基于HAR—RV模型的中国证券市场异质性研究

随着高频数据可以越来越方便的获得，过去的ARCH模型及GARCH模型已经不能满足高频数据研究的需要。相比于模型波动率，已实现波动率能够更直接、准确地描述波动率的特征。研究表明，已实现波动率具有长记忆特性，其解释之一即为异质市场假说，即市场中存在异质交易者。文章选取了HAR—Rv模型对已实现波动率进行建模，并且通过回归分析，证明了我国股票市场交易者存在异质性。

Researching of Heterogeneity in Chinese Securities Markets Based on HAR-RV Model

Because of the convenient acquisition of high frequent data, the ARCH ＆ GARCH model cannot satisfy present research anymore. Compared to model volatility, realized volatility has many better characteristics. Some research shows that realized volatility has long memory. One of the possible reasons comes from the heterogeneous market hypothesis, which insists that there are heterogeneous traders in financial market. This paper chooses HAR-RV model in modeling realized volatility and proves that there are heterogeneous traders in Shanghai stock market.