Model Risk Management

Arrayo was engaged by a leading financial institution to enhance its Model Risk Management practices and ensure compliance with the Federal Reserve Board’s SR 11-7 guidelines.

Challenge

The primary objective of this project can be split into two parts. The first part of the project was to vet internal and third-party models for compliance with Federal Reserve Board SR 11-7 (“Supervisory Guidance on Model Risk Management”). This included performing tests to evaluate the model under normal and stressed market conditions. The second part of the engagement was to produce up-to-date documentation on a proprietary model suitable for reference by users and for review by internal audit.

Delivery

Arrayo’s team coordinated closely with model developers, the market risk function, and head office oversight to ensure seamless collaboration across all stakeholders. Arrayo’s expert led team also liaised effectively with support personnel from vendors to gather critical insights and data necessary for successful outcomes. As a result of these comprehensive efforts, Arrayo delivered two key outputs:

1. Comprehensive Model Reports: For each model, our team produced a detailed, structured report that meticulously assessed and documented key components, including model assumptions, limitations, and conceptual soundness. This report also included a thorough evaluation of ongoing performance, benchmarking results, and a retrospective analysis of outcomes through back testing. The accompanying documentation adhered strictly to internal model risk standards, with all findings organized in standardized templates for easy review and future reference. This robust documentation ensured that all stakeholders had a clear understanding of the model’s efficacy, limitations, and potential risks.
2. Detailed Configuration and Procedure Report: In addition to the model reports, our team delivered a professional report that included a comprehensive step-by-step explanation of the most critical procedures involved in model development, implementation, and evaluation. This report provided clarity by defining key terms that were previously implicit in the original documentation. Furthermore, Arrayo enhanced the report by adding an appendix with an exhaustive listing of configuration files, ensuring that every setting, parameter, and version was transparently recorded for future reference. To further enrich the documentation, screenshots were included to visually capture the configuration details and functionalities, providing a clearer understanding of the model’s operational mechanics and facilitating a smoother integration and troubleshooting processes.

Value Creation

The Model Risk Management project delivered immense value by:

• Documenting Historical Model Testing: We organized and documented all previous model testing, ensuring a clear historical record and aiding future decision-making
• Evaluating Model Risk: Through strategic testing, we evaluated model performance under both normal and stressed conditions, identifying potential vulnerabilities.
• Identifying Model Issues: Our work uncovered previously unknown limitations and bugs, leading to crucial resolutions that enhanced model accuracy and performance.
• Effective Communication: We effectively communicated testing results and model risks to senior management, ensuring that potential risks were properly addressed
• Audit-Ready Documentation: All final reports were fully audit-ready, providing valuable assets for regulatory inquiries, such as when the Federal Reserve requests model documentation or testing results.
• Ensuring Compliance: We supported the regulatory compliance process, ensuring all testing and reporting aligned with SR 11-7 guidelines.
• Improved Documentation: Enhanced documentation that clarifies key aspects of the models making it more user-friendly and accessible for a variety of internal stakeholders, including model users and auditors.
• Increased Readability: We improved document accessibility by adding tables of contents, enlarging figures, and highlighting key features, making complex details easier to understand.
• Precise Language: Replaced aspirational language (e.g., “the model will do…”) with precise, factual descriptions of the model’s actual capabilities (e.g., “the model does…”), ensuring that the documentation accurately reflects the current state of model implementation and avoids potential misunderstandings or misrepresentations.
• Enhanced Methodology Transparency: We strengthened explanations of the model’s methodology for commonly traded instruments, based on specific requests from internal audit, which helped provide a deeper understanding of how the model works and the rationale behind key assumptions.
• Smoother Audit Process: The improved documentation ensured smoother internal audits and aligned model implementation with both internal and external audit requirements.