Do Auditors Reveal Their True Opinions? Embedding Going-Concern Opinions for Corporate Default Prediction
This study investigates whether auditors’ going-concern opinions (GCOs) truly reflect their private assessments of firms’ financial health. By embedding textual representations of GCO disclosures into predictive models, we examine the extent to which these opinions contain incremental information for forecasting corporate defaults. The research bridges auditing judgment with machine learning techniques, offering insights into the credibility of auditors’ reports and their predictive value in credit risk assessment.