Beyond Value at Risk: A Stochastic Dominance Framework for Risk Management in Banking

Abstract

Objective: This study examines the application of stochastic dominance as a distribution-based framework for improving risk evaluation in banking beyond traditional metrics. Research Design & Methods: A quantitative analytical approach is employed using simulated and banking-style portfolio datasets. The study applies first-, second-, and third-order stochastic dominance to compare asset distributions and benchmark results against Value at Risk and Expected Shortfall. Findings: Results show no first-order dominance; however, second-order dominance consistently identifies conservative portfolios as optimal under risk aversion. Stochastic dominance reveals distributional differences not captured by conventional measures. Contributions: The study extends risk management literature by integrating nonparametric dominance techniques into banking portfolio evaluation. Novelty: This study introduces an empirical application of stochastic dominance in banking and demonstrates its superiority in capturing full distributional risk.