Background
A mortgage lender uses a combination of an internal fraud score, business rules and 3rd party alerts to rank every loan application they receive by fraud risk. Their Fraud Exposure has has been steadily increasing and now averages $9MM per month. They are also experiencing a decline in their Fraud Prevention Rate, which is down to 65% from 74%, resulting in annualized fraud losses of $38MM!
The Process
The lender securely transfers an application-level data file with no PII data included. The data is processed through the Optimization Engine and within seconds results are provided showing fraud detection performance for every 50 point segment for the internal fraud score, each of the 30+ business rules, each of the 5 vendor alerts, as well as fraud score & business rule combinations, fraud score & vendor alert combinations, and business rule & vendor alert combinations. The results also provide incremental performance, ranking all indicators based on incremental fraud detection and clearly identifying the best indicators and those that do not provide any incremental value. Reports and output data are delivered to the lender who determines the changes to make to their fraud targeting strategy based on business objectives.