投资者注意的风险补偿对股市回报的预测

利用历史累积交易金额数据,本文构造了中国股票市场增量注意风险补偿和存量注意风险补偿,并检验其对中国股票市场收益率的预测能力。样本外检验结果显示,以上两种注意风险补偿均能显著预测下个月中国股市的超额收益率,其R²分别达到了2.68%和2.50%;与中国股票市场中其他预测变量相对比,增量注意和存量注意风险补偿表现出更强的预测能力。此外,基于不同的样本外检验期、不同的风险厌恶参数以及五种不同的变量构造方式,投资者注意风险补偿均产生显著的预测能力。围绕着经济周期波动,本文对注意风险补偿的预测能力进行了解释,同时还发现,相较于经济衰退期间,经济繁荣期间的投资者注意风险补偿样本外预测能力更强。

Risk Premium of Investor Attention and the Predictability of Stock Return

Using the history accumulative trading amount data, this paper constructs the risk premium of change in investor attention and level of investor attention, and test its predictive power for aggregate stock market excess return in Chinese stock market. Under the out-of-sample predictability test, the results show that the risk premium of two investor attention significantly predict the next month excess return on Chinese stock market with R2 being 2.68% and 2.50%, respectively. Compared with other predictors in the Chinese stock market, the change in investor attention and the level of investor attention risk premiums show a stronger predictive power. In addition, based on alternative forecasting windows, alternative risk aversion choices and alternative five constructing methods, the risk premium of investor attention also significantly predicts excess returns. Economically, the predictive power of investor attention risk premium for stock market comes from the changes in the business cycle. The predictive power is further improved during economic expansions.