Model risk is the potential for a loss when financial models for valuing financial instruments, measuring risks, or making business decisions are misused or inaccurate. Historically, model risk has played an important role in major financial losses; examples include the London whale, Long-Term Capital Management (LTCM), and the subprime crisis of 2008–2009.

Generally, hundreds to thousands of models are being used in financial institutions to manage their business. To mitigate model risk, risk teams need to perform various tasks, including:

• Model documentation
• Model validation and monitoring
• Scenario analysis and stress testing
• Benchmarking and challenging models with machine learning
• Model risk reporting

Popular tools include MATLAB^®, Statistics and Machine Learning Toolbox™, Risk Management Toolbox™, MATLAB Report Generator™, and MATLAB Production Server™.