中文摘要：

利用理论分析和计算机仿真相结合的方法研究一个随机多主体的股市模型，理论分析得到基础价值均衡、非基础价值均衡、周期和混沌四种市场形态的典型参数设置，基于多主体的计算机仿真产生对应参数的价格序列．对此数据的统计分析发现：股市的所有市场形态都呈现收益率分布和波动时间依赖的标度行为，其中基础价值均衡形态下收益率累积分布指数和波动时间依赖的自相似指数最大，非基础价值均衡形态下两指数最小，周期和混沌形态下居中．

英文摘要：

This paper studies a stochastic multi-agent stock market model by both theoretical analysis and computer simulation. Theoretical analysis provides the parameter settings for different market states including fundamental equilibrium, non-fundamental equilibrium, periodicity and chaos. Agent based computer simulations produce the stock price series. Sta.tistical analysis of these data shows： markets of all market states present power law scaling of the return distribution and temporal dependence in volatility; the fundamental equilibrium state has the largest scaling exponent of the cumulative distribution for return and smallest self-similarity exponent of temporal dependence in volatility, and non-fundamental equilibrium state has the smallest and largest, with periodicity and chaos states in between.