To the Editor:
Your June 1 article "
I applaud these organizations' willingness to explore new approaches to combat debit fraud, but I would caution that these efforts will best succeed when a solid data infrastructure is in place to analyze behavior and measure results at the account level.
We have conducted purchase limit analyses using millions of transactions from hundreds of financial institutions.
Our research reveals that the vast majority of debit-card holders use a small fraction of the daily purchase limit allotted to them, but fraudsters use much more. Therefore, the opportunity exists to reduce portfolio risk and loss through reduced daily purchase limits.
Our experience has shown that daily purchase limits can be materially reduced without cardholder impact, but only if purchase volume is analyzed and tracked at the individual cardholder level.
This involves trending debit purchase volume over time, aggregating purchases by day, and finding the peak day utilization for each cardholder.
Each account is then assigned a purchase limit based on historical behavior.
After daily purchase limits have been adjusted, cardholder behavior should be reviewed for changes in approval/decline rates, as well as changes in purchase behavior. There is some indication from our research that elements such as life cycle and seasonality will require additional adjustment to purchase limits over time.
In summary, debit purchase limit management can be considered roughly analogous to credit card credit limit management.
It is a data-driven process that has both revenue and risk/loss implications. Financial institutions and processors who successfully apply analytics to the exercise will be best able to quantify the impact of their approach and refine it over time.
Jeffrey M. Trachtman
Vice president, fraud risk analysis
Analytic Innovations LLC
Chicago
