Risk Assessment in Financial Auditing Using the E-Bayes Estimation Method

In financial auditing, accurately predicting the time to the first material error in financial reporting systems is critical for risk-based audit planning. This paper proposes a multi-model reliability framework to evaluate internal control failure risk using the Rayleigh (increasing hazard rate) and Exponential (constant hazard rate) distributions under Type-II progressive censoring that is a realistic representation of audit data constraints. We apply the E-Bayesian estimation method under SELF and LINEX loss functions, with three distinct hyper-prior structures to account for auditor uncertainty. Performance is evaluated using two robust metrics: Mean Squared Error (MSE) and Mean Absolute Error (MAE). Results from both real audit data and Monte Carlo simulations consistently show that the proposed approach enhances audit precision and aligns with ISA 315’s risk-based philosophy. This study provides auditors with a statistically rigorous yet practically feasible tool for internal control risk assessment.