Model risk is the risk of loss or poor decisions arising from errors in a model’s design, data or implementation, or from using a sound model outside the conditions it was built for. Banks and financial market infrastructures use models to price instruments, set margin through an initial margin model, measure risks such as value at risk, score credit and detect fraud, so a single flaw can spread into losses or too little collateral.
In the United States, the Federal Reserve and the Office of the Comptroller of the Currency issued joint supervisory guidance in 2011, known as SR 11-7, that treats model risk as something to be managed through sound development, independent validation and governance. In April 2026 the Federal Reserve, the OCC and the FDIC replaced it with revised, more risk-based interagency guidance, issued by the Federal Reserve as SR 26-2. The practice is described under model risk management (MRM). Typical controls include backtesting against actual outcomes, sensitivity analysis and comparison with other models. The PFMI also expect clearing houses to validate their margin models and stress tests regularly.