Saturday, June 6, 2009

Deep misunderstanding about efficient markets.

Is efficient market theory « passé »? That’s a very popular argument at the moment. The financial and housing markets boom and bust supposedly “prove” that these markets are inefficient and that the “Chicago School” of economics, where one finds many prominent advocates of free markets, has been proved wrong.

Here is the demonstration, according to a New York Times journalist (“Poking Holes in a Theory on Markets”, June 5):

“First came the rise of behavioral economists, … who convincingly showed that mass psychology, herd behavior and the like can have an enormous effect on stock prices – meaning that the market isn’t so efficient after all.”
Wrong: it just shows that individuals do not use perfect or complete information in all their decisions (limited rationality). This behavior in itself can be rational: it does not pay to spend much time, and the same amount of time, for all decisions, because the time spent at gathering information is costly and anyway most relevant information is not available for free. If these costs exceed possible benefits, then it is rational to decide under limited information. But individuals could adopt limited information behavior and nevertheless markets could be efficient, because the definition of an efficient market is a market that incorporates into prices all the available information. If individuals and firms decide that it is not worth producing more information, then the total amount of information incorporated into market prices will be limited to that amount.

But the article goes on:

“Then came a bit of more tangible proof: the dot-com bubble, quickly followed by the housing bubble. Quod erat demonstrandum.”
Wrong again: bubbles could be rational if investors produce for themselves different amounts of information, accumulate it progressively rather than all at once, and even, in the case where information production is very costly for them, decide to imitate other investors’ behavior, because they think that those other investors are better informed (the firm’s insiders for instance).

What efficient market theory concludes is that there is no “easy money”, i.e. it is difficult for an investor to beat consistently the market, and there is no obvious way to forecast market prices. Interviewed by the journalist, that’s exactly what Burton Malkiel said: “The problem with bubbles is that you cannot recognize them in advance.” In other words an efficient market is a market the evolution of which is very difficult, or impossible, to predict.

But the popular demand addressed to economists is precisely the impossible: tell me what will happen tomorrow, without any “if” or “conditional upon”. What serious economist can do, on the other hand, is precisely to predict what will happen if ….

As Burton Malkiel said in the same article:

“If you are leveraged 33-1 and you’re holding long term securities and using short-term indebtedness, and then there’s a run on the bank …. How can you blame that on efficient market theory?”

A last word regarding the Chicago School. First it is an exaggeration to pretend that all economists there consider that all markets are perfect. Imperfect markets are currently analyzed at Chicago as they are elsewhere, and it is widely recognized that the study of imperfect information is necessary. Second, one could both recognize that imperfect markets are a reality, depending on the part of the economy under review, and still conclude that an imperfect market is preferable to no market at all or even to a market that is imperfect but heavily administered by a very imperfect political process.

To conclude, it is obvious that information is imperfect, especially that of journalists and other economic commentators.

No comments: