Topic tones of analyst reports and stock returns: A deep learning approach

We present a novel approach that analyzes topics and tones of analyst reports using a deep neural network in a supervised learning approach. By letting trained classifiers evaluate topics and tones of the reports, we find that incorporation of topic tones significantly enhances the accuracy of predicting cumulative abnormal returns, increasing adjusted ${\mathrm{R}}^2$ from 6.1% without considering textual information to 17.9% with detailed topic tones. This improvement is primarily driven by the inclusion of opinion and corporate fact type of topics. Our findings highlight importance of topic assessment to make the most use of analyst reports for informed investment decisions.