To challenge or not to challenge that is the question. In a second-line risk function, challenging financial and mathematical models is not optional. The question then becomes how best to challenge.
Within financial services, accounting and regulatory models abound, think capital, impairments and reserving. These first-line models typically require considerable investments in data, systems and highly skilled people. Second-line functions, including internal audit, model risk management and group risk functions, are then faced with the burden of how best to review, monitor and challenge these first-line models with a fraction of the resources.
One option is the use of challenger models. In the recent model risk supervisory statement SS1/23, the UK Prudential Regulatory Authority (PRA) notes,
Challenger models provide an independent benchmark against which to measure the performance of first-line champion models. Especially when faced with a universe of models, challenger models enable second-line risk functions to parallel run independent estimates. Where the difference between champion and challenger models increases beyond acceptable limits, the champion model can be flagged for detailed review. Another benefit of challenger models is that they can be used for continuous monthly rather than once-off reviews, minimising the time from model drift to detection. Advanced challenger models also provide graphical insights into emerging trends, providing lead indicators for emerging risks.
Given the capacity constraints facing the second-line risk function, talk to Analytix Engine about how we can help provide independent challenger models to support your business. We have a suite of model building software, including credit scoring and IFRS 9 models, available for on-premise deployment. Further, we are happy to meet your bespoke requirements and have a large team of data scientists available to augment your capacity when necessary.