Site visit information content and return predictability: Evidence from China

In this paper, we use frequency of related phrases in site visit summary reports to denote the site visit content, and study whether site visit content reflecting institutional investors' market concerns can predict Chinese stock market return. We find that site visit content has greater forecasting power in Chinese stock market returns than other economic predictors after comparing out-of-sample R-2. The predictability is both statistically and economically significant. Additionally, our results also suggest that the particular information content has better forecasting power than general content in site visit summary reports.