Apply the SR 11-7 guidance for model risk management
SR 11-7 is a supervisory guidance for model risk management (MRM) published by the Federal Reserve (Fed) and the Office of the Comptroller of the Currency (OCC). In 2011, the Fed and the OCC jointly published SR 11-7 as a supervisory guidance for model risk management (MRM). The “principles-based” guidance articulates the elements of a sound program for effective management of risks that arise when using quantitative models in bank decision-making.
SR 11-7 applies to national banks, state banks, bank holding companies, and all other institutions for which the Fed or OCC is the primary supervisor.
Recognizing the importance of model risk and its impact on financial systems, other countries have published their local guidance similar to SR 11-7, including:
- SS 3/18 by the UK PRA in April 2018
- TRIM by the ECB in February 2017
- E-23 by the Canadian OSFI in September 2017
The SR 11-7 guidance covers the entire model lifecycle including model development, implementation, and use; model validation; and model governance, policies, and controls. Figure 1 shows different components of the model lifecycle with a centralized model inventory and the stakeholders aligned with each component.
Since SR 11-7 was published, banks have invested significant time and capital implementing the guidance and still face several challenges, including:
- Handling data and documentation
- Reproducing, interpreting, and validating model results
- Accelerating and scaling implementation
- Monitoring model performance in real time
The challenges related to SR 11-7 are described in the white paper Effective Model Risk Management. The paper also highlights how transitioning from a waterfall-based model lifecycle to an agile model lifecycle reduces cost and facilitates compliance, as shown in Figure 2.
Considering the challenges and global regulatory practices—including SR 11-7, EU TRIM, OSFI E-23, SS 3/18, and industry best practices on MRM—MathWorks developed the MATLAB® Model Risk Management solution (MATLAB MRM) to develop, validate, deploy, monitor, and manage models quickly and efficiently.
MathWorks implemented MATLAB MRM at HSBC’s Group Risk Analytics division for access by users globally, while helping the bank in transitioning from waterfall-based MRM to agile MRM, as shown in Figure 3.
Ray O’Brien, HSBC Global COO, shared his vision of agile MRM and how HSBC uses MATLAB to adapt to strategic and regulatory changes: Financial Risk Management and Model-Based Design (1:30).
Because of its efforts, MathWorks was named a category leader in the Chartis RiskTech Quadrant for OpRisk solutions in 2021, as shown in Figure 4.