摘要
利用可以有效提取日内信息的"已实现"波动来度量高频金融时间序列的波动,使用排列熵方法分析"已实现"波动序列的顺序模式与序列之间的广义同步,利用全概率理论,在已知历史"已实现"波动顺序模式的情况下,预测下一个交易日的"已实现"波动处于不同水平的概率。使用上证综指与深圳成指的5分时收盘价进行实证研究,验证方法的可行性与有效性,发现这两个指数的"已实现"波动序列之间基本不存在广义同步,确定了它们的主要顺序模式,并基于主要顺序模式对"已实现"波动水平进行预测,结果显示主要顺序模式的条件顺序模式仍然占主要地位。
The realized volatility which can effectively exploit the information in intraday return data was applied to measure high - frequency intraday financial volatility, then permutation entropy methods were firstly introduced to analyze the ordinal struc- ture of realized volatility series and general synchronization between two series. The total probability theorem was used to forecast the next day volatility level after knowing the order patterns of history realized volatility. The ability and effectiveness of the meth- ods proposed were tested by Shanghai Composite Index and Shenzhen Component Index high frequency data whose sample period is 5 minutes before market close. The dominating order patterns were determined and it was found that these two indexs' realized volatility series have quite low generalized synchronization. Moreover, the next day volatility level was forecasted based on princi- pal order patterns. The forecasting results indicate that the conditional order patterns of the main order patterns are still dominant.
出处
《武汉理工大学学报(信息与管理工程版)》
CAS
2013年第4期599-603,共5页
Journal of Wuhan University of Technology:Information & Management Engineering
基金
国家自然科学基金资助项目(70971097)
关键词
“已实现”波动
排列熵
顺序模式
广义同步
realized volatility
permutation entropy
order patterns
generalized synchronization