Is Your Model Risk Rating Source Agnostic?
11/29/2021
Are vendor models riskier than internally developed models? The 2021 RMA MVC Vendor Survey wants to know and is already yielding interesting insights from our members.
Model risk managers may validate vendor models internally or externally with vendors, depending on the model impact and available resources.
According to the U.S. Federal Reserve Bank's SR 11-7 regulatory guidelines, all models, including vendor models, should be validated. How banks validate the models, and when they should reach out to third-party validators, may differ depending on the organization.
Some banks engage a third-party model validator when they view a vendor model to be riskier than other models. According to early results from the 2021 MVC survey, over 26% of respondents’ organizations consider vendor models to be inherently riskier, due to the limited information available on model development, data, processing, and implementation.
However, about 62% of the respondents said their organizations treat vendor models equally in risk, as their model risk rating is source agnostic. Many model risk managers determine model risk by the likelihood and impact of an outcome due to model failure or inaccuracy. The fact that models were developed by external sources is not a variable in their risk assessment.
What factors go into your organization’s model risk assessment? Come join the discussion December 1 and 2 at our inaugural Model Risk Management Summit.
Join our live session “Supervisory Roundtable on Model Risk Management - FDIC, OCC, and Federal Reserve” on December 2. David Palmer, lead supervisory financial analyst in the Division of Banking Supervision and Regulation at the Federal Reserve Board, Alireza Ebrahim, senior financial economist of the OCC, and Jitendra Rathod, senior quantitative risk specialist at FDIC, can answer your most pressing model risk management questions.