Journal of Resource Management and Decision Engineering In Press 2027 09 01 A Data-Driven Model for Fraud Risk Assessment in Financial Statement Auditing: Empirical Evidence from Iraqi Commercial Banks A Data-Driven Model for Fraud Risk Assessment in Financial Statement Auditing: Empirical Evidence from Iraqi Commercial Banks 1 18 EN Journal Article 2026 06 16 <p>This study aimed to develop and empirically validate a data-driven model for assessing financial statement fraud risk in Iraqi commercial banks by integrating financial reporting indicators, audit-related signals, and corporate governance characteristics within a longitudinal panel-data framework. A quantitative, applied, longitudinal design was employed using secondary panel data from 18 Iraqi commercial banks over the period 2015–2024, yielding 180 bank-year observations. Fraud risk was modeled using financial statement indicators including accrual intensity, earnings quality, asset growth, revenue growth, changes in receivables relative to revenue, leverage, liquidity, return on assets, and bank size, together with auditor change, modified audit opinion, board independence, and audit committee effectiveness. Pooled OLS, fixed-effects, and random-effects estimators were examined. Model selection was performed using the Hausman test, while multicollinearity, heteroscedasticity, serial correlation, and cross-sectional dependence were assessed through VIF, modified Wald, Wooldridge, and Pesaran CD tests. Robust fixed-effects estimation and temporal out-of-sample validation were applied. The Hausman test supported the fixed-effects specification, , . The final model was statistically significant, , , with a within- . Accrual intensity was the strongest positive predictor of fraud risk, , , followed by changes in receivables-to-revenue, , , modified audit opinion, , , leverage, , , asset growth, , , auditor change, , , and revenue growth, , . Earnings quality, liquidity, profitability, bank size, board independence, and audit committee effectiveness significantly reduced fraud risk. Validation produced an , accuracy of 77.8%, and AUC of .821. The proposed model demonstrated that financial statement fraud risk in Iraqi commercial banks can be effectively assessed through an integrated combination of accounting anomalies, financial pressure indicators, audit signals, and governance mechanisms.</p> https://journalrmde.com/index.php/jrmde/article/download/431/398