ID Analytics Pokes a Hole in Address Fraud Fight

Bank Technology News | December, 2009

Banks and data firms that handle address changes, new account openings and billings may be fighting the threat of address change fraud with one hand tied behind their backs. A new study from ID Analytics suggests the traditional variables used to gauge the ID theft risk connected to real and fake address changes are not as predictive as once thought.

Using data from its ID Network (which includes 700 billion total aggregated data elements, 2.6 million reported identity frauds, and 1.4 billion consumer transactions,) ID Analytics found that conventional measures to distinguish authentic address changes from a fraud attempt—such as distance of reported moves, comparative area income levels and geographic move patterns—do not provide comprehensive ID risk coverage, and extensive and automated scoring methods have proven more effective.

For example, high to low income moves (more expensive to less expensive address), do not necessarily stand out as a particular target for ID thieves. That debunks long-held conventional wisdom, given the expensive-to-less-expensive move’s perceived susceptibility to “address takeover” fraud, in which a crook falsely reports an address change to gain access to a consumer’s bank account. “High to low is a big fraud risk, but even the inverse has risk,” says Mike Cook, chief strategy officer, ID Analytics. “It doesn’t’ really matter where you go from an income perspective.”

ID Analytics contends that using distance of a reported move does not provide comprehensive fraud protection. The study found that while distance is more predictive than income comparisons when identifying address change fraud for reported moves of more than 100 miles, more than 95 percent of reported consumer address changes occur between addresses that are less than 50 miles apart—making the use of distance as a tool to distinguish real from fradulent address changes less effective. Additionally, the study suggests network-based ID risk scoring is predictive at identifying the risk of a reported address change, with more than 30 percent of fraud found in the top one percent of identity theft risk scores.