Barclays Bank's head of fraud prevention has identified data mining as the key element in the bank's battle against organised fraudsters.
Fraud-busters can apply similar techniques to those used by supermarkets in tracking customer buying patterns, according to Barclays' fraud prevention chief Tony Thomas.
"The key to fraud prevention is to put data in one central point, such as a data warehouse," he said. "Gone are the days when we looked through documents all the time. Now, nine times out of 10 the initial referral for fraud will be as a result of something that we have found in our data."
Speaking at this week's SAS E-Intelligence Fraud Detection and Prevention forum, Thomas said organised crime poses a significant threat to the banking sector. "Criminals can have an intimate knowledge of processes. We have caught people who have worked for the bank. Fortunately, our systems and audit trails are very good," he said.
Fraud costs UK banks an estimated £2bn a year. Figures from the British Bankers Association suggest that counterfeit fraud alone cost £101m in the first six months of last year.
According to Thomas, banks need to focus on the information they already have. In addition to internal and external data matching, this includes credit referencing and fraud scorecards. He also urged closer links with government departments and suggested using workflow techniques to get the most out of systems.