高频金融数据的长记忆SV模型分析——基于FFF回归

High-frequency Financial Data's Long Memory SV Model Analysis Based on FFF Regression

高频金融数据的分析与建模是金融计量学的一个全新的研究领域。与低频数据不同,高频数据通常具有“日历效应”和波动长记忆性。本文在使用弹性傅立叶形式(FFF)回归技术消除“日历效应”的基础上,针对高频数据的波动长记忆性建立了长记忆SV模型,结果发现高频数据的波动持续性大大降低。

High-frequency financial data analysis and modeling is a new research field in financial econometrics. Unlike low frequency data, high frequency data has the calendar effects and long memory volatility. The paper uses Flexible Fourier Form Regression to fit the calendar effects, and constructs long memory SV model for Shanghai Stock Index. Through this research,it can be discovered that the volatility persistence of high frequency data is very low.