Efficient markets and football

It's important to have our priorities straight. Tonight President Obama gave a major economic policy speech, but he made very sure that he was done in time for the Packers-Saints kickoff. (I suspect that part of the President's political base feels betrayed because, in a concession to Republicans, his speech preempted a re-run of Big Bang Theory.)

But back to priorities, and by that I mean football. Tim and I are in good moods because Michigan State and Northwestern won their first games of the year, and this makes us happier than fans of Oregon, TCU, and Notre Dame. However, as a fan of Northwestern, I have less hope for the future than do fans of these other teams (and not just relating to the specific history of Northwestern's post-season performance). And this gets to the Efficient Markets Hypothesis (EMH).

The EMH has been getting kicked around lately, but its basic idea is simple. It is that arbitrage takes away free lunches, that we shouldn't expect to find consistently profitable trades in asset markets. The simplest form of the EMH (usually called the "weak form") is that charting prices should be useless, that is, the current price of an asset is the best predictor of its future price, and adding information about past prices doesn't help our prediction accuracy at all.

Behavioral economists find some evidence for deviations from the EMH based upon "loss aversion", where people are sometimes reluctant to sell an asset for less than they paid for it. Perhaps this explains something that the brilliant Nate Silver documents about the power of pre-season NCAA football polls. In some fun econometric work, he finds that the final football rankings depend on how the teams did in terms of wins and loses over the season...but also on how they were ranked in the pre-season poll, before the first game is even played!

So, yes, if Michigan State ends up ranked ahead of Northwestern, I'm going to blame it on those pre-season polls.