RiskTech Forum

Not So Splendid Isolation: Operational Risk Model Risk

Posted: 21 February 2013  |  Author: Shane Murray  |  Source: Chartis

Model risk has become a more prominent subject in recent years, following the failure of numerous risk models to prevent the financial crisis. While financial institutions are attempting to reduce model risk across all risk disciplines, model risk is particularly acute for operational risk for a number of reasons.

The chief cause is that operational risk is not as mature a field as other risk disciplines. There is no commonly accepted benchmark model and financial institutions have developed their models in isolation. As a result, there is a great deal of variation around the use of scenarios, external data, Business Environment and Internal Control Factors (BEICFs), Units of Measure, and how to combine inputs into a single operational risk measure.

This means that operational risk is not at all standardized within the industry and financial institutions are unable to benchmark their models and develop best practice. The idiosyncratic use of Units of Measure means there will be wide variation of the measures of operational risk reached by different organizations. The reliability of these models might be called into question. Financial institutions need to co-operate with regulators to increase standardization and benchmarking, as only supervisors view models from all institutions. Doing this will allow firms to decrease model risk.