BANKRUPTCY PREDICTION AND FINANCIAL RISK ASSESSMENT IN EMERGING MARKETS: EVIDENCE FROM NIGERIA

Abstract

This study evaluates the predictive performance and associated risks of four prominent bankruptcy prediction models within the Nigerian business environment: the Altman Z-score, Ohlson O-score, and the locally validated IN01 and IN05 indexes. Utilizing a comprehensive dataset of Nigerian firms, the models are assessed across multiple metrics, including overall accuracy, sensitivity, specificity, precision, and F1 scores. Our empirical results demonstrate that the Nigerian-validated IN01 and IN05 models outperform the traditional Altman and Ohlson models, highlighting the critical importance of contextualizing bankruptcy prediction tools to local economic conditions. The analysis further includes Receiver Operating Characteristic (ROC) curves and confusion matrix heatmaps to provide nuanced insights into model discriminative power and classification errors. Findings underscore the practical implications for financial institutions and regulators in improving early warning systems and mitigating systemic risks in emerging markets. Limitations related to data quality and model scope are acknowledged, with recommendations for integrating machine learning and macroeconomic variables to enhance future predictive frameworks.