Tuesday, December 8, 2009

Randomness and Herds


Economists used to believe that market prices were determined by all available information, and this was the most efficient way to do so. This is called the rational market hypothesis. I used to believe this, and it is nice because it gives us a kind of utopian world where zillions of "rational actors" guide the economy in the most efficient way via a  "magic hand." Thus, the markets may bob and weave, but they are not actually random -- actually the opposite, completely determined by billions of bits of information.

I don't think the old-time economists really believed this, but it formed a theoretical basis for economics that lasted for many years. I liked it because it was tidy.

On the opposite side are the technical analysts, who are like numerologists, studying the trends of the prices in an attempt to make money. These guys don't think about basic value, just analyze the market movements. It is the opposite of rational markets.

Other people argue that the stock market makes a random walk between more giant market crashes or big run-ups. This math is  called a Levy Flight. See diagram at right.  Empirical evidence shows the Levy flight model is a little more extreme than reality, but it fits better than the random walk does.

I like this because it has a lot of randomness followed by occasional bouts of rationality.  (Of course, it could be randomness followed by irrational panic too.)

The modern treatment is that the markets depart from rationality in a few known ways. One of these is herd behavior, which is the notion that buyers and sellers all think the same way. That is the club of buyers and seller is too homogenous. It seems US stock traders are more homogenous than European ones.

Anecdotes are easy to find. Recently everyone believed that real estate could go up in value forever. In the 1990's the dot.com boom & bust was also group-think gone mad.