几种异质自回归实现波动率模型的有效性——基于中国股票市场高频数据的实证研究

The Effectiveness of Several Models of Heterogeneous Autoregressive Model for the Realization of Volatility:An Empirical Study Based on the High Frequency Data of the Stock Market in China

本文对HAR-RV-CJ模型了进一步的改进,提出的假定模型回归系数为1的整合的IHAR-RV-CJ模型,实证对比了HAR模型,含跳的HAR-J和HAR-CJ模型在系数通过合理的约束为整1后以及原始模型对中国股市波动率的刻画和预测问题。实证结果显示,在四种损失函数综合对比下,IHAR模型同样适用于刻画对中国股市波动率,相对于原始模型提高了模型的预测精度,且我们所构建的IHAR-RV-CJ模型预测的有效性要高于IHAR-RV-J模型。

This paper further improves the HAR-RV-CJ model,puts forward the integrated IHAR-RV-CJ model which the assumed model regression coefficient is 1.The empirical comparison of the HAR model is carried out.The jumpy HAR-J and HAR-CJ models are compared when the coefficients are reasonably constrained to be 1.The description and prediction of the volatility of Chinese stock market by the original model are put forward.The empirical results show that the IHAR model also can be used to describe the volatility of Chinese stock market with the comprehensive comparison of the four kinds of loss functions.Compared with the original model,the prediction accuracy of the model is improved,and the validity of IHAR-RV-CJ model is higher than that of IHAR-RV-J model.