|
|
|
|
|
|
|
Jens 'n' Frens
Idle thoughts of a relatively libertarian Republican in Cambridge, MA, and whomever he invites. Mostly political.
"A strong conviction that something must be done is the parent of many bad measures."
-- Daniel Webster
|
|
|
|
|
|
Monday, March 10, 2003 :::
I hate this post, Steve, because it takes all the good things I wanted to say. "Oh," I think, "let me say something about hyper-rationality", and there it is at the very top of the piece. So to some extent, I'm left restating what he says, but in my own way. (I think there's something novel here, but probably not much. Search his article for "exhausted gold mine" to find the passage most relevant to what I've written; I jumped in, as is my habit, with the talking before finishing the listening, and most of what I wrote was written by him around there.) There is evidence that more energy is spent doing research and trading than is actually efficient; this assertion doesn't even make since if you don't accept that requiring markets to be rational doesn't require that market prices be perfect, merely that they be close enough that to profit off of them requires more work than the abnormal return from such activities can justify. (Even if this is true, I think it's kind of impressive how much of our economy is rationally allocated to price discovery; in fact, I think one of the attractions of Marxism is the belief that this can't possibly be right. Nobody, in a Marxist society, is "wasting" their time collecting tolls — a straight transfer at that point — or changing price labels on goods; these activities are only required to determine what the more directly productive part of the economy should produce. If I didn't believe in the value of such activity, I might become a bit despondent; I write software that communicates orders on securities that derive their value from enterprises that may or may not produce actual consumer goods, which means I'm something like five levels removed, depending on how you count.)
He notes that markets can be rational even if not all investors are; if 2% of investors are perfectly rational, and the other 98% makes sufficiently random and uncorrelated mistakes, the entire 2% weighing in on the same side can often be enough to correct the mistakes. I find this often to be believable, though I'm pretty true some mistakes were being made in a correlated manner 3 years ago today. (Happy anniversary, NASDAQ investors.)
I've been reading some Karl Popper — he appears to be the source of the notion that for a theory to be scientific requires that it be in principle falsifiable, which struck me the way that the student was struck who complained that Shakespeare used too many cliches; it's a familiar enough idea to me that, when I was confronted by its provenance, I was probably less impressed than I should have been. Some of the discussion of behavioral psychology made in this article evokes Popper commenting on theories he classes as pseudo-scientific, such as Freudian psychology, into whose rubric just about any conceivable observation can be crammed and thereby explained, just as easily as its opposite could. It also reminds me of a popular quip in the world of theoretical physics that a good theorist can explain any data, whether correct or not. (Some of modern theoretical physics was as much as called pseudo-science by Nobel Laureate Sheldon Glashow in the mid eighties.) This is meant (I hope) as a bit of self-critique, partially intended to warn against a tempting mistake, much as is the story repeated by Rubenstein of the economist who stops his friend from picking up a $100 bill lying in the street; "If there were really a $100 bill there, someone would have picked it up."
My take on market efficiency in general is similar to my take on evolution (where I may have a less sanguine view of evolution than does Rubenstein): an investigation driven by asking of each feature of an animal "What is the evolutionary advantage of this?" will find some features for which there is none, but it will have a nonobvious answer more often than it has no answer, and to simply assume that anything that isn't obvious to you must be wrong is both incorrect and denies you a lot of interesting investigation. (Indeed, the most elegant results are almost always nonobvious.) I think laymen assume erroneously that they see true market inefficiencies far more often than economists make the opposite mistake; Rubenstein points out that one doesn't see a lot of $100 bills lying around, for exactly the reason given by the skeptical economist.
::: posted by dWj at 9:48 AM
|
|
|
|
