Operational risk modeling under the loss distribution approach: estimation of operational risk capital by business line versus risk category

Hrair Danageuzian and Re-Mi Hage

Abstract

This research models operational risk data via the loss distribution approach under Basel II using open-source data consisting of 3192 operational loss events between 2009 and 2018. The approach is implemented per business line and a second time per risk category. The capital requirement is determined for each case using the RSTUDIO environment in the R programming language. Significant differences are identified between the yearly capital requirements obtained for each of the two cases. The total ten-year period is also considered, and a weighted average of the yearly capital charges calculated. The business-line method records a capital charge that is around 15% lower than the risk-category method. Ultimately, to diminish the impact of operational risk, the larger of the two capital charges is recommended for the next year after our sample period. Our findings contribute toward a better understanding of the composition and distribution of operational risk data across risk classes and the corresponding operational risk capital requirements.