Operational Risk: The Sting is Still in the Tail But the Poison Dependson the Dose
IMF Working Papers, October 1, 2007
Source details
- Canonical URL
- Operational Risk: The Sting is Still in the Tail But the Poison Dependson the Dose
Other formats
Bibliographic details
- Authors: Andreas Jobst
- Published: October 1, 2007
- Series: IMF Working Papers
- DOI: https://doi.org/10.5089/9781451868036.001
Summary
- Paper investigates generalized parametric measurement methods of aggregate operational risk in compliance with the regulatory capital standards for operational risk in the New Basel Capital Accord ("Basel II").
- Operational risk defined as the risk of loss resulting from inadequate or failed internal processes and information systems, from misconduct by people or from unforeseen external events.
- Analysis informs an integrated assessment of the quantification of operational risk exposure and the consistency of current capital rules on operational risk based on generalized parametric estimation.
Analytical focus and methods
- Focus on generalized parametric estimation approaches for aggregate operational risk.
- Techniques and concepts referenced in the paper (as listed in source keywords):
- Advanced Measurement Approaches (AMA)
- extreme value theory (EVT), generalized extreme value (GEV) distribution, generalized Pareto distribution (GPD)
- g-and-h distribution
- peak-over-threshold (POT) method
- percentile level, threshold quantile, threshold-quantile surface
- Value-at-Risk (VaR)
- MLE estimation algorithm
- estimation method, estimation risk, descriptive statistics, surface estimation
- Application domain: Banking; Capital adequacy requirements; Financial regulation and supervision; Operational risk measurement.
Key findings and themes (as described)
- The "sting is still in the tail" — emphasis on extreme tail behavior and fat tail behavior in operational risk losses.
- Assessment centers on consistency of current capital rules under Basel II when confronted with generalized parametric extreme-tail estimations.
- Highlights estimation risk and the implications of tail modeling choices for capital estimates and risk management.
Policy implications and considerations
- Importance of choosing appropriate extreme-tail models (EVT, GPD, GEV, g-and-h) when quantifying operational risk for regulatory capital.
- Relevance of estimation technique selection (including MLE algorithms and POT methods) to the reliability of operational risk capital estimates.
- The paper's analysis is intended to inform the evaluation of Basel II operational risk capital standards and their consistency with parametric tail estimation outcomes.