Textual analysis of sentiment in finance and accounting: A comprehensive review

Virginia Sotiropoulou¹ and Vasileios Sogiakas²

¹Postdoctoral researcher, Accounting Lab, School of Social Sciences, Hellenic Open University, Greece

²Professor, School of Social Sciences, Hellenic Open University, Greece

Correspondence: Virginia Sotiropoulou, Accounting Lab, School of Social Sciences, Hellenic Open University, Greece. Tel: 697.6878125. E-mail: virginia.sotiropoulou@ac.eap.gr

Abstract

The purpose of this study is to examine critically the literature of textual analysis of disclosures’ sentiment in finance and accounting. Specifically, this research work investigates, compares, and contrasts the various methodologies that have been implemented to date to measure textual sentiment of financial and accounting information. Furthermore, we present the attributes of textual information of disclosures by investigating thoroughly established research in the field. We summarize the significant and influential empirical findings regarding textual analysis of sentiment in finance and accounting. Moreover, emerging opportunities that hold promise for future research are appearing from the incorporation of artificial intelligence techniques and the extensive use of computer linguistic methods to extract the sentiment in finance and accounting research. The adoption of these methodologies for analyzing textual information in finance and accounting literature could enable researchers to evaluate more efficiently the quality of financial reporting.