Co evolution of Genetic Programming Based Agents in an Artificial Stock Market

The complexity of the financial markets, represents a big challenge to the specialist in the area. The traditional way of coping with the analysis of such systems is the use of analytical models. However, the analytical models present some difficulties and this has leaded to the development of alternative methods for the analysis of such markets. In this paper we analyze the different conditions under which the statistical properties of an artificial stock market resembles those of the real financial markets. The different types of agents that we use in the simulations are technical, fundamental and noisy. Changes in some parameters and agents’ behavior produce different properties of the stock price series. We analyze the wealth distribution of the agents after several periods of trading in the different simulation cases.