基于非理性交易行为的证券收益分布特征研究  

A Research into the Distribution Characteristics of Securities Return Based on Irrational Trading Behavior

作　　者：赵鹏举[1,2] 

机构地区：[1]西北工业大学,陕西西安710072 [2]中原工学院经济管理学院,河南郑州450007

出　　处：《经济经纬》2009年第2期149-152,共4页Economic Survey

基　　金：国家自然科学基金项目(70572050)和(70771102)的资助

摘　　要：笔者使用一个异质交易者并存的市场收益模型,证明非理性的反馈交易导致证券收益的长程相关,收益序列的长程相关性引起了异常扩散,由于扩散的方差大于标准布朗运动的方差,造成收益概率密度函数的厚尾现象。在模型的基础上,利用蒙特卡罗仿真方法生成了股市收益时间序列样本,计算了仿真数据的峰度、尾指数和自相关系数,并与我国上证指数真实收益序列的相应指标进行了对比分析。模型的仿真结果显示了证券收益的尖峰厚尾特征以及收益序列的自相关性。This paper establishes a market return model with the participation of irrational traders and rational traders and proves that irrational feedback trading causes the long range correlation of securities return, the long range correlation of return series causes unusual diffusion. That the variance of diffusion is bigger than that of standard Brown Motion causes the fat tail of the return probability density function. Based on the model, the author uses Monte Carlo simulation method to produce a time series, computes the kurtosis, tail index and the autoeorrelation eoefficient and compares it with the corresponding indexes in the real return series of Shanghai stock market. The simulation result of the model shows the characteristic of excess kurtosis and fat - tail of securities return and autocorrelation of stock market return series.

关 键 词：非理性交易者 反馈交易 尖峰厚尾分布 收益自相关性 

分 类 号：F830.91[经济管理—金融学]