I don't disagree with most of what you say here, but the drop in money
supply is something which can be accounted for "conventionally". I had
a couple comments, which I'll give out of order.
Horace Heffner wrote:
> This theory seems to be working out to some degree. It appears M has
> been driven strongly toward zero.
This may have nothing whatsoever to do with program trading.
Very roughly speaking, the money supply is determined by the amount of
"high powered money" in existence (direct loans and gifts from the Fed)
times a multiplier. The multiplier is determined by the reserve
requirement in effect on banks.
If the reserve requirement is 5%, then of each dollar placed in the
banking system, 95 cents is lent out. The borrowers may stuff a little
of the money they borrow into a mattress (and hence take it out of
circulation), but *nearly all* of it is going to be either spent or
invested. If it's spent, it goes to a company, and if it's invested, it
also goes to a company, and ultimately it works its way back to the
banking system.
Once back in the banking system, it's subject to the same reserve
requirement, and once again 95% of it is lent out.... and around and
around and around we go.
The result is that the money supply can be viewed as, *very* approximately,
high powered money * sum_{n=0,inf} (1 - a)^n
where "a" is the reserve requirement. The sum on the right can be
evaluated by subtracting from it the same sum taken from 1 to infinity
and dividing through by (1 - (1 - a)) to obtain
sum_{n=0,inf} (1-a)^n = (1-a)/a = 1/a - 1
With our "strawman" reserve requirement of 5%, that evaluates to a
factor of 19. Thus, with a 5% reserve requirement, the money supply
might be expected to be (*very* approximately!) 19 times the amount of
money deposited in the system by the Fed.
Now, suppose there's panic or breakdown in the financial system. If the
bank lending rate drops because of fear among the bankers, *or* if
people panic and start removing their money from the banks (a "run"),
that factor of 19 can drop very quickly. And when the "lending
multiplier" drops the money supply drops just as fast.
Something like this happened at the beginning of the 30's depression,
and about 2/3 of the money supply simply vanished, virtually overnight.
(Or maybe it wasn't 2/3 -- I've forgotten the actual fraction, as
calculated by Friedman in some essay or other.)
>> The problem with modern portfolio theory is its fundamental
>> assumption, that the market activity is actually based on stochastic
>> processes. It is assumed that all fundamentals are known by all the
>> participants and very quickly "priced into the market". All that is
>> left is due to random fluctuations.
This is indeed the common assumption, and it is patently false.
Stock prices continue to be set largely by emotional factors, and stocks
which are "in style" are consistently overvalued while those which are
"out of style" are consistently undervalued. By "overvalued" I mean
over a long term (say, a year) the value of "stylish" stocks tends to
drop relative to the market, and by "undervalued" I mean that over a
long term (say, a year) the value of "out of style" stocks tends to rise
relative to the market.
This has been true for as long as anyone's been studying the market and
it continues to the true today, as far as I can tell.
P/E ratios (and, more recently, PEG ratios) tend to tell the "style"
story pretty clearly. As one trivial example, back when Sun
Microsystems' PE hit ~ 90 it was clear that it was a "stylish" stock and
its price was being supported not by perfect information, but by the
usual herd mentality among brokers. That was a few years back but I
don't see any evidence that much has changed since then, save for the
massive increase in volatility which results from program trading and
from the loss of some securities regulations which "damped" things (and
which were removed under pressure from the program traders).
As far as I know there is no model which correctly predicts company
growth. Unlike option prices, stock prices have no firm theoretical
foundation.
