News sentiment and stock return: Evidence from managers’ news coverages

Abstract

In this paper, we construct a monthly news-based manager sentiment (SM) based on the tone of managers’ news reports. Statistically, SM has excellent predictability for the subsequent month's return in both in- and out-of-sample periods. we find that SM contains additional information to forecast stock returns compared to popular economic predictors. After analysing the prediction performance at different sentiment levels, it is found that the prediction power of SM is far better in the high sentiment period than in the low sentiment period. In terms of investing, SM also generates considerable economic value for investors who use forecasting information to optimise their stock portfolios.