Stout was engaged by a financial institution to perform a model validation of the institution’s customer risk rating model used to monitor and assess the levels of risks that are associated with its customers.

When assessing the bank’s risk rating model, we focused on the model’s conceptual soundness, data validation, system functional testing via model replication, process verification, and outcome analysis.

We assessed the conceptual design by reviewing the bank’s most recent AML customer risk assessment to validate that the system was appropriately designed, reviewing model documentation, and assessing whether the model appropriately considers various customer risk types.

Additionally, we performed control testing for data inputs to validate accuracy and completeness and analyzed data quality of key data elements.

As part of the system functional testing, we tested the risk scoring methodology to confirm it was appropriate and customer risk scores and risk categorizations were accurate.

We also performed an outcome analysis to verify the outputs from the system were in line with the bank’s objectives.

Through our validation, we were able to assist the bank by highlighting key findings and improvement opportunities; thus, the bank can enhance their controls for the customer risk rating model to enhance model effectiveness.