We were engaged by a regional bank to perform a model validation of their AML transaction monitoring and sanctions screening system, which was used by the Bank to monitor transactions for unusual or suspicious activity and for possible sanctions matches.
As part of this validation, we assessed the Bank’s AML and sanctions model design, the conceptual soundness of the model, alignment of the model to the Bank’s AML and sanctions risks, the configuration of the model components and model governance. We also performed a validation of the data imported into the model for accuracy and completeness, performed functional testing of the model’s suspicious activity detection rules and scenarios, sanctions screening configuration and analyzed the parameters and thresholds used to verify if they were set appropriately to maintain model effectiveness and efficiency. Additionally, we analyzed model output through a review of cases and alerts for evidence of reasonable review, annotations and timely disposition.
Our validation resulted in several key recommendations that assisted the bank enhance key controls related to model governance, model configuration and enabled the bank to improve model effectiveness and efficiency.
Note: This work was performed prior to joining Stout.