CREDIT EXPOSURE, FUNDING STABILITY AND FRAUD DETECTION IN QUOTED NIGERIAN COMMERCIAL BANKS

This study examined the effects of TLTA and NSFR on fraud detection in quoted Nigerian commercial banks, with firm size included as a control variable. The study adopted an ex post facto design and used quarterly observations for ten quoted commercial banks over 2014–2023, yielding 400 bank-quarter observations. Fraud detection was proxied by a binary indicator derived from discrepancies in net income. Binary logistic regression was estimated from the dataset reported in Appendix M of the underlying thesis. The results show that TLTA had a coefficient of 0.7247 (p=0.3115), while NSFR had a coefficient of 0.2667 (p=0.4286). Neither focal variable was statistically significant at the 5% level. Firm size was positive and significant (coefficient=0.1041, p=0.0077). The likelihood-ratio test for the model was significant (p=0.0267), although McFadden’s pseudo-R² was 0.0166. The study concludes that TLTA and NSFR, when considered with firm size, are better treated as components of a broader fraud-risk screening framework rather than stand-alone evidence of fraud