行业关注度与股票横截面收益率——基于百度行业搜索指数的研究

Industry-wide Attention Based on Internet Searches and the Cross-section of Stock Returns

本文以行业名称为关键词爬取百度搜索指数作为行业关注度的代理变量,研究投资者行业关注度与股票横截面收益率的关系。结果表明,行业搜索量能正向预测下一天的股票收益率,基于这一指标构建的多空策略每年可获得4.89%(4.28%)的等权(市值加权)收益率。与投资者关注度假说一致,行业关注度导致的价格压力效应在卖空约束强的股票中更突出,且长期来看价格压力会消退。进一步的研究表明,行业关注度对股价的预测能力强于个股关注度。

This study investigates how investor attention on industries affects the cross-section of stock returns,using aggregate search frequency from the Baidu Index based on industry names as a proxy for industry-specific attention.We find that an increase in the index predicts higher stock returns the next day.A long-short strategy based on this attention proxy yields an equal-weighted(value-weighted) annual return of 4.89%(4.28%).The buying pressure which will revert is stronger among stocks with more binding short-sales constraints.Further analysis shows that industry-level attention proxy is dominant over individual stock-level search volume index in predicting stock returns.