The Effect of the Informed and Noise Traders Perceptions from the Financial Reports on Stock Returns: Text Mining Approach

The purpose of this study is to investigate the effect of informed and noise investors' perceptions of financial statements on stock returns. In this regard, the qualitative data of the reports of the board of directors of 116 companies listed on the Tehran Stock Exchange during the period 2011-2019 have been used. To qualitative analyzing the reports of the board of directors and extracting the textual elements considered by the two groups of informed and noise traders, the text mining and Lasso regression approach was used and to separate the capital market traders into two groups of informed and noise traders, Kalman filter was used. Findings of the study show that both groups of informed and noise traders can achieve abnormal returns by using the information of board reports and basing words and separating them into words with real and fact basis. In the case of mix-meanning based words, the perception of informed traders can affect abnormal stock returns, while noise traders are unable to distinguish mix-meaning words and do not seem to pay much attention to them in their decisions. The results of the research are generally in line with the theory of noise traders and show the behavioral basis (imitative effect) in the Tehran Stock Exchange.