We're within five months now of the initial EMV (chip) card liability shift for POS transactions. Most people in the industry have held the belief that as the ability to create counterfeit cards is shut down, the criminals will shift their focus primarily to the card-not-present (CNP) environment, where they can continue to use payment card data they take from the magnetic stripe or other data breaches. In fact, my colleagues and I have been broadcasting this message in our presentations and posts for quite some time. Our assessment, along with most other industry experts, was based on the statistics released by banking groups in major countries that had already gone through the EMV migration. The chart illustrates one view of their experiences. It seems to leave no doubt about what we can expect.

Chart_cnp_fraud_losses

But does it mean what we think it means? While the chart clearly shows an increase in the CNP channel in fraud losses, did the ratio of CNP fraud to overall sales increase? Unfortunately, definitive data is not readily available to provide that answer. Using some confidential sources and partial—but significant volumes of—payment data, we were able to determine that during the period from 2010 to 2013, as a percentage of overall sales, CNP fraud in Canada actually held relatively steady. But was that stability created due to the large increases in the recurring billing segment in the CNP environment, which has a relatively low rate of fraud? At this point, we just don't have data granular enough to tell us.

I don't think this means that there isn't a reason to be concerned about CNP fraud as the EMV migration in the United States continues. For one thing, the experience of others is no guarantee that we will experience the same. But perhaps the biggest reason for us not to relax about the issue is that, even if the levels hold flat through our migration, CNP fraud is still quite significant and has a major negative financial impact on merchants and issuers. The 2013 Federal Reserve Payments Study found that CNP fraud by volume is three times that of card-present fraud.

This situation also demonstrates the need to be able to collect detailed and accurate data on fraudulent payments activity. Fraud has been a real challenge in this country because of the large number of payments stakeholders that end up saddled with the loss. The Federal Reserve is interested in working with the industry to develop a process for collecting such information for the benefit of all.

Photo of David Lott By David Lott, a payments risk expert in the Retail Payments Risk Forum at the Atlanta Fed