Pre-conference workshop 1: March 14th 2016

WORKSHOP 1: Operational risk losses and the macroenvironment

8:30 Registration and breakfast

9:00 U.S. banking sector operational losses and the macroeconomic environment
Using regulatory data from large financial institutions, we document a significant association between U.S. bank holding companies' operational losses and the U.S. macroeconomic environment. Banks face greater aggregate losses in a weakening environment, with such a relationship driven by specific types of losses. We also document a link between individual loss severity and the macroeconomy showing an association exists only at the upper end of the severity distribution. Macroeconomic conditions are related to other aspects of operational losses as well. In a strengthening economy, although banks take longer to detect the occurrence of operational losses, they are able to recover proportionately more. Our analysis makes a significant contribution on the link between the macroeconomic and financial institutions' risk beyond what has been documented in the finance literature so far.

  • There is a significant relation between banks' aggregate operational losses and the macroeconomic environment
  • The upper end of the individual loss severity distribution is linked with the state of the economy
  • Recoveries are also significantly affected by the macroeconomic conditions


10:00 Morning coffee break 1

10:15 Operational risk capital uncertainty under the loss distribution approach
In this paper we explore the uncertainty associated with operational risk capital estimates for US bank-holding companies (BHC) that are required to follow the Advanced Measurement Approach (AMA). Basel's AMA for operational risk requires banks to hold enough capital to cover the 99.9th quantile of their annual operational loss distribution. AMA banks are required to use four data elements in estimation: internal loss data, external loss data, scenario analysis, and business environment and internal control factors (BEICF). Most banks use the Loss Distribution Approach (LDA) to account for internal and external loss data in AMA operational risk capital estimates. The LDA simplifies the estimation of operational risk exposure by breaking it down to two sub-components, frequency and severity, to be estimated separately from historical losses. To obtain the distribution of total operational losses, the frequency and severity distributions are combined, typically through a Monte Carlo simulation.

  • Main sources of uncertainty
  • Measuring uncertainty
  • Which factors affect the heavy-tailness of severity distribution

Marco Migueis, Banking Supervision and Regulation - Quantitative Risk, FEDERAL RESERVE BOARD GOVERNORS

11:15 Morning coffee break 2

11:30 Improving the robustness of operational risk modeling through external data scaling
One of the biggest challenges that banks face in modeling operational risk is the substantial instability of risk estimates from one period to the next. The key elements responsible for this instability are the heavy-tailness of loss distributions and insufficient loss data. To mitigate data limitations, the advanced measurement approach (AMA) under Basel II requires that the operational risk measurement system complements internal bank's loss data with relevant external data. However, the heterogeneity of external data makes combining it with internal data a challenging task. In this paper, we propose a scaling method to combine external and internal data, focusing on the tail of the loss distribution. The method is based on our finding that the loss severity of tail losses is related to bank size. We demonstrate that our method improves the robustness of operational risk estimates using supervisory operational loss data. Our method can also be applied for modeling losses under stressed conditions.

  • How to incorporate external data
  • Proposed scaling methodology
  • Application to CCAR


12:30 Lunch

1:30 Benchmarking operational risk models
The 2004 Basel II accord requires internationally active banks to hold regulatory capital for operational risk, and the Federal Reserve's Comprehensive Capital Analysis and Review (CCAR) requires banks to project operational risk losses under stressed scenarios. As a result, banks subject to these rules have measured and managed operational risk more rigorously. But some types of operational risk - particularly legal risk - are challenging to model because such exposures tend to be fat-tailed. Tail operational risk losses have significantly impacted banks' balance sheets and income statements, even post crisis. So, operational risk practitioners, bank analysts, and regulators must develop reasonable methods to assess the efficacy of operational risk models and associated equity financing. We believe benchmarks should be used extensively to justify model outputs, improve model stability, and maintain capital reasonableness. Since any individual benchmark can be misleading, we outline a set of principles for using benchmarks effectively and describe how these principles can be applied to operational risk models. Also, we describe the benchmarks that have been used by the Federal Reserve in assessing Advanced Measurement Approach (AMA) capital reasonableness and benchmarks that can be used in CCAR to assess the reasonableness of operational risk loss projections. We believe no single model's output and no single benchmark offers a comprehensive view; and practitioners, analysts, and regulators must use models combined with rigorous benchmarks to determine operational risk capital and assess its adequacy.

  • Principles of benchmarking
  • AMA benchmarks
  • CCAR benchmarks


3:00 Afternoon coffee break

3:30 Using generalized additive models for stress testing operational risk losses
In this paper we investigate the main driving factors of operational risk losses. Specifically, we propose a model that allows us to link operational losses with both macroeconomic and firm-specific factors. Our findings show that there is a significant relationship between the severity distributions and certain internal and systemic factors, suggesting that factors can be used to improve the projection of future operational risk losses. In addition, our proposed model can be used to scale external loss data

  • Generalized additive models for location, scale and shape
  • Relation between individual loss severity parameters with both macroeconomic and firm-specific factors
  • Application to CCAR

Filippo Curti, Financial Economist, Supervision, Regulation and Credit, THE FEDERAL RESERVE BANK OF RICHMOND

5:00 End of workshop

Follow us
Lead Sponsor
Panel Sponsor
Cyber Risk Sponsors 2016
CPE Accredited
Hosted by
Media Partner