基于高频数据的深圳基金市场日内周期性研究

本文利用深市基金指数高频数据,采用Anderson和Bollerslev(1997)提出的弹性傅立叶回归(FlexibleFourierFormregression,即FFF回归)方法首次对深市基金市场进行了日内周期性的研究。通过对高频收益的定性分析,发现基金市场具有同股票市场相似的周期性,并对这一周期性进行了初步的理论解释。通过FFF方法,将该周期因子进行滤波处理以后,基金指数高频绝对收益不再具有明显周期性。FFF回归能较好地确定日内周期因子。

Intra-day Periodicity of Fund Market: a Case Study on High Frequency Data in Shenzhen

By using high frequency data of fund index in Shenzhen Exchange Market, FFF regression, which was put forward by Andersen and Bollerslev in 1997, is first applied to analyze the intraday periodicity of fund index return in Shenzhen Exchange Market. The qualitative analysis of high frequency return indicates that the fund market has the similar periodicity with the stock market, which is analyzed theoretically. And after the intraday periodicity factors are filtered through FFF regression, the high frequency absolute return of fund index doesn't have obvious periodicity any longer. Furthermore, FFF regression works well in determining the intra-day periodicity.