Abstract :
The price fluctuations of the stocks in the financial markets are the result of the individual operations by many individual investors. However for many decades the financial theory did not use directly this “microscopic representation” of the markets. The main difficulties preventing this approach were solved recently with the advent of modern computer technology: – massive detailed data on the individual market operations became available; – “microscopic simulations” of the stock markets in terms of their individual participating agents allow very realistic treatment of the problem. By taking advantage of the modern computer processing and simulation techniques, we are now able to confront real market data with the results of simulating “microscopic” realistic models of the markets. These models have the potential to include and study the effects on the market of any desired feature in the investors behavior: departures from rationality, herding effects, heterogeneous investor-specific trading strategies. We propose to use the comparison of computer simulations of microscopic models with the actual market data in order to validate and enhance the knowledge on the financial behavior of individuals. Moreover we hope to explain, understand (and may be predict and control) macroscopic market dynamical features (e.g., cycles of booms and crashes, investors wealth distribution, market returns probability distribution etc.) based on realistic models using this knowledge.
