Model Risk Governance and Audit Preparation

Model Risk Governance Services^SM

Best-in-class methodologies and expertise create a sound approach to model risk governance provides organizations with support to meet the Federal Reserve Board (FRB) and Office of the Comptroller of the Currency (OCC) Supervisory Guidance on Model Risk Management (Guidance).

A strategic approach

Model Risk Management business review and audit preparation

Experian’s business consultants have established regulatory experience and best practices to provide clients with an independent, third-party review of their model risk management practices and to assist with audit preparation for compliance with the Guidance.

By conducting an in-depth review of client risk exposures, business activities, and the complexity and extent of model use for either consumer or commercial purposes, we provide a comprehensive gap analysis report and best-practice recommendations to meet regulatory compliance.  

The gap analysis report identifies discrepancies between current actions and regulatory requirements, such as:

• model documentation and specifications
• implementation standards and protocols
• consistency of model assumptions with utilizations
• value and usage of monitoring reports

The gap analysis is followed by an action plan, or “blueprint report which identifies and prioritizes actions for bridging gaps.

Model Risk Validation

Our consulting and modeling teams provide expert strategic direction to offer all required tests to meet model risk governance on several types of consumer and commercial models, including credit risk, econometric and pricing models.

Our approach to model validation combines Experian’s modeling expertise, data and regulatory understanding to provide four key model governance analytical services for:

• Back-testing
• Benchmarking
• Sensitivity analysis
• Stress testing

Experian’s model governance protocol includes four specific analyses that can be executed on both generic and custom credit risk models.

Back-testing and benchmarking

Experian^® will conduct multiple tests of your models to validate performance and accuracy. During this analysis, statistical metrics will be produced to assess model performance in relation to previous years’ performance and to an industry benchmark model.

Sensitivity analysis and stress testing

Experian will evaluate the impact of changes in inputs and parameter values on the model’s ability to rank order and ensure it falls within expected ranges for model stability given a set of unexpected changes to inputs.

The Guidance requires that both small changes and more extreme changes to input be tested to verify that the model is robust and stable. Experian will conduct a series of characteristic-level stress tests to the model to determine the sensitivity of the model’s performance to small changes in expected input values, as well as sensitivity to more extreme changes, such as changes in economically based conditions that may cause the model to become unstable or inaccurate.