跳跃强度对股市波动率建模与预测的影响研究

Impact of Jump Intensity on the Modeling and Forecasting of Stock Market Volatility

以上证综合指数为研究对象,基于C_TZ统计量估计出波动率的连续成分、跳跃的大小及次数,并通过自激点过程拟合跳跃强度。基于此,构建包含跳跃强度的HAR-TCI模型和HAR-TCJI模型,通过样本内和样本外两种预测方式考查跳跃强度在波动率建模与预测中的作用。实证结果表明,跳跃强度对已实现波动率有着显著的正向影响。在考虑跳跃对已实现波动率进行建模时,将跳跃强度和跳跃大小同时作为解释变量是最好的方式。上述结论在股票指数发生跳跃较为频繁的时期尤为明显。

By studying the Shanghai composite index and based on the C_TZ statistics,we make estimation of the continuous components,the size and the times of jumps of volatility,and then we get jump intensities through the self- exciting point process. Based on the estimated jump intensities,HAR- RV- TCI model and HAR- RV- TCJI model are constructed,and the HAR- RV- TCJ model is chosen as the benchmark model. The function of the jump intensity in modeling and forecasting volatility is studied through both in- sample and out- of- sample prediction methods. The empirical results show that the jump intensity has a significant positive impact on realized volatility. The best way to model and forecast realized volatility by jump component is to take both jump size and jump intensity as explanatory variables. The conclusions are particularly evident during the periods when jumps occur frequently.