Commonality in the textual sentiment of corporate disclosures: Model and applications.

It is well established that the communication between firm management and stakeholders goes beyond the quantitative information of the numbers reported by the firm in the earnings report, balance sheet or other corporate disclosures. In fact, the firm stakeholders also process and analyze the sentiment of the qualitative information of corporate disclosures while making their decisions. However, no consensus has been reached on how to accurately measure sentiment. It has often been left to the sole discretion of the researcher to define the sentiment of a financial disclosure, which has led to the development of different methodologies, different underlying data sets or different libraries. Given the relevance of the field of textual analysis in finance, an important research question is how to estimate the sentiment of a corporate disclosure accurately in order to maximize its informativeness for future firm performance. The objective of this project is to fill this gap and analyze the additional information that can be extracted from the sentiment data by exploiting the longitudinal nature of corporate disclosures of a cross-section of firms observed at different points in time.

Keywords:textual sentiment, corporate disclosures

Disciplines:Management