This paper investigates whether the text of the analyst report could provide additional value beyond the information on the recommendation or target price. Using Korean analyst reports from 2016 to 2018, sentiment dictionary was used to measure the opinion of the text. After that, the relationship between the opinion variables and the stock announcement returns is examined. The result shows that the textual opinion of the analyst report has a significant relationship with the stock return after the announcement, which was maintained even after the recommendation or target price are controlled. In particular, the opinion of text shows stronger information in negative tones. The results mean that even if the analyst's __EXPRESSION__ of intention is limited in the recommendation, text analysis can reveal more detailed information provided by the analyst.​

Keywords: textual analysis, analyst report, textual opinion, recommendation, target price