The introduction of the EMV or chip card standard for credit cards is being touted as the solution to the huge card fraud problem in the United States.
But it's important to realize that EMV may result, at least in the short-term, in more fraud involving card-not-present transactions (i.e., transactions completed over the phone or online).
In the UK, EMV was rolled out in the early 2000s, and led to a clear increase in card-not-present fraud, even as card-present and overall credit card fraud decreased.
We believe fraud patterns in the UK in the 2000s are an important indicator of what we should expect to see in the U.S., as it adopts the chip card standard.
Here's what we saw in the UK:
- In the UK, the volume of card-not-present fraud losses overtook card-present fraud sometime around 2006, precisely when EMV card penetration reached about 100%.
- Total card fraud actually grew from £512 million in 2006 to £692 million in 2008 in part due to the growth in card-not-present fraud, but also due to fraud perpetrated abroad, and expanded credit card transaction volume overall.
In other words, the new chip card standard did wonders for reducing card present fraud — which occurs when payment cards are authenticated at a physical payment terminal. It also reduced total fraud. But the difficulty of committing successful card-present fraud clearly drove criminals to card-not-present fraud (see chart, below).
Click here to download the charts and data in Excel.
To make sure we captured the effect of chip card implementation accurately, we constructed an indicator that controls for both inflation and the growth in credit card transactions in the U.K. during the period in which chips card were implemented. Mainly, we wanted to make sure that the rise in card-not-present fraud was independent of the overall rise in credit card volume.
The raw data comes from Financial Fraud Action UK and the Bank For International Settlements. The indicator captures the fraud rate in British pounds per thousand card transactions.
Here are the key data points:
- Our card-present fraud Indicator dropped from £83 per thousand card transactions in 2002, to £24 in 2009.
- But the card-not-present fraud indicator grows from £29 in 2002 to a peak of £49 in 2008, before dropping to £36 in 2009.
- From initial EMV implementation in 2004 to the point at which it neared full implementation in 2006, the card-present indicator crashed from £75 to £38.
This suggests that e-commerce merchants should expect to see rising fraud costs as the U.S. moves to the chip card standard.
That said, the growth in card-not-present fraud in the U.S. probably won't be as strong as it was in the U.K. because the card networks have improved their algorithms for detecting card-not-present fraud. That's partly why card-not-present fraud eventually decreased in the U.K. after peaking in 2008.
EMV cards carry a "chip," which is an embedded microprocessor. Chip cards are difficult to copy, unlike stripe cards that can be cloned. In addition, in the U.K., card users are required to enter a PIN to complete a transaction, which makes it even more difficult for criminals to use stolen cards for card-present transactions.
