In their book on game theory John von Neumann
and Oskar Morgenstern observe that it is the
study of "free fall which brought forth mechanics".
But, contrary to conventional wisdom, the study of
free fall did not rest solely on observing the
fall of an apple.
The author aptly notes that it required the observation
of a whole set of falling bodies in order to
get rid of a bunch of unessential factors and to
single out the influence of gravity. And of course
it required Newton's genius to include into this
set of falling bodies the "fall" of the moon.
In short, by describing in detail a single
stockmarket crash one will fail to understand what is
common to ALL stockmarket crashes. The author extends
the argument even further: by focusing solely on stock
markets one will fail to recognize what is common
to a broad variety of speculative episodes.
One might think that there
is a big difference between a price peak for postage
stamps and one for stock prices but this difference may
be more appearant than real as indeed
is the difference between the fall of an apple and that
of the moon.
What kind of results are brough about by this
At the present time (October 2002) we are particularly
interested in real estate prices. Will they continue to
rise, will they stabilize or are they going to fall in the
The sections on real estate price peaks provide several
1) It is in the most expensive markets
(e.g. San Francisco as opposed to say Pittsburgh, PA) that
prices begin to rise first at the start of the
peak; it is in these markets that prices peak first
which gives a hint about when the downturn will occur in
other markets; and it is in expensive
places that the price increases are the strongest.
2) Price peaks for real estate resemble those for
stocks in that they are almost symmetrical with respect
to the maximum; which means that where prices experienced
strong increases, they also experience
The last part of the book provides a theoretical framework.
It begins with an insightful chapter that points out that
there are in fact two classes of speculative peaks: the
U- and the S-class. The first includes all bulky commodities,
the second comprises real property, collector books,
diamonds, bonds and in a general way
all items for which transportation cost plays no
role and for which investors can (at negligible cost)
select the item that they want to buy.
The last chapters show how these effects can be accounted
for provided one uses a model which includes a memory effect.
This means that one should use a dynamical equation of order
at least equal to two. In addition,
a basic parameter
is the proportion of investors,i.e. people who buy in order
to sell with a profit within a few years, with respect
to the number of "users"
who buy a house in order to live in it for say
A superb book. On almost every page, good ideas
come swirling and densily packed.