In recent years, the use of complex models has grown exponentially in the financial services industry. With this increased usage comes a greater need for oversight and risk management. To ensure that banks and other institutions are properly managing their model risk, the Federal Reserve has developed a supervisory framework to assess model risk management practices. Let’s take a closer look at what this framework entails.
The Supervisory Framework Overview
The Federal Reserve’s supervisory framework is designed to assess three key aspects of model risk management: governance, development and validation activities, and ongoing monitoring processes. This includes assessing whether a firm has defined roles and responsibilities for model-related activities; established policies, procedures, and systems; documented evidence of the integrity of models; and proper reporting mechanisms to monitor the performance of models over time.
With regards to governance, firms should have clear lines of accountability between those responsible for developing models and those overseeing their performance. The board of directors or senior management should also be involved in setting expectations for model quality as well as providing adequate resources to develop, validate, maintain, and monitor models. Additionally, firms should have clear policies for how data used in models should be collected and managed over time.
In terms of development and validation activities, firms must be able to demonstrate that they have thoroughly tested their models prior to deployment. This includes conducting backtesting or “out-of-sample” tests with data not used during model development; assessing potential biases; understanding any limitations of the underlying assumptions; regularly validating existing models against new data sets; simulating forecasts using different scenarios; monitoring performance metrics such as accuracy rates; documenting any changes made to the model over time; and ensuring that only qualified personnel are developing or maintaining the models.
Finally, firms should also have processes in place for ongoing monitoring of their models’ performance over time. This includes regular reviews by internal teams as well as external auditors or consultants if needed. Any issues identified through these reviews should be addressed promptly with appropriate corrective actions taken if necessary. Additionally, firms should periodically review their assumptions around regulatory capital requirements based on changes in economic conditions or business strategies that may affect their exposure levels going forward.
Properly managing model risk is essential for all organizations operating within the financial services industry today—and it’s critical that they adhere to the Federal Reserve’s supervisory framework when doing so. By following these guidelines on governance structure and controls; development, validation testing methods; and ongoing monitoring processes—firms can ensure they are properly managing their risks while meeting all applicable regulatory requirements along the way.
Looking to manage your model risk? Check out Connected Risk’s model risk management module today.