Case Studies for Fraud Risk Analysis

Purchase Limit Management
The credit card industry long ago discovered the benefits of managing purchase limits (credit lines) at the cardholder level. The same approach can be applied to debit card portfolios – resulting in material impacts on contingent risk and fraud losses. (click here to see example)

• Cardholder-Level Limits: This analysis will quantitatively derive appropriate purchase limits at the cardholder level and calculate the reduction in risk at the portfolio level.
• Adaptive Control: Borrowing from the successful credit card practice, debit purchase limits are periodically reviewed based on cardholder lifecycle and seasonality. Through continuous purchase limit adjustment, risk is managed and incremental purchase volume is enabled.

Mass Data Compromise
Build a framework for data-driven reissuance decisions in the case of a mass data compromise. (click here to see example)

• Behavioral: Segments of a debit portfolio have different levels of sensitivity to a fraud reissuance. This analysis looks at pre- and post-behavior and helps determine how portfolio segments react.
• Profitability: This analysis looks at all the hard and soft costs associated with a large card reissuance. Combined with the cardholder behavioral change, we can determine the net profitability of such a portfolio action and provide a framework for future reissuance decisions.

Debit Fraud Detection
Data can be leveraged to analyze fraud and build new detection strategies.

• Fraud Score Evaluation: By crossing confirmed fraud with scored transactions, we can calculate how effectively your neural network is predicting fraud, and how efficiently your case cut-off score is set, based on your capture rate and false-positive ratios. This will allow for proper oversight of the service for management and audit purposes.(click here to see example)
• Data-Driven Fraud Detection Rule Development: By analyzing your data and adding in the impact of internal and/or VAA fraud scores, we can help you generate fraud detection rules. Our quantitative approach is a mathematical process for rule development and measuring the effectiveness of deployed rules before and after implementation. (click here to see example)
• New Fraud Service Implementation Support: As new fraud services become available, quantitative support is vital to business case development and implementation.