Comparing the Informed and Noise Investors’ Perception of the Tone of Financial Statements and Its Impact on Stock Returns: A Text-Mining Approach

Fluctuations in the companies’ stock returns depend deeply on the nature of informed and noise traders’ behavior and inclinations. In qualitative accounting reports, managers try to influence the investors’ behavior and inclinations by choosing words with specific semantic orientations. The purpose of the present study is to compare the impact of informed and noise traders' perceptions on the company's stock returns; it also investigates, from the informed and noise traders’ perspective, the effect of the board of directors’ report tone on the stock market reaction. In this regard, qualitative data from the board of directors’ reports, text mining approach, LASSO Regression have been used. In order to separate capital market traders into two groups of informed and noise ones, Kalman Filtering has been applied. All words employed in each report have been divided into three categories: words based on fact, words based on emotion, and words with a mixed meaning. The study thus compares the effect of informed and noise traders' perceptions of each type of word on stock returns. The statistical population of this study includes all companies listed on the Tehran Stock Exchange during the period 2012-2020; the statistical sample includes 116 firms selected through the systematic removal method. The research findings indicate that informed and noise traders can achieve abnormal returns by using fact-based and emotion-laden words, respectively, Also the tone of the board of directors’ reports is an influential factor just for informed investors; it does not affect the noise traders’ behavior and reaction.