Re: Poll on market efficiency

From: John Conover <>
Subject: Re: Poll on market efficiency
Date: 28 Jul 2000 12:35:57 GMT

FWIW, efficiency in the sense of the EMH, (depending on who is telling
the story, of course,) gets larger as the system's entropy gets
larger, (the EMH assumes infinite entropy, i.e., 100% efficiency, as a
first order approximation; probably in the name of mathematical
expediency, so the variances exist, which are assumed to be linearly
proportional to risk-which is a good idea since it saves us from
having to program all that median and interquartile stuff.)

The persistence, (i.e., the likelihood of what happened today,
happening tomorrow,) is a 50/50 crap shoot if the entropy is infinite,
and, by definition, the system would be efficient. The system would
also be fair in the game-theoretic sense; it would be impossible for
anyone to have an advantage in making money over anyone else,
(excepting those that cheat, of course.)

I keep a running tab on the persistence for the stocks in the US
exchanges, (daily closes-whatever daily and closes means these days,)
and the persistence has been running about 0.52, averaged across the
stocks in the exchanges, since about 1983, which is not efficient,
(its close, but not close enough not to be exploited by the PTs.) For
the historical databases of the 20'th century, it is only slightly
lower, but still not 100% efficient.

So there.

        John writes:
> jim blair <> wrote
> > Exactly correct.  The (probably unresolvable) question then is: when the
> > price of the stock changes (up or down) is that because the "estimate" of
> > its true value changed?  Or because changes in the economy/world changed the
> > true value?  Changes in taxes or laws or technology or whatever.
> >
> > I will just cite changes in the laws that were passed or took effect (or
> > were seen as likely) just before the US stock market crash of 1929 and the
> > Japanese decline that started in 1990.
> >
> > And I comment that the original post assumed that "efficiency" ranges from 0
> > to 100%, but did not define the term.  If he means "accuracy", then an
> > "error" can be in either direction (too low or to high).  So why is
> > "efficiency" limited to 100%??
> Well, any discipline is allowed some leeway with its terms-of-art.  The
> physicist's definition of "efficiency" is not the same as the layman's,
> and neither does the economist's have to be.  I do grant you that fuzzy
> definitions make for fuzzy thinking, so it would be reassuring to hear a
> real economist define "market efficiency" in a non-fuzzy way.
> The only definition I remember from Econ 101 (actually, it was 14.001 in
> MIT-speak, and the lecturer was Robert Solow) had to do with the notion
> that all sellers are asking the same price, and all buyers are bidding the
> same price, in an "efficient" market.  I don't know how that squares with
> the pricing of airline tickets nowadays.
> To answer your substantive question, though:  it's obviously the
> "estimate" of discounted future earnings that fluctuates 10% a day (or
> more, for some stocks), not the "intrinsic value" based on laws,
> technology, and so forth.  By "obviously" I mean "that's what I think,
> anyway".
> -- Tony Prentakis


John Conover,,

Copyright © 2000 John Conover, All Rights Reserved.
Last modified: Fri Aug 4 21:41:52 PDT 2000 $Id: 000804213845.4670.html,v 1.0 2001/11/17 23:05:50 conover Exp $
Valid HTML 4.0!