The Intraday Periodicity and Long-memory Characters in High-frequency Data of China Stock Market
中国股市高频数据中的周期性和长记忆性

By FFF regression of Andersen and Bollerslev(1997), we analysis the periodicity of Shanghai stock index 5-min high frequency data and the long memory characters in filtered absolute returns. We document that the periodicity of intraday absolute returns is stronger than that of raw intraday returns, and FFF regression is an efficient way of determining the periodicity. After comparing the previous results, we found that the long memory in high-frequency absolute returns is stronger than that of daily returns.