基于符号收益和跳跃变差的高频波动率模型
发布时间:2018-01-01 10:14
本文关键词:基于符号收益和跳跃变差的高频波动率模型 出处:《管理科学学报》2017年10期 论文类型:期刊论文
更多相关文章: 高频波动率模型 跳跃 已实现半变差 符号跳跃变差
【摘要】:基于符号收益率的视角,对现有的HAR-RV类及其跳跃扩展模型进行相应分解,构建新型的HAR-RV类波动率模型.进一步,结合符号收益和不同的跳跃识别检验方法,提出了包含符号跳跃变差的HAR-RV类模型,并利用样本外滚动窗预测技术和"模型信度设定"(MCS)检验法评价了各种新旧HAR-RV模型对我国沪深300股价指数波动的预测能力.结果表明:基于C_TZ跳跃识别检验的符号跳跃变差能显著改善波动率模型的短期预测能力,但在中长期波动预测时,符号跳跃变差未能明显提升HAR-RV类模型的预测精度;新提出的HAR-S-RVTJ-TSJV模型和HAR-S-RV-TJ模型分别在对短期(未来1天)和中长期(未来5天和20天)的波动预测检验中,展现出了最高的预测精度.
[Abstract]:Based on the symbolic rate of return, the existing HAR-RV class and its jump extension model are decomposed accordingly, and a new HAR-RV volatility model is constructed. Combined with the sign income and the different test methods of jump recognition, the HAR-RV class model including the symbol jump variation is proposed. And using the prediction technology of rolling window outside the sample and "model reliability setting" MCSs). The new and old HAR-RV models are used to predict the fluctuation of Shanghai and Shenzhen 300 stock price index in China. The results show that:. The symbolic jump variation based on CSTZ Jump recognition test can significantly improve the short-term prediction ability of volatility model. However, in the medium and long term fluctuation prediction, the symbol jump variation can not obviously improve the prediction accuracy of HAR-RV model. The new HAR-S-RVTJ-TSJV model and the HAR-S-RV-TJ model are tested for short-term (next 1 day) and medium (5 and 20 days) volatility forecasting, respectively. It shows the highest prediction accuracy.
【作者单位】: 西南交通大学经济管理学院;
【基金】:国家自然科学基金资助项目(71371157;71671145) 教育部人文社会科学基金规划资助项目(15YJA790031;16YJA790062) 四川省科技青年基金资助项目(2015JQO010) 四川省社会科学高水平研究团队资助项目 中央高校基金科研业务费专项资金资助项目(26816WCX02)
【分类号】:F832.51;O21
【正文快照】: 0引言对金融资产波动率的描述和预测是现代金融学理论和实务界研究的热点和难点问题.近十多年,随着计算机技术的飞速发展和金融高频数据获取性逐渐增强,基于日内高频数据的波动率测度和预测受到了国内外学者的广泛关注.其中,具有代表性的是Andersen和Bollerslev[1-2]提出的已,
本文编号:1364068
本文链接:https://www.wllwen.com/jingjilunwen/touziyanjiulunwen/1364068.html