reasonable to have expected them to have correctly made the opposite prediction? Hindsight and all that.)

The remaining question is this: if the strong-form version of EMH depends on rational actors making full use of aggregate information, and yet this picture does not accurately describe the world, is that because a). the actors are irrational; b). information is not complete in aggregate; c). individuals mostly act rationally most of the time, and information is mostly complete in aggregate most of the time, but black swans can kill.

If the correct answer is c)., as I believe it is, then Fama is correct: EMH is a solid view of the world for almost all practical purposes. This is the weak-form version of EMH. But under the same set of assumptions it is also true that investors should not behave as if it were. If all investors did, and made every investment decision by playing the averages, then they would miss opportunities to maximize the return on their investments. That, of course, would be irrational. But because no individual has full access to all aggregate information, and because some people act irrationally some of the time, we should expect EMH to never be true according to the tenants of EMH. So investors must act as if EMH were wrong, by doing so they demonstrate that it is wrong (or at least doesn't have strong microfoundations), but by missing the mark mostly randomly they demonstrate that weak-form EMH is actually correct approximately all of the time, and that strong-form EMH is approximately correct almost all of the time.

The problem begins when errors are not random. But this takes us right back to the question we just answered: are errors non-random because investors are irrational, because information is imperfect in aggregate, or both? We're going in circles. As Konczal says, this is a strange theoretical place to be. And as Matthew Yglesias pointed out awhile ago, there's no easy way to square this circle.

So what are the implications for policy? Should we expect regulators to be more or less rational than investors, and to possess more or less information? Interpreted one way, Konczal's argument can mean that regulators are both less rational and possess less information than investors. Interpreted another way, if regulators are rational they have no choice but to regulate as if EMH were true, but if EMH were actually true then there would be no cause for regulation. Is this an argument in favor of tying the hands of regulators (and investors) with inflexible rules in the hopes of preventing non-random errors from accumulating into a financial crisis? If so, what kinds of regulations are these? Should these regulations assume EMH or the absence of it?

A good answer to that question is worthy of the Nobel Prize.