BaffledExperts by Adam Norman

The efficient markets hypothesis

The Efficient Markets Hypothesis says, if I am correct, two things:
1) All available information is priced into the components of the market
2) Outsiders cannot win.

Point 2 follows from 1; insider information is unavailable to the rest of us. It is also illegal to act upon. 2 follows because information travels fast, and people have the incentive to both research and act upon valuable information. Thus, if a company looks likely to double its profits in the coming year, early-movers will bid up the price of the stock until it is no longer a better deal than any other stock on the market. Therefore, unless you are one of the early movers (and you're not), you can't win. If you do win (or lose, for that matter) it is only because you happened to act on unavailable (or unknowable) information--in other words, because you were lucky.

There has been much discussion lately about whether the EMH is correct. The recent stock market crash is said to disprove point 1. Wild swings in the market show, supposedly, that the market is not 'rational' or 'efficient'. Why this is so is not quite clear; the EMH does not say that everyone will win or that prices will be stable. It says that all available information is priced in. Unavailable information (and information about the future is certainly unavailable) can still influence prices.

Behavioural economics is supposed to offer an alternative. By understanding how people work, we are able to better predict their actions. But, this implies two things, either of which makes the EMH, or its corollary, relevant again.

The very premise of Behavioural Economics must be this: If BE works, we can understand the collective actions of irrational actors using reason. We can predict what irrational people will do. Of course, as soon as behavioural economics gets polished up, people will use it to invest in the market. Thus, the market will again be 'rational'--although this time, in a meta way, or a meta-meta way: people will use reason to understand reasoning about the unpredictable actions of individuals. Or something like that.

Of course, there is another possibility: perhaps we will not be able to use BE to understand the market. Perhaps, fundamentally, people are unpredictable. If that's true, though, the market moves randomly, and point 2--that outsiders cannot win--is again true. Nobody can win except by luck in an unpredictable market. Placing money on the outcomes of unpredictable events is the very definition of gambling.

In the past, EMH2 followed by deduction from EMH1, but I think we have something stronger here: If EMH1 is true, then EMH2 is true. If EMH1 is false, then EMH2 remains true. If the market is rational, outsiders cannot win but from luck. If the market is irrational, outsiders cannot win but from gambling.