Modeling And Analyzing Of Stock Market Based On Multi-agent

With the continuous development of economic, the stock market has become the most active and fastest growing one of capital markets in China, so the study of stock market becomes a focus by many experts and scholars. There are a lot of widespread existing stylized facts in the stock market, such as fat tail(leptokurtosis), volatility clustering and so on, which are hard to be explained by traditional financial theories. As a developing multi-agent based complex adaptive system, it provides a new angle to study the economic system that is considered as a complex system which is composed of numerous economic agents which are interacted, autonomous and having learning ability. The evolution of the economic system is a result of which agents adapt the surrounding environment, and the emergent behaviors of interaction between agents at the macro level. This bottom-up modeling idea is based on multi-agent modeling method.The stock market is a typical complex adaptive system. In this article, we emphasize on the study of the influence factors to the volatility clustering of the multi-agent model based on the information communication. The results obtained here indicate that the change of topology of the market has a further effect on the behaviors of price, and provide a good reference to study influence factors. In order to explain the "stylized facts" of the stock market effectively, we establish a new model by introducing the spectator into Lux model, and use Swarm software platform to simulate, then analysis the data from the model. After that, we do research on the GARCH effect of the returns produced by the model, find that the return GARCH effect is obviously when the market is in the periodic state.on the contrary, when the market is in the bubble equilibrium state, the return GARCH effect is not obviously. These results provide a certain theoretical basis to help stock investors and market traders make decisions, enrich the existing financial theory.