Overshoot self-correction to collapse in theS&P 500Mar-Aug 07

Posted by Marcus G. Daniels on
URL: http://friam.383.s1.nabble.com/Overshoot-self-correction-to-collapse-in-the-S-P-500-Mar-Aug-07-tp524586p524608.html

sy at synapse9.com wrote:
> just because independent behaviors of emergent natural systems are not susceptible to deterministic analysis of the usual sort doesn't mean they're not observable, dangerous and generally predictable by other more general means
In a high-dimensional dynamical system characterized by a single summary
statistic, like an index price of widely held stocks, the standard for
`that's really weird but I know why' gets more strict, not less.  
Occam's razor would point toward those few factors shared by all of the
components of the system that enter into this summary statistic.  One
class of factors are the intrinsic properties of the system (having
many, potentially reinforcing, instances).  Another class of factors are
common perturbations to the whole system.    Unless the time series
phenomena occurs all of time in many sorts of contexts, it seems to me
that external perturbations are the sort of explanation to be seeking.  
Otherwise, convincing evidence of an intrinsic explanation would be a
price prediction algorithm based only on price with odds of success
sufficient to make money!