Model Risk Management was outlined in 2011 by the Federal Reserve and the Office of the Comptroller in the Supervisory Guidance on Model Risk Management (SR 11-7). This regulatory standard remains the standard guideline for MRM today.
While MRM is a regulatory requirement, its value extends far beyond compliance. The consequences of incorrect model usage or defective models can be detrimental to banks and their clients. Financial losses in the billions have occurred as a result.
Effective model risk management starts with building a solid governance framework that outlines how model risks will be managed throughout the model development and validation processes.
Model Risk Governance
Each firm needs to have an MRM framework in place – set of policies, procedures, and defined roles that formalize MRM implementation. A vigorous model governance framework should provide support and structure for model risk management.
This framework should be governed by the board of directors and senior management to assess the effectiveness and ensure that risk is within their level of tolerance.
Model Development
Model risk management begins with sound model development and implementation that are consistent with the goals of the end user and bank policy. Each piece of the development process must be documented thoroughly, including the statement of purpose, design, theory, logic, and methodologies.
Prior to implementation, data must undergo quality assessments, and the model components must be tested for accuracy, stability, limitations, and behavior across varying inputs.
Model Validation
Model validation should verify that models are performing as expected, in line with their objectives and uses. It should also identify potential limitations and their potential impact. Validation should be completed by people who are not involved with the development of the model.
Initial validation is completed before a model is implemented. During this process, validators should verify data and ensure there are controls in place to manage issues with data quality. The model should go through rigorous testing and benchmarking to check performance over a range of inputs and observe the effect of changing assumptions.
A key component of validation is the ongoing monitoring of models once they are implemented. Ongoing monitoring ensures that a model is performing as intended. By continuously tracking models, you can evaluate whether different activities and exposures require model changes or redevelopments.
Conclusion
Model Risk Management is a complex, time-consuming function that is critical to preventing severe consequences of model misuse and malfunction. At Evalueserve, we have experts who have decades of experience in risk and top-of-the-line technology that can elevate your MRM function.
Talk to an expert about how we can support your MRM teams.