ISBIS4 Abstract

Contact Author's Name: Pavel Shevchenko
Title of Abstract: Operational Risk Modelling and Quantification
Author(s): Pavel Shevchenko
Affiliation: CSIRO Mathematical and Information Sciences, Sydney, Australia

The development of Basel II Advanced Measurement Approach (AMA) for quantification of Operational Risk is of strong importance for the banks. The AMA allows to quantify operational risk capital charge using internal measurement approaches. It is expected that institutions receiving AMA accreditation will benefit by having lower capital requirements for operational risk. There is no established measurement methodology for AMA and the Basel II allows for flexibility in the approach requiring, however, addressing some important elements such as: internal data, external data, scenario analysis, and factors reflecting business environment and internal control systems. Emerging best practices share a common view that AMA based on a Loss Distribution Approach (LDA) is the soundest method.

The LDA is based on modelling of frequency and severity of individual operational risks. Given estimated frequency and severity of individual risks, the distribution of loss over all risks and corresponding regulatory capital charge (difference between Value at Risk at the 0.999 level and expected loss) can be calculated using Monte Carlo simulation method. Development of a meaningful quantitative approach to operational risk is a challenging task and rich field for new research. In this talk we address a number of important aspects in the quantification of operational risk using LDA such as:

- Data thresholds (available data are collected above some reporting threshold levels)
- Dependence between risks (e.g. copula method)
- Scaling of the external loss data
- Modelling of the risk distribution tail and insurance