Measurement and Quantification of Operational Risk an Application of Covariate–Var

Abstract

Oft-repeated catastrophic business failures and Basel II requirement of capital charges for operational risk (OR) have increased the awareness about OR. While the policy authorities are keen on banks adopting international best practices, yet measurement and quantification of OR has sparked a debate worldwide. While highlighting the difficulties in quantifying OR, we briefly recall the present methods for measuring and modeling OR as prescribed by the Basel committee. Using full information content of real-time operational loss data of a large Indian public sector bank, we demonstrate the superiority of covariate-VaR compared to standard Peak-Over-Threshold (POT) based approach for OR measurement. We conclude in saying that as the operational loss databases are built across banks, an effective methodology for risk modeling becomes imperative.