Model Validations

Model Validations
Model Validations


Under SR11-7 / OCC 2011-12, AB2013-07, OSFI 23 and other regulatory guidance, independent model validations play a key role in providing assurance that model risk is properly understood and well-managed. Models must be independently validated, and the validations must meet requirements in a wide variety of areas including input data quality, conceptual soundness, process verification, and outcomes analysis. The validation work must be completed to high standards.


Protiviti’s Model Risk team performs model validations for large and small institutions, ranging from the largest global investment banks to midsized commercial banks and Federal Home Loan banks. We have validation models for retail, commercial, trading desk, risk management, and other lines of business.  


We consistently receive high ratings from our model validation clients. Clients know they can rely upon the quality of our work, and choose us year after year.

Our areas of expertise include:

Stress Testing & Forecasting Models

In recent years stress testing requirements have led to development of large number of models. Proper validation is crucial, and regulators carefully review not only the models but also the validations. Protiviti has validated hundreds of stress testing / forecasting models, often meeting tight timelines. Our stress testing validations have been well-received.

Credit Risk 

Scorecards are used not only for retail exposures, but also for C&I and other product lines. We have extensive expertise in a wide variety of modelling methodologies used in scorecard models. Use of such models is rapidly widening beyond traditional areas to include e.g. online transactions. Each year we validate increasing number of these models for wider variety of uses.

Allowance Models

Allowance Models have been in use longer than other model types. There are well-developed modelling methods in common use. Protiviti has validated hundreds of allowance models spanning a wide variety of retail and wholesale exposure types. Our approach spans all areas required to be validated under regulatory guidances.

Operational Risk 

Advanced approach Operational Risk models are particularly complex and require considerable statistical sophistication. Protiviti’s Models team has validated a large number of Operational Risk models, most of which have been for use in Stress Testing and Basel programs. Such models must be carefully tested for stability and performance – our validations include special testing to address these aspects and other aspects mandated by Model Guidances.

AML Transaction Monitoring and Customer Risk Rating

Our approach to AML Customer Risk Scorecards (Model) validation is fully consistent with the Joint OCC and Federal Reserve Guidance on Model Risk (OCC 2011-12 / FRB SR 11-7), the FFIEC BSA/AML Examination Manual 2014-12, and also takes into account industry leading practices. We have extensive experience validating various types of Customer Risk Scorecards, including commercial products offered by AML industry leading vendors and Customer Risk Scorecards developed in-house by the bank.