Financial institutions (FIs) are increasingly using decision models across credit lifecycle processes such as origination, credit monitoring, portfolio management, collection and recovery.
The analytics used is getting increasingly complex, driven by exponential proliferation of data and computing ability. Consequently, models have become more complex and difficult to understand.
These models also expose the FIs to risks associated with decisions based on incorrect models, financial losses, capital liquidity shortage, and loss of customers. Such risks are collectively referred to 'model risk'.
We believe this is a major risk and needs to be proactively managed instead of reactively.
Though model usage has seen a sharp increase among FIs, the measurement of risk associated with their usage is largely overlooked. This can be attributed to the dearth of regulatory suasion, unlike in other risk areas.
Regulators in the developed markets have become cognizant of the costs associated with model risk and have therefore formulated guidelines around it, such as the US Federal Reserve's SR 11-7 and the European Banking Authority’s guidance note CP/2014/14.
However, regulators in the emerging economies are yet to do so despite model usage in their markets reaching the scale of the developed world.
That’s why we believe FIs in emerging markets should put in place a holistic model risk management framework for systemic mitigation.
In this report, we highlight the challenges ahead of model-driven organizations in emerging markets, underscore the missed opportunities, and present CRISIL Risk Solutions’ framework for alleviation through effective quantification of model risk.