Predicting Listed Company Profitability From Annual Report Narratives: Explanatory and Predictive Modeling

Abstract

The research combines explanatory and predictive modeling to examine the impact of annual report tone in predicting publicly traded companies’ profitability in Vietnam, an emerging Southeast Asian market. SGMM regression shows that this year’s narrative tone affects next year’s profitability. The study also used Scikit-learn Python machine learning algorithms to forecast profitability. The tone-based forecasting model that incorporates the company’s general and financial features predicts profitability is the most effective model. This study provides stakeholders such as investors and creditors with an approach to predict future profitability based on the narrative tone and expands theoretical understanding of its predictive power.