Organizations Are Getting Too Little Value From Analytics
“Through 2022, only 20% of analytic insights will deliver business outcomes.” -Gartner  Organizations are getting too little value from analytics. Billions are being…
FI was tasked with validating a large financial institution’s stress models.
Client: Large Financial Institution with $130+ Billion Dollars in Assets
Challenge: A large bank contracted FI Consulting to support its Model Risk Management Group by performing baseline validations of three of their Probability of Default (PD) and Loss Given Default (LGD) CCAR (Comprehensive Capital Analysis and Review) Stress Testing models. These models forecast the expected loss for the client’s mortgage and HELOC portfolios within the Residential Real Estate (RRE) lines of business as part of the capital planning process. Key deliverables of the project included three comprehensive validation reports and an accompanying presentation of the validation approach, techniques, and key findings.
FI Consulting provided expert advisory support to the Model Risk Management Group in the areas of model theory, data quality, testing, implementation, and regulatory compliance. Model Validation activities were consistent with regulatory guidance, internal validation policies, and industry best practice. Throughout the validation process, FI Consulting met with model developers, conducted effective challenge, and performed full replication and testing of all datasets and model components. FI Consulting’s efficient discovery of model findings and significant issues allowed the Credit Policy Committee to remediate issues and implement changes in production in time for CCAR reporting season.
FI Solution: FI conducted code review and systemically replicated key processing components. FI performed sensitivity analysis and bootstrapping using R, and tested goodness-of-fit, parameter stability, and discriminatory power. Non-linear machine learning techniques (Gradient Boosting, Random Forests) were used to validate the variable selection process.
FI Impact: FI Consulting helped identify and mitigate risks associated with the CCAR models in the areas of methodology and design, data integrity, deployment and monitoring, and regulatory compliance. This will help to ensure the reliability of RRE mortgage and HELOC loss estimates and prevent the Fed from issuing Matters Requiring Attention (MRA) related to the models. Specific recommendations supported by independent analysis assisted stakeholders in remediating model deficiencies. Documentation of the Model Validation process also served to increase model performance and mitigate model error. FI Consulting validated models over the span of 3 months, giving the Credit Policy Committee ample time to remediate issues and submit their capital plan by the January CCAR deadline.