About this paper
Customer due diligence begins with verifying each customer’s identity and assessing the associated risk. Assessing customer risk is an essential component of a comprehensive Bank Secrecy Act/Anti-Money Laundering (BSA/AML) monitoring program.
To meet risk governance regulatory expectations and accurately assess higher-risk customers, financial institutions are modernizing their customer risk rating models and moving their heuristic, rule-based customer risk rating models to statistical models, specifically ordinal logistic regression models.
These statistical models perform better than rules-based models, are easier to justify to the regulators and are easier to update, validate and maintain because they use an established and understood framework. They are quickly becoming standard due to the regulatory pressure to use more scientific approaches.