From: John Conover <john@email.johncon.com>
Subject: forwarded message from John Conover
Date: Mon, 28 Oct 1996 12:55:29 -0800
Predictions like the attached are interesting. If you look at the
statistics of such prognistications, 52% of the time they will be
correct, 48% incorrect. (Likewise, saying that the market value will
increase in November would carry about the same statistics-you would
run about the same chance of being correct, also.)
When such things were pointed out by Brian Arthur in the early 80's,
there was a lot of argument in economics. Want to prove it correct?
Here is how:
Take a piece of paper that is divided into two equal parts,
labeled "go down," and "go up." Every day, (week, month, year,
whatever you want your time period to be,) and every time period,
throw a dart over your shoulder at the paper. Your prediction will
be in which area the dart lands. Do it enough times, like over 16
years, and you will find that your forecasting mechanism is as
good as the experts. (If a CEO's decisions determine a company's
earnings, then a stock's price will be determined by the earnings
and the prime rate-if you believe in the Rational Market
Hypothesis. So, we would expect that a CEO's decisions would be
about the same as the up movements in the stock's price-which is
52%, which is measured over a long enough time.) So much for MBO.
One pundit economist quipped, "I have predicted 9 out of the last 5
recessions," which is a humorous statement of the same thing. (In 10
predictions, we would expect about 56% to be correct, or so, using the
above statistics.)
For more details, see the current Harvard Business Review, an article,
"Increasing Returns and the Two Worlds of Business," Brian Arthur.
As it turns out, the RMH is correct-sometimes. And some times it is
not. (Arthur argues, based on an enormous empirical study-backed by
complexity-theory, that there are two kinds of business
environments-and the rules are different for both. High tech company's
have to by managed by margins on the fly, low tech company's actually
do fit well into the RMH model.)
John
BTW, in case you are interested in the historical perspective, by
1939, what is now called Keynesian economics, had been largely
discounted. This period is called the pre-Keynesian and Keynesian
period of economics. It was replaced by what is called neo-classical
economics, which was based on Von Neumann's
game-theory. Unfortunately, to be useful, it required what is called
utility-theory, and no one was ever able to concisely define the term,
much less quantify it. Most MBO, Operations Research, etc., that we
use today comes from this theory. Unfortunately, neo-classical
economics was incapable of proving that markets will expand, which,
empirically, almost all do. In the early 80's, there was a revolt,
(commonly called the "revolt of the Turks at Stanford,) where
complexity-theory was used for market and GDP analysis. Economics is
a very conservative field, and complexity-theoretic solutions were not
well received-except by a few folks that though programmed equity
trading was possible. (Most of the prestigious journals would not
print Arthur-for close to a decade.) neo-classical economics predicts
that diminishing returns is the macroscopic economic agenda for any
economic system. Complexity-theoretic solutions would seem to indicate
that increasing returns are the economic agenda, at least in
contemporary industrial countries-thus the name of the contemporary
prevailing concept in economics-increasing-returns. Most work is being
done with this theoretic concept-to the tune of several hundred papers
being published a month on various aspects of the subject. (A decade
ago, HBR would not print Arthur-now it is an invited paper.)
------- start of forwarded message (RFC 934 encapsulation) -------
Message-ID: <"M3Zx82.0._M6.buGTo"@netcom3>
From: John Conover <conover@netcom.netcom.com>
To: John Conover <john@email.johncon.com>
Subject: Business News Headlines [Oct 28 3 am PST]
Date: Mon, 28 Oct 1996 5:50:58 PST
Reuters Business News Highlights
FRANKFURT - Market guru Elaine Garzarelli, best known for
predicting the October 1987 market crash, said in a magazine
interview that she still believes Wall Street could drop about
15 to 25 percent by the end of the year.
Garzarelli, in an interview published in Germany's Der
Spiegel magazine, said in particular that she expected the
market drop to occur after the November 5 presidential election.
------- end -------
--
John Conover, john@email.johncon.com, http://www.johncon.com/