Benchmarking Operational Risk Stress Testing Models

17 Pages Posted: 17 Jul 2019

See all articles by Filippo Curti

Filippo Curti

Federal Reserve Banks - Federal Reserve Bank of Richmond

Marco Migueis

Federal Reserve Board

Rob T Stewart

Federal Reserve Bank of Chicago

Multiple version iconThere are 2 versions of this paper

Date Written: 2019-05-28

Abstract

The Federal Reserve's Comprehensive Capital Analysis and Review (CCAR) requires large bank holding companies (BHCs) to project losses under stress scenarios. In this paper, we propose multiple benchmarks for operational loss projections and document the industry distribution relative to these benchmarks. The proposed benchmarks link BHCs' loss projections with both financial characteristics and metrics of historical loss experience. These benchmarks capture different measures of exposure and together provide a comprehensive view of the reasonability of model outcomes. Furthermore, we employ several approaches to assess the conservatism of BHCs' stress loss projections and our estimates for the conservatism of loss projections for the median bank range from the 90th percentile to above the 99th percentile of the operational loss distribution.

Keywords: Benchmarking, Operational Risk, Stress Testing

JEL Classification: G28, G21, G32

Suggested Citation

Curti, Filippo and Migueis, Marco and Stewart, Robert T, Benchmarking Operational Risk Stress Testing Models (2019-05-28). FEDS Working Paper No. 2019-038. Available at SSRN: https://ssrn.com/abstract=3418947 or http://dx.doi.org/10.17016/FEDS.2019.038

Filippo Curti (Contact Author)

Federal Reserve Banks - Federal Reserve Bank of Richmond ( email )

P.O. Box 27622
Richmond, VA 23261
United States

Marco Migueis

Federal Reserve Board ( email )

20th Street and Constitution Avenue NW
Washington, DC 20551
United States

Robert T Stewart

Federal Reserve Bank of Chicago ( email )

230 South LaSalle Street
Chicago, IL 60604
United States

Register to save articles to
your library

Register

Paper statistics

Downloads
19
Abstract Views
88
PlumX Metrics