运筹与管理 ›› 2022, Vol. 31 ›› Issue (2): 99-103.DOI: 10.12005/orms.2022.0049

• 理论分析与方法探讨 • 上一篇    下一篇

基于Expectile和Realized GARCH模型的波动率预测

高雷阜, 李伟梅   

  1. 辽宁工程技术大学 运筹与优化研究院,辽宁 阜新 123000
  • 收稿日期:2020-08-07 出版日期:2022-02-25 发布日期:2022-03-11
  • 作者简介:高雷阜(1963-),男,辽宁阜新人,教授,博士生导师,研究方向:最优化理论与方法及应用;李伟梅(1988-),女,山西朔州人,博士研究生,研究方向:最优化理论与方法及应用。
  • 基金资助:
    辽宁省教育厅重点攻关项目(LJ2019ZL001)

Volatility Prediction Based on Expectile and Realized GARCH Model

GAO Lei-fu , LI Wei-mei   

  1. Institute for Optimization and Decision Analytics, Liaoning Technical University, Fuxin 123000, China
  • Received:2020-08-07 Online:2022-02-25 Published:2022-03-11

摘要: Realized GARCH模型是预测波动率的经典模型之一,最小化非对称二次损失函数的Expectile对收益率尾部分布更加敏感,我们在Realized GARCH模型的基础上引入Expectile提出Expectile-Realized GARCH模型。以沪深300指数的高频收益率为例建模分析,对比不同模型下的波动率预测效果,发现Expectile-Realized GARCH模型较Realized GARCH模型对波动率预测能力更好。其中,当风险水平为95%时,对应的Expectile-Realized GARCH波动率预测能力最好。

关键词: 波动率预测, Expectile, Realized GARCH模型, 高频数据

Abstract: The Realized GARCH model is one of the classical models to predict the volatility, and the expectile based on minimizing asymmetric quadratic loss function is more sensitive to the tail distribution of return rate. We establish Expectile-Realized GARCHmodel based on Realized GARCH and expectile. We model and analyze the high frequency return rate of Shanghai and Shenzhen 300 index,and compare the volatility prediction results under different models. It is found that the Expectile-Realized GARCHmodel is better than Realized GARCHmodel in forecasting volatility. When the risk level is 95%, the predictive volatility is the best.

Key words: volatility prediction, Realized GARCH model, expectile, high frequency data

中图分类号: