Verbal Qualitative Information from Social Networks and Stock Performance of Tunisian Financial Companies

Abstract

Social networks have invaded our world and changed the way we communicate. Even the financial world has turned to social media, with Twitter leading the way. As a result, social media is becoming an undeniable tool that affects the stock market. In this study, we consulted the articles published on the facebook page ilboursa of Tunis. The published articles contain an announcement of the news of the movements of the shares of the listed companies by using negative and positive words. Our database is composed by a manual counting of these qualitative verbal information. We used the list of words of the psychological dictionary "Harvard-IV-4". Our research focuses on 26 Tunisian financial companies listed on the Tunis Stock Exchange, over a one-year horizon, from 01 January 2015 until 31 December 2015. We used the GMM (Generalized Method of Moment) on dynamic panel. The generalized method of moments analyzes two models in which five days of delay of the dependent variable appear as explanatory variable. The results of the study are twofold. First, the Tunis stock exchange seems to react positively to positive qualitative information. Second, it reacts negatively to negative qualitative information. Among other things, the impact of stock returns on information is quite important. It is always necessary to master the tool of social networks to disseminate good and relevant qualitative financial information on the financial market.