chaoplexity and institutions

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

     This is being sent to three lists, although it is an
immediate followup to a thread on pkt (longish).
     Dave Colander asked if institutions can limit the
instabilities associated with chaos and complexity
(chaoplexity), a la the famous "corridor of stability" idea
of Axel Leijonhufvud.  I said maybe, but then said that
they may also lead to greater instability.  I wish to
follow up:
     1)  The corridor of stability has inherent in it the
possibility that by widening the corridor a bit, thus
allowing more local volatility, one may gain more
boundedness or global stability of the system.  This has a
counterpart in ecological theory in the alleged tradeoff
between "stability" and "resilience" enunciated by C.S.
Holling.  A simple example is the oak tree (stable but not
resilient in a hurricane) versus the palm tree (unstable
but resilient in a hurricane) or elephant populations
(stable but unresilient) versus sheep blowfly populations
(potentially chaotic but resilient). There are many chaotic
systems that are actually very resilient, if locally
unstable, including our brains (when we aren't bonkers).
     2)  An example of an institution trying to increase
stability but reducing resilience might be the the
government bank insurers, or in Japan and "non-opaque"
banking systems the central authorities more generally.  In
the US this was posed as the "moral hazard" problem of
FSLIC in that when banks or lenders or borrowers or all of
them felt safe because of the government agency covering
for them they engaged in reckless behavior that eventually
brought the system to a much greater degree of crisis,
especially as we are now seeing in East Asia where lots of
lenders and borrowers counted on their governments to
maintain pegs to the US dollar.  Similar such complacency
is what we see at the late stages of speculative asset
bubbles a la the Ponzi analysis of Minsky.
     2)  Nevertheless the fear of actual financial market
chaos is currently playing a role in some major institution
changing going on right now, in particular European
monetary unification.  One of the most prominent advocates
for the EMU all along has been Paul de Grauwe of the
University of Leuven in Belgium.  I gave a talk there in
1990 on chaos theory right after he and Kris Vansanten at
CEPR in London had published a paper showing how chaotic
dynamics could easily arise in forex markets.  This was
followed up by a book by de Grauwe, Embrechts, and Wachter
in 1993 on Chaotic Dynamics in Foreign Exchange Markets
that developed the argument much further and has been much
cited.  I personally am convinced that de Grauwe sees the
EMU as the way to KILL THE CHAOS in the European forex
markets.
     4)  Finally I note that in the kinds of models of
financial markets where there is a struggle between
"fundamentalists" and "chartists" or something like that
(which is what is going on the de Grauwe et al models) much
more complex dynamics are possible than mere chaos, e.g.
fractal basin boundaries (eeeeek!) and some other pretty
hairy stuff. A paper giving the full array of this stuff is
William A. Brock and Cars H. Hommes, "A Rational Route to
Randomness," _Econometrica_, 1997, vol. 65, pp. 1059-1095.
Barkley Rosser
Professor of Economics
James Madison University
Harrisonburg, VA 22807 USA

--
Rosser Jr, John Barkley
rosserjb@xxxxxxx