论文部分内容阅读
在充分解析价格扩散过程的基础上,本文构建了融入时变噪音因素的新过程,并将其离散化后转换成状态空间模型。然后,利用Kalman滤波方法并借助EM算法估计未知参数,实现了有效度量噪音收益的目的。最后,以上证综指1991年1月4日至2012年2月24日的周数据为样本探析中国股市噪音收益情况,结果表明:期间中国股市的噪音收益水平处在-23.00%~83.51%,且存在右偏及尖峰特征,进一步分析表明投资者理性程度及监管是影响噪音收益的重要因素。
On the basis of fully analyzing the process of price diffusion, this paper constructs a new process that incorporates the time-varying noise factors and transforms them into a state-space model after being discretized. Then, Kalman filtering method and the EM algorithm are used to estimate the unknown parameters to achieve the purpose of effectively measuring the noise benefit. Finally, using the week data from January 4, 1991 to February 24, 2012 of the Shanghai Composite Index as an example, the paper analyzes the noise return of the Chinese stock market. The results show that during the period, the noise return of the Chinese stock market is between -23.00% and 83.51% And there is a right-deviation and spike characteristics. Further analysis shows that investor rationality and regulation are important factors that affect the noise yield.