高频数据连续部分杠杆效应的估计

对股票价格与其波动率之间的负相关性的发现,引发了对高频金融数据杠杆效应的研究热潮.对于高频数据连续时间条件下满足伊藤半鞅模型的对数价格过程和波动率过程,定义了连续部分杠杆效应(CLE),并用临近窗口和向下截断方法,采用二次变差来构造相应的估计量,进一步研究了该估计量的相合性和渐近正态性,最后给出了定理证明.

Estimation of Continuous Partial Leverage Effect on High Frequency Data

The discovery of the negative correlation between stock price and its volatility has triggered an upsurge of research on leverage effect of high-frequency financial data. For the logarithmic price process and volatility process satisfying ITO semi martingale model under the continuous time condition of high frequency data, the continuous partial leverage effect (CLE) is defined, and the method of near window and downward truncation is used, The quadratic variation is used to construct the corresponding estimator. The consistency and asymptotic normality of the estimator are further studied. Finally, the theorem is proved.