Master's – Revisiting the relationship between reputation and financial performance: an analysis based on text mining and machine learning
Advisor: Profa. Dra. Adriana Backx Noronha Viana
Comission: Profs. Drs. Rui Manoel Teixeira Santos Dias, Fábio Claro Coimbra and Tatiana Albanez
Link youtube: https://youtu.be/abAZlAN9ryU
Corporate reputation is a valuable, rare, difficult to imitate and non-replaceable intangible asset, which leads to a sustainable competitive advantage. Previous empirical research confirms that there is a positive relationship between corporate reputation and financial performance. Based on the theory of Resource Based Vision and Signaling theory, this study explores the effects of corporate reputation, measured through news in the online media, on the profitability and market value of listed public companies on B3 - Brasil Bolsa, Balcão. The media acts as an intermediary in the signaling process between companies and stakeholders, which mitigates information asymmetry. In addition to these aspects, this research also examines the interaction between the dimensions of corporate reputation, which can reinforce its consequences on financial performance. The techniques of linear regression and regression through the machine learning algorithm Gradient Boosting Machines are used to empirically test a theoretical model that links corporate reputation to financial performance in a multidimensional way. The research data set consists of financial indicators and online news texts from all companies listed on B3. The results of the research showed that corporate reputation influences the profitability and market value of companies, in accordance with the Resource-Based View theory, since it is a strategic resource socially built from the perceptions of stakeholders.
*Abstract provided by the author