>Predictive markets are
>based on the Efficient Market Hypothesis, which suggests the collective
>wisdom of a group is greater than the individual wisdom of a single
>person.
>
>To illustrate, the process of valuing a company is a fairly arduous
>task for a single person. That person might speak with company
>executives, customers, and competitors in order to make an educated
>guess about the company's future cash flows. Alternatively, the public
>stock markets do a good job of valuing companies by aggregating
>information that a lot of different people have. The market quickly
>"prices in" new information by driving a stock price up or down.

 From a 1994 Doug Henwood article:

The U.S. economy also served as a laboratory for another economic
experiment, efficient market theory, which, its elaborate
mathematical structure aside, boils down to the touching faith that
markets know all, and the stock market in particular knows most of
all. Stock markets, since they instantly reflect the collective
wisdom of "sophisticated" investors, are marvelously prescient
instruments that guide capital towards its most productive uses. And
if the real world were reformed to resemble the stock market, if
impediments like regulations and unions and national borders that
block the immediate adjustment of prices to available information,
then efficiency would be served.

That doctrine too, which Michael Jensen once called the
best-established fact in all the social sciences, is now on the
ash-heap. A series of studies have demonstrated convincingly that
market volatility far exceeds the volatility of the underlying
fundamentals, that financial market prices swing to great extremes
only to revert back to the mean, that anomalies in pricing (like
better-than-average performance shown by stocks that are cheap
relative to underlying corporate profits) persist despite the fact
that theory claims they shouldn't, and that not all market
participants are equally well-informed and can never be. Much of this
is common sense, again, to anyone who watches Wall Street from
anywhere but a university campus. But the fact is that efficient
market theory has been heavily discredited in the very environment
where it was once revered. And if stock markets are not the rational,
efficient places that they were thought to be 10 or 15 years ago,
what of the project of making the real world more like the stock
market?

Curiously, the broader inference -- that if the stock market is a
madhouse, those who want to make the real world more like it are
engaging in a pretty strange enterprise -- has hardly been drawn. Far
from it. Robert Shiller, the Yale economist who has done about as
much as anyone to demonstrate that stock prices are far more volatile
than underlying corporate profits, has just published a book in which
he argues that we should establish futures markets for nearly every
important economic number -- GDP, employment, housing prices, you
name it. In other words, make the real world more like the stock
market.

Full: http://www.leftbusinessobserver.com/Anti-market-forces.html