Investor Simulation Research Based On Lmproved Artificial Stock Market Model

The modem finance system is a complex system with the property of exoteric, adaptive and non-linear. Being faced such a huge and complex macroscopical finance system, we have cognizance that the traditional analysis of the financial system can no longer meet the research needs. There apper many new theories among which the Multi-agent simulation has been widely used in the social and economic research. MA, characterized by its bottom-up nature, picture the economy as a complex system composed of lots of interacting intelligent agents and presents economic agents at the microscopic level to achieve the macroscopic effects.Based on Swarm platform, the paper models and implements an order-driven stock market using multi-agent simulation method. This artificial stock can be securitized from the complex system perspective and examines how the interaction between traders the microscopic level can result in macroscopic behaviors. Evolutionary Game Theory describes a continuous game of agents,it masters the key feature of agents with boundary rationality contesting for limit resources in economic society. The data produced by simulation is highly conformed to the empirically observed data from the real market. The innovation lies in the combination of the newly-borne multi-agent simulation methodology with computer science. Moreover, in spite of the basis of the mainstream microstructure theories on which the model is built, this artificial stock market with relatively sophiscated trading mechanism is comprised of the two types of traders:fundamentalists and technicalist interacting with each other, differing from its predecessors that consist of the only one type of heterogeneous agents interacting in a simple trading mechanism.