论文部分内容阅读
通过利用三种主流的跳跃检验识别方法,结合已实现异质自回归模型和基于已实现正负变差的思想,构建不同类型的高频波动率模型,以此评价不同跳跃检验的跳跃成分对波动率预测精度影响的优劣.MCS实证结果发现,跳跃作为解释变量有助于提高波动率模型的预测能力.另外,本文还发现LM检验得到的跳跃成分更有助于提高波动率模型的预测精度.最后,基于已实现正负变差和LM跳跃检验构建的新的波动率模型,HAR-S-RV-JV,相比其他的高频波动率模型,具有更好的预测表现能力.“,”By using three mainstream methods of jump tests,combining the HAR-RV model and based on the thought of Realized Semi-variance,we construct different types of high frequency volatility models in order to evaluate the effect of volatility prediction accuracy for jumping components in different JUMP TEST.The empirical results show that jumping,as an explanatory variable,helps to increase the predictive ability of the volatility model.In addition,we have found that the jumping component from LM test is more helpful to improve the prediction accuracy for volatility models.Finally,HAR-S-RV-JV,the new volatility model that is based on the HAR-RV model and LM jump test,performs better predictive ability compared with other high frequency volatility models.