Anti Fraud Tools Score at MasterCard, SAS

Bank Technology News  |  March, 2011

By John Adams

Print Email Reprints Feedback

In a pair of moves on the fraud prevention front, MasterCard’s developed a new service that attacks one of the most dangerous strains of card fraud via real-time scoring of account data compromise (ADC) exposure, while SAS won a patent for its fraud detection modeling.

At MasterCard, the Expert Monitoring Compromised Account Service will be available to all of MasterCard’s issuing banks as part of the card firm’s suite of fraud detection and prevention capabilities.  

Using MasterCard’s internal technology and data management techniques, the service scores the probability of fraud on ADC-exposed accounts by analyzing transaction information, merchants, confirmed fraud data and behavior of similar accounts. The score predicts the probability of fraud over the following one to two weeks, helping issuers determine when, or if, to close the exposed account and reissue a new account number to the cardholder.

ADC is considered a particularly damaging threat, exposing banks to fraud losses, re-issue expense and attrition. MasterCard’s internal research found that accounts exposed to an ADC are three times more prone to fraud than non-exposed accounts.

Spotting these events and determining the actual level of corresponding fraud risk is a beneficial pursuit—Javelin has found that cardholders who have to receive a new card because of an ADC event often use the card less or abandon it completely.

At SAS, the firm’s “Computer-Implemented Predictive Model Generation Systems and Methods,” which is part of SAS’s fraud management suite, received an approval from the U.S. Patent office.

SAS’s fraud management employs a self-organizing neural network arboretum (SONNA) modeling. The system is designed to improve fraud detection performance by creating risk-based factor groups within an institution, giving the financial institution’s fraud operations and analytics groups greater insight and context on the risks and reasons behind its predictive fraud scoring.