From: John Conover <john@email.johncon.com>
Subject: News: Futurist: markets, attacks, point to possible depression
Date: Mon, 05 Nov 2001 06:34:36 -0800
Always, in times of tribulation, doomsday prognostications capture
the media's attention.
We might be headed into a depression, and we might not. Whether we
are, or not, is not knowable.
However, the odds of a depression being imminent are not very
significant.
The standard deviation of the increments in the US GDP, over the
20'th century, was about 10% per year, (0.099733 from 1930-1995.)
During the US Great Depression of 1930-1933, the GDP dropped about
50%, (in 1930, it was $96.8 billion, in 1933, it bottomed at $56.8
billion-the GDPs in the Asian Contagion in the late 1990's did about
the same.)
The standard deviation of the decrease in the US GDP during a
depression that lasts at least three years is about sqrt (3) * 0.1 =
0.17.
Or, a 50% decrease would be about a 0.5 / 0.17 = 2.887 standard
deviation incident, which has a chance, on average, of once in 526
years, or about twice a millennia-quite a rare event, indeed.
John
BTW, assumptions: Black-Scholes paradigm, erf (3) = 1, statistical
independence in the marginal increments of the US GDP-all reasonable
first order approximations, (depending on who is telling the story,
of course.)
Statistical estimate on the standard deviation of marginal increments
of the US GDP being 10% is about 1 / sqrt (95 - 35), or about 12%, or
a confidence level of about 88%. Call it a 90% confidence level.
Data references: 1996 US Federal budget,
gopher://sunny.stat-usa.gov:70/11/BudgetFY96, and the 1997 US Federal
budget, http://www.doc.gov/BudgetFY97/index.html. Data from Table 1.2,
"SUMMARY OF RECEIPTS, OUTLAYS, AND SURPLUSES OR DEFICITS(-) AS
PERCENTAGES OF GDP". Data not adjusted as constant 1987 dollars, nor
inflation.
So there.
http://www.computeruser.com/news/01/11/05/news15.html
--
John Conover, john@email.johncon.com, http://www.johncon.com/