The Pandemic has created an environment of uncertainty to a level not seen by the industry, increasing the risks faced by financial institutions, including the risks associated with the use of models. For many institutions, models play an integral role in conducting day to day operations. Prominent examples of its use include evaluating risks and capital, defining funding requirements, understanding customer behaviors, managing data analytics, and making investment decisions.
The traditional guardrails to manage model risk may not be adequate given the unprecedented level of connectivity and nonlinear outcomes owing to what is described commonly as “fat tails” or statistically remote events. Models built on data from past recessions with known outcomes are inadequate for modeling pandemic tail risk, which has become reality in the current unknown environment.
Regulatory focus on model risk will likely be increased given the important role models play in day to day bank operations across all three lines of defense and the increased inherent risk due to the pandemic.
Sound model risk management involves having strong practices at various levels of the organization. Truly managing model risk relies on applying proper judgment and appropriately evaluating qualitative information, in addition to employing proper quantitative expertise. SR 11-7, interagency US federal agency guidance issued in 2011, incorporating lessons from the financial crisis, discusses several elements for model risk management. These elements are intended to fall into the same framework for managing traditional risks such as credit, liquidity, operations, and others. Model risk management should be viewed as a process, not an event, utilizing the following key risk management practices consistent with managing other traditional risks:
• Risk Identification, Ongoing Monitoring, Assessment
The rapidly changing environment of the pandemic has created a strain in the process of model development and independent validation functions. Discussed below are selected challenges institutions may face in its model risk management process and mitigants to address some of these challenges, leveraging its traditional infrastructure for managing all risks on an enterprise wide level. Also attached is an Appendix with a brief list of FAQs regarding models.