基于HAR模型对中国股市已实现极差的研究  

The Research on Realized Range Based on HAR Model in China Stock Market

作　　者：黄旭东[1] 葛靖[1] 

机构地区：[1]安徽师范大学数学计算机科学学院,安徽芜湖241003

出　　处：《安徽师范大学学报（自然科学版）》2016年第1期5-10,共6页Journal of Anhui Normal University(Natural Science)

基　　金：国家自然科学基金项目(11201006);全国统计科学研究项目(2015LZ54)

摘　　要：在20支上证A股股票高频日内数据的基础上,考虑成交量、交易次数和各种形式的隔夜回报对已实现极差的影响.为了考查影响效果,我们将加入这些滞后变量的增广HAR模型同传统HAR模型进行比较.研究结果表明,在样本内预测上这些滞后变量都对已实现极差有一定的影响,然而在样本外预测效果方面,加入这些滞后变量后的增广HAR模型同传统HAR模型相比并没有显著提高.Based on the high--requency intraday data of 20 stocks of A-share index of Shanghai Stock Exchange,we investigate the effects of trading volume and number of trades as well as in various specifications of overnight returns for realized range. For this purpose,the augmented HAR models by adding these lagged variables are compared with the traditional HAR model. The results show that these variables exhibit some in-sample forecasting power,but the accuracy improvement of out-of-sample forecasts is non-significant.

关 键 词：波动率 HAR模型 已实现极差 成交量 交易次数 隔夜回报 

分 类 号：F830[经济管理—金融学]