The Federal Reserve recently announced a new FraudClassifier™ model designed to improve the consistency of fraud classification and create a consistent taxonomy, leading to effective quantification of fraud losses. The purpose of the model is to classify fraud independent of payment type, payment channel, and other payment characteristics. It implements a decision graph beginning with who initiated the payment to differentiate payments initiated by authorized versus unauthorized parties.

The result is a holistic view of fraudulent events, allowing for better management in the following areas:

• Helping financial institutions standardize classification and make better fraud management decisions.
• Enabling allocation of appropriate resources for various fraud characteristics.
• Providing consolidated reporting to the Federal Reserve for the creation of a macro view of industry fraud trends.

The new model has a multi-year roll-out plan determined by the Federal Reserve. ARGO’s fraud solution, OASIS™ (Optimized Assessment of Suspicious Items), will support the implementation of the FraudClassifier model by:

• Pre-populating information and using decision trees to enable easy classification.
• Providing enhanced reporting to institution management.
• Enabling fine tuning of machine learning decision engines to solve optimization problems based on this new classification.
• Automating reporting to the Federal Reserve.

ARGO OASIS customers get two classes of benefits: hard measured monetary, and customer relationship. OASIS provides value by more accurately detecting and preventing more fraud through multiple channels, at multiple points in time and contact with customers, through a single platform. More intelligent detection and machine decision-making allows for streamlined and accurate reporting, and improved strategic decisions by financial institutions with respect to the direction they take their fight against increasingly challenging fraud perpetrators.