Stochastic Modeling of Expected Credit Loss Under IFRS 9: A Monte Carlo and Scenario-Based Approach

Abstract The adoption of International Financial Reporting Standard 9 (IFRS 9) has significantly transformed the way financial institutions manage credit risk, moving away from the retrospective Incurred Credit Loss (ICL) model to a forward-looking Expected Credit Loss (ECL) approach. This paper presents a detailed examination of the IFRS 9 framework from both regulatory and quantitative standpoints. We compare advanced credit risk estimation techniques, including stochastic simulations and macroeconomically weighted probability of default models, to evaluate risk-adjusted losses under different economic scenarios. These models are implemented through numerical optimization to assess their effectiveness in estimating credit loss across various risk stages. The findings indicate that a multi-model approach enhances the precision and resilience of ECL estimates, particularly in volatile environments. Analysis of portfolio segmentation further reveals that retail secured loans exhibit lower credit loss rates in advanced risk stages, highlighting the role of collateral in mitigating losses. This study offers practical value for financial institutions seeking to strengthen risk management practices while adhering to IFRS 9 standards.