Benjamin Graham, on Security Analysis, and the Efficient Market Theory

[LinusKS]

I've come across three different statements of EMT so far - other than the "strong," "weak," etc. -

The first is that information is quickly, or immediately incorporated into market prices.

For practical purposes, I think this is pretty much indisputable. Certainly you couldn't read an article about gas shortages, and call your broker, and hope to make a profit. By the time you placed your order, prices would have long ago responded to the story. Arguably, prices would have reponded even before the story hit the presses, even if you were somehow the first person to get the paper, since insiders, or friends of insiders, would have moved already.

The second statement is that equities are always fairly priced - or alternatively, that they always closely approximate intrinsic values.

This is a much more interesting statement, and much more subtle & complicated.

The third statement, sometimes said to follow from one or both of the others - is that you can't systematically make a beat the market - that excess profits are the result of luck or good fortune.

What's interesting is the the first two statements are saying entirely different things. And that while the third necessarily follows from the second, it doesn't from the first.

The problem with the second statement - that markets always fairly value securities - is that there any number of historical examples of markets not fairly valuing securities. The dot-com craze, the tulip bulb mania, etc. are all examples of speculators driving prices well beyond any reasonable approximation of value.

The problem of speculative bubbles, though, is that they're very hard to take advantage of.

The Graham & Dodson investor stays away from speculative bubbles, because he buys securities that represent value. He might be able to avoid losing money in a bubble, but he can't profit from one. The problem for a fundamentalist is that he can short bubbles, and he can buy put options, but those things don't pay dividends, and you can't hold them indefitely. Those two things - the willingness to hold a security indefinitely, and ability to profit from an investment, whether or not the market gets around to valuing it correctly - are what makes value investing so profitable in the first place.

To put it differently - that a bubble will pop might be pretty close to a certainty. When it will happen might be something else entirely. Speculators might well bankrupt you, long before they run out of money themselves.

Buffet - who stayed out of the internet bubble - did not, to my knowledge, figure out a way to profit from it.

In other words, it's possible markets might sometimes - or even often - behave irrationally, without giving investors opportunities to take excess profits, so long as markets confine themselves to excess exuberance & confidence.


BTW, according to Buffet's recent newsletter, he's beat the market in all but 6 of the past 40 years. That's a pretty frikken good record. According to my - possibly entirely inaccurate calculations - the chances of doing that well or better is only about 1 in 25,000.