Model Validations

Model Validations

Model Validations

Suspicious activity monitoring systems empower institutions to monitor their accounts’ transactional behavior systemically. Anti-money laundering (AML) models rely on skilled judgment coupled with sophisticated scoring algorithms and have other quantitatively derived components. Examiners recommend in the FFIEC AML Examination Manual that the rules and parameters of your AML monitoring programs should be periodically and independently reviewed to ensure your program addresses the risk at your institution.

• With the constant evolution of new money laundering trends along with new regulatory requirements, it is critical to re-evaluate the effectiveness of your anti-money laundering programs on a regular basis. And, as public and regulatory expectations increase, it is even more important that BSA/AML professionals can rely on the systems they use to identify suspicious activity.
• Not all alerts/cases will result in a suspicious activity report filing, so it becomes difficult to evaluate or tune the effectiveness of the rules and scenarios based solely on the alert to SAR ratio. A thorough scenario tuning process, coupled with strong subject-matter expertise, will ensure your institution’s monitoring systems are effective and sustainable by combining both analytics and an expert judgment approach.
• Validation involves a degree of independence from model development and use. Arc-Serv is an independent consulting firm with consultants who have over 35 years’ experience in the industry. Arc-Serv validations have successfully been through regulatory scrutiny.
• Effective validation helps ensure that models are sound. It also identifies potential limitations and assumptions and assesses their possible impact. Arc-Serv will review to ensure the parameters of your rules in your model are successful in identifying suspicious activity without creating excessive noise.
• The rigor and sophistication of validation should be commensurate with the bank’s overall use of models, the complexity and materiality of its models, and the size and complexity of the bank’s operations.
• BSA/AML Model Validation includes various components. Most model validations can be broken into three main projects (although institutions may customize the project to fit their needs).
  • The three main project components are:
  • Program Evaluation
  • Transaction Testing (testing of the parameter algorithms)
  • Data Integrity Review

Depending on the size, risk, and needs of an institution, they may request Arc-Serv to perform any of the various components. The validation teams at Arc-Serv will identify weaknesses and rigorously support conclusions regarding the acceptability of model results.