摘要
利用高频股票指数数据,本文构造了中国股票市场的已实现偏度,并检验了其对中国股票市场收益率的预测能力。实证结果显示,当前较低的已实现偏度可以显著预测下个月中国股票市场较高的超额收益率,样本内和样本外的R^2分别达到了3. 39%和2. 24%。在控制了一系列的其它股票预测变量之后,该结论依然成立。此外,基于四种不同的构造方法,已实现偏度对上海和深圳两个股票市场都具有显著的预测能力。在将所有不同的已实现偏度指标进行组合之后,预测能力得到了进一步提升。从经济解释上,本文发现已实现偏度对股票收益率的预测能力是通过影响股票市场的交易活跃程度,从而传导到股票市场收益率。
With high frequent stock index data, this paper constructs the realized skewness of Chinese stock market and test its predictive power for aggregate stock market excess return. Results show that the realized skewness significantly predicts the next month excess return on Chinese stock market. In- sample and out- of - sample R2 are 3.39% and 2. 24% , respectively. After controlling for a set of variables, this conclusion still holds. In addition, based on alternative 4 constructing methods, the realized skewness significantly predicts excess returns on both of Shanghai and Shenzhen stock markets. The predictive power is further improved when combine all 4 skewness measures together. Economically, the predictive power of realized skewness for stock market return comes from its ability to predict future trading activity of the Chinese stock market.
作者
陈坚
张轶凡
CHEN Jian;ZHANG Yifan(Department of Finance,School of Economics,Xiamen Universit;Hebei Branch,The Export-Import Bank of China)
出处
《金融研究》
CSSCI
北大核心
2018年第9期107-125,共19页
Journal of Financial Research
基金
国家自然科学基金面上项目(71671148)中期科研成果
关键词
已实现偏度
中国股票市场
样本外预测
Realized Skewness
China Stock Market
Out - of - sample Predictability
