Detecting Financial Statement Fraud Using Beneish M-Score and Financial Ratio Analysis: Evidence from Nigeria’s Consumer Goods Sector
Reuben Osita Nwosu *
Department of Auditing and Forensic Accounting, College of Public Sector Accounting. ANAN University, Kwall, Plateau State, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
This study examines the determinants of financial statement fraud among listed non-financial firms in Nigeria using the Beneish M-Score model and logistic regression analysis. The research draws on a purposively selected sample of 11 firms from an initial population of 19, covering 110 firm-year observations between 2015 and 2024. Fraud likelihood was modelled as a binary dependent variable derived from Beneish’s framework, while explanatory variables included leverage, profitability, liquidity, capital turnover, and earnings manipulation score (MSCORE). The analysis employed descriptive statistics, correlation matrices, and logistic regression with cluster-robust standard errors to control for firm-level heterogeneity. The findings reveal that leverage and MSCORE significantly and positively influence fraud likelihood, whereas liquidity and capital turnover exert significant negative effects. Profitability shows a marginally significant inverse association, implying that better-performing firms are less prone to engage in manipulation. The model demonstrated high predictive accuracy with an area under the ROC curve of 0.97, indicating strong discriminatory power. The study underscores the need for robust internal controls, stronger audit oversight, and enhanced regulatory surveillance to deter manipulation and promote reporting integrity in Nigeria’s capital market.
Keywords: Financial statement fraud, Beneish M-Score, leverage, profitability, logistic regression, Nigeria