Banks that can’t stay on top of credit card fraud risk losing customers as well as incurring reputational damage over and above the financial implications of being defrauded. Moreover, credit card fraud appears to be on the rise.

These are the views of Colin Hill, senior solutions manager for financial crimes and risk management at analytics company SAS.

Hill says in light of social media and an increasingly connected consumer base, banks that fall victim to giant frauds often find customers moving their business elsewhere.

“[Consumers] feel unprotected,” says Hill. “Just look at Postbank. I can’t imagine the damage [recent fraud there is] going to do to its business and reputation.”

Postbank was defrauded to the tune of R42m at the beginning of January.

It’s not only the theft of money that worries consumers, but the increasing prevalence of identity fraud. Hill says customers are beginning to question not only whether financial institutions are sufficiently equipped to protect their money but also their personal information. He says syndicates are increasingly attempting to steal identity information to use for later, often financial, crimes.

Hill says the primary method of detecting fraud is via data analysis. That is, looking at a customer’s banking habits and profile and detecting anomalies that might suggest something is awry.

However, one of the problems of this approach is that banks often have separate divisions dealing with different aspects of a customer’s finances and data can thus be disparate and disconnected. Hill says it’s crucial banks work to create a “single customer view” that encompasses all of these.

According to Hill, although this puts more of a customer’s data in one place — making it all the more essential that this data be protected — it’s necessary to build early-detection systems with sufficient information in them, particularly as fraudsters are themselves becoming more savvy in response to banks’ attempts to curtail fraud.

“Syndicates are hiring students to do quantitative analysis and build models of customer behaviour so as to be better able to sidestep the preventative measures,” says Hill.

The recent Postbank debacle is particularly worrying because the bank’s security systems should have been alerted the moment there was a repeat transaction on a card, he adds. “It’s astounding that the fraudsters were able to make so many withdrawals before the fraud was detected.”

He says, however, that fraudsters often plan these sorts of attacks months and even years in advance and there are often insiders involved making it more difficult to detect, particularly if they are familiar with the bank’s inner workings and how to sidestep detection methods.

A recent statement from the South African Banking Risk Information Centre (Sabric) claimed credit card fraud rose by over 50% in the past year and that the banking sector lost R403m to fraudulent transactions in 2011, up an enormous 53% from R263,8m in 2010.

“According to Sabric, the majority of this can be attributed to higher volumes of cards, card-not-present fraud and counterfeit cards,” says Hill.

The growing volume of transactions that banks are dealing with means a substantial increase in the available data they need to analyse in order to detect fraudulent transactions. Hill says data and quality analysis is invaluable in preventing fraud and that this is where the key to fraud prevention lies.

“Without the ability to perform real-time analytics on every single transaction and with turnaround times in the order of less than 30 milliseconds, organisations will not be able to reduce the magnitude of fraud and detect fraudulent transactions proactively,” he says. — Craig Wilson, TechCentral