以往的文献对于股票市场之间的相关性研究大多是基于低频数据的.本文则采用高频金融数据,对上海和深圳股票市场的相关性进行研究.首先提出了基于高频数据的赋权已实现变差估计量和赋权已实现协变差估计量.然后,给出了基于赋权已实现变差和赋权已实现协变差的金融市场之间相关系数的定义.最后,采用Bayes方法来检测相关系数时间序列的变结构点,来发现金融市场间的相关性是否存在显著变化.通过对上海和深圳股票市场进行的实证研究发现,两个市场的相关系数在变结构点存在显著变化.
Currently a large literature about the correlations between stock markets are based on the low frequency data. This paper studies the correlations between Shanghai stock market and Shenzhen stock market by using high frequency data. Firstly, a weighted realized variance and a weighted realized covariance which are based on high frequency data are put forward. Then the definition of the correlation coefficient which is based on the weighted realized variance and weighted realized eovariance is given. Finally, the structure change points in the correlation time series are tested by using Bayes method. Through the empirical research on Shanghai stock market and Shenzhen stock market, it is found that the correlations coefficients between the two stock markets have remarkable changes before and after structure change points.