Create a framework for validating financial models using MATLAB

Model validation is the iterative process used to verify and validate financial models to ensure that they meet their intended business use and perform within design expectations. To ensure transparency and independency, model validation is sometimes performed by a third party who neither develops nor uses the models.

Additionaly, model validation is one of the essential elements for sound model risk management. Risk managers need to work with users, developers, integrators, and other stakeholders to assure that the model works as intended without any error.  Common model validation activities include:

  • Creating and maintaining a model inventory
  • Performing quantitative and qualitative assessment
  • Backtesting the model
  • Documenting and reviewing data, results, and code implementation

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

See also: bank stress test, model risk, Basel III, Solvency II, IFRS 9, CECL

Financial Risk Management: Improving Model Governance with MATLAB