July 29, 2005
Can Markets Both Be Efficient and Make Mistakes?
Personally, I think this is one of those questions that answers itself. If one believes in efficient markets, one has to admit that something very mistake-like happened in 1929, 1987, 1998, 2000, and 2001.
And yet, many of us would still argue that it's difficult-to-impossible for most investors to beat the market.
How can this be true?
First off, let me point out that you don't have to believe in Modern Portfolio Theory to believe in Efficient Markets. These really are two separate things. I personally think Modern Portfolio Theory is largely a crock, based on a distribution of returns that just doesn't hold long enough to be useful.
That, however, is an argument for another day. For the moment, let me just reiterate that one can dismiss MPT, and still believe in Efficient Markets.
Not perfectly efficient markets, mind you. Data clearly shows leakage prior to major announcements, it shows momentum, mean reversion, and all sorts of other tantalizing bits that people have been following for years. That few, if any of these turn out to be investable doesn't refute their existence, and may in fact account for their persistence.
Still, how can an efficient market overshoot targets time after time?
I think the answer comes from another book that's awaiting review: The Wisdom of Crowds by James Surowiecki. He points out two things about collective, aggregated, independent crowds. First, that they are consistently smarter than even their smartest members; although individual members will inevtiable beat the crowd each time, these winners will invariably differ from run to run. And second, the decisions, the valuations, if you will, have to be independent.
The first observation results from efficiency under normal conditions. You and I may beat the market from time to time, but on the whole, the market is smarter than either of us.
The second observation is what causes mistakes. Since at any given moment, we assume that the market is smarter than we are, the tendency is to go along for the ride. And most of the time that's the right thing to do. The warning sign is when people stop looking ahead, and start looking side-to-side at what everyone else is doing.
That's when decisions stop being independent. First, people start looking for a greater fool. Then, they all head for the exits. That's what happened in 2000 during the tech bubble, and it's also what happened in stages to Long-Term Capital Management in 1998.
Even during a run, the market is smarter than we are. Even during Tulipmania, the market at each point is probably smarter than we are. There's a point where decisions stop being independent. But since nobody really knows when the fever breaks, it's hard for even the smart, cold-blooded investors to forgo certain profits and get off the merry-go-round before he's thrown off.