基于Agent异质行为演化的人工金融市场及其非线性特征研究

通过构建基于Agent的人工金融市场,试图从交易者个人异质行为演化的角度研究金融市场非线性特征的形成。市场中,Agent依赖个人行为特征,如:情绪、记忆长度等,来同时考虑基本面信息与价格趋势,针对当前市场状态,基于经验认知权衡二者后形成价格预期与交易行为。权重的自适应性更新揭示了个人行为的演化,其通过遗传算法与生成函数进化预测规则来实现。模拟实验表明,在做市商的价格生成机制下,当市场由自信的基本面分析者、技术分析者和自适应性理性交易者组成时,人工金融市场呈现出与真实市场相似的非线性特征——尖峰、厚尾,波动聚集性,长期记忆性以及混沌特征。这为探究导致市场产生非线性特征的行为因素提供了一个计算实验平台。

The Artificial Financial Market Based on Evolution of Agent's Behavioral Heterogeneity and Nonlinear Characteristics Analysis

After an Agent-based artificial financial market was built,the formation of financial market's nonlinear characteristics has been researched in this paper from the view of the evolution of investor individual's heterogeneous behavior.In our market,Agent will consider fundamental information and price tendency simultaneously relied on personal behavioral characters,such as mood,memory length and so on,make the tradeoff between them based on empirical knowledge,then form price expectation and trading behavior...