股票市场呈现出的多种波动形态,对其内在机制的探讨备受学者和业界的关注。建立多主体股市演化模型,并抽象出一个常微分方程系统描述其核心动力机制。对该系统的定性分析和数值实验表明,受交易主体模仿学习倾向、价格追踪倾向和交易策略转移倾向等市场环境参数影响,模型呈现出基础价值均衡、非基础价值均衡、周期和混沌等市场形态。
The fluctuation states of stock market draw much attention of researchers and practitioners. This paper constructs a multi-agent evolutionary model of stock market. A nonlinear dynamical system of ordinary differential equations(ODEs) is abstracted to depict the core dynamical mecha- nisms of market evolvement. Analysis on the ODEs by both analytical technologies and numerical experiments show that with different settings of parameters, such as dispositions of agent imitation learning, price chasing, and trade strategy alternation, the market model appears four kinds of states: fundamental equilibrium, non-fundamental equilibrium, period and chaos.