Most studies on financial market using the method called’Agent-based Computational Finance’are based on the assumptions that the market has only one risk asset and there is no information interaction among investors. However, with the development of information technology, information carrier is constantly increasing, information exchange between people is increasingly close and the possibility of linkage between different assets is growing. Therefore, the study on the wealth dynamics of heterogeneous investors who exchange information each other in a multi-assets market provides not only strategic guidance to investors, but also a reference for supervision and regulation department to establish the system and policy.This paper applies the method of agent-based computational finance, modeling an artificial multi-assets stock market where investors exchange information with each other. Some of the investors place much emphasis on fundamentals of assets, while the others pay close attention to historical prices; some of them are mark-to-market, however, the others pay limited attention to the market. Therefore, investors are divided into four groups:fundamental-mark-to-market, fundamental-limited attention, technology-mark-to-market and technology-limited attention. The simulation result concludes that fundamental-investors have more effects on stability of the market than technology-investors have as the number of’fundamental’ investors is approximate to the number of’technology" investors. In addition, regardless of whether the market is stable,’fundamental’investors always outperform than the’technology’investors. The main innovations of this paper lie in two aspects. On one hand, the study of Agent-based Computational Finance based on a multi-assets market is a big vacancy in domestic academic research at present. Even in foreign countries it is also very rare. This paper enriches the study result. One the other hand, the model used in this paper contains a variety of assets and sets flexible information interaction mode between investors. Compared to the model in the past research which based on the extremely strict market restrictions and assumptions, the model in this paper is wider in scope. |