ChaoticMarket 47The speed of intrinsic time may differ from that of chronological
time. Price changes would then move ahead of information changes.
Investors and other market participants conform perfectly neither to
the linear assumption of homogeneous expectations nor to the ubiq-
uitous irrationality of noise theory.
Instead, investors have heterogeneous expectations that may or
may not be rational and that may be defined according to a number
of variables. Chief among them are investor time horizons that range
from the very short term (for day traders and minute traders and
market makers, say) to the very long term (for central banks, say).
The range of different time dimensions contributes to the Joseph
and Noah effects, persistence, discontinuity, and premature and de-
layed adjustments to information. Short-term traders react more
quickly to new information; long-term investors react more slowly.
Therefore, information changes will not produce proportionate price
changes. Indeed, volatility will increase when there are greater num-
bers of short-term traders (day traders) than long-term traders.
Changes that are produced constitute new information, produc-
ing another round of price changes again defined according to a
range of discrete time dimensions. Adding further complexity to this
mix of investor heterogeneity and time dimensions is the increasingly
global nature of financial markets: News itself is dynamic, traveling
around the world, usually in 24-hour cycles, and impacting Tokyo,
then London/Frankfurt, then New York, and around again.
In this reality, it seems implausible to claim instantaneous, un-
biased market adjustment to new information and it is not necessary
to attribute all market preadjustment or readjustment to irrational
noise trading. Incremental information changes in a perfect market
would be expected to produce proportionate price changes.
But informational changes produce disproportionate changes. In
terms of chaotic dynamics, these disproportionate changes may be
seen as a result of initial measurement error that (as in the hockey
puc kexample and the butterfly effect generally) leads to exponen-
tially greater price changes over time.
BEHAVIORAL FINANCE
Not only do these systemic complexities and the stickiness of prices
show new reasons to be skeptical of EMT, they also suggest partial
explanations for the observed nonlinear dependence of stoc kprices
and the possible presence of chaotic phenomena. They certainly