Experimenting with a market simulation.

Michael de la Maza and Deniz Yuret (1994) ( PS )
Experimenting with a market simulation. The Magazine of Artificial Intelligence in Finance, 1(3).

Abstract:

At the turn of the century, Texas wildcatters would decide were to dig for oil by taking a pinch of sand from the ground and tasting it. Their methods were uninformed by any understanding of the causal processes that led to the formation of oil deposits. Everett Lee De Golyer, known as the Father of American Geophysics, was the first to introduce scientific methods into the oil discovery field on a wide scale. As a result, he become a multimillionaire in an age in which a million dollars was still a lot of money.

The money management industry today is similar to the oil industry at the turn of the century. Virtually all of the trillions of dollars that are actively managed are managed by seat of the pants methods that are similar to digging for oil by tasting sand. However, in recent years a growing number of individuals and companies have dedicated themselves to applying the methods of science and engineering to the stock market. Those who have been successful in doing so have made fortunes and there are many more fortunes to be made.

In this article, we take another step in this direction by applying our understanding of genetic algorithms to the market. This article describes a simulation of a market in which individuals evolve trading strategies and compete against each other to maximize profits.