Re: Sweden -- A few reactions on the N.Y. Times article,lbo-talk@panix.com

["Rosser Jr, John Barkley" <rosserjb@xxxxxxx>]

     Well, Doug, you're sounding like Paul Davidson here.
All we need to know is that the economy is nonergodic and
maybe some simple linear algebra, and all the rest is just
a bunch of flimflam hooey, or whatever.  Right?
     Well, I have already argued that a chaos/complexity
view of financial markets implies that they are inherently
and endogenously unstable.  Now, it may be that "we" who
have read Marx, Keynes, Galbraith, Minksy, Kindleberger,
Shiller, and other assorted types (and even Henwood's _Wall
Street_), may "know" this.  But there are a lot of
economists and financial types who don't, who claim that
the market is efficient, that it is endogenously stable,
and that the only reason we see volatility is due to
exogenous "noise" (Brad De Long and others have shown that
rational market actors must chase the noise traders).  I
would note that even on a seminar on _Wall Street_, a lot
of people seemed to not know very much... .  .   .    .
     Now, since you already "know" about the inherent
instability of the market, you may not be impressed that
some mathematical theory is out there that backs up what
you already know.  Who needs it (and maybe nobody does)?
But it may well be useful in dealing with those who are
more math oriented and who think otherwise in general (and
who view Marx, Keynes, Minsky, et al as hopelessly outdated
and nonrigorous, if not completely wacko).
     Rather than the chaos theory stuff, let me cite a
paper from _The Economy as an Evolving Complex System II_,
ed. by Arthur, Durlauf, and Lane, 1997, SFI,
Addison-Wesley. It is a paper with a bunch of authors of
whom the lead is Brian Arthur and it is a genetic algorithm
model of a bunch of agents in the stock market.  It allows
these agents to evolve strategies over time.  In some ways
the underpinnings of the model are very neoclassical in
that there is a fundamental equilibrium value for stocks.
But the model shows the system evolving over time with
bubbles and crashes breaking out from time to time
separated by periods of quiescence and stable behavior.
Now, again, of course "we" may all "know" this and this
particular computer simulation version (as opposed to some
analytical chaos model) may not "prove" anything and may
not convince anybody who is not already convinced.  But, it
is a pretty impressive piece of work, IMHO.  It looks a lot
closer to "reality" than the analytical models.
     But, again, if one already knows it, one may not care.
But, Doug, not everybody "knows" what you do, not even
everybody who subscribes to LBO or is on lbo-talk or has
even read _Wall Street_, :-).
     BTW, I did not claim that you thought the Tobin tax is
a panacea, only that you made "favorable noises" about it.
I quote from p. 319 of WS: "Few things, aside from the
threat of direct appropriation of their property, make Wall
Streeters scream more loudly than the assertion that their
pursuits are pointless or malignant, and that their
activities should be taxed like a noxious effluent.
Listening to those screams would be another positive
benefit of a transactions tax." [several others having been
listed earlier]
Barkley Rosser
On Tue, 11 Aug 1998 14:16:47 -0400 Doug Henwood
<dhenwood@xxxxxxxxx> wrote:

> Rosser Jr, John Barkley wrote:
>
> >(a
> >source of much anger to many mathematicians to this day,
> >although Doug thinks that mathematics is useless for
> >economics
>
> Now, Barkley, I don't think that at all. Math is essential to studying,
> say, labor markets and income distribution, and I've read a fair amount of
> that literature. But it's pretty simple math. What I'm skeptical about is
> all the whiz-bang machinery econometricians use, and the latest fads for
> complexity/chaos theory. I'm at a loss to understand just what they
> contribute, and your defense of applying those fads to financial markets
> was hardly convincing. Yeah, the stock market is nonlinear, even allowing
> for noise. What, really, does that tell us?
>
> Doug
>
>
>

--
Rosser Jr, John Barkley
rosserjb@xxxxxxx