As a complex giant system, financial market has complexity characteristics with distinctive forms, which shakes theoretical basis of effective market hypothesis to challenge traditional paradigm for financial research. Based on agent-based computational economics, the work established artificial stock market by computer’s high-speed information processing ability and bottom-up modeling method. The real financial markets were highly simulated to guide financial market regulation and policy development. The artificial stock market model in the work realized the simulation of real stock market to a certain extent. This showed that it was effective to improve artificial stock market modeling method based on agent and scale-free network. Therefore, the application value analysis of model has important practical significance.