FICO has used non-monetary variables in two recent online banking models for international banks, leading to significant added fraud detection.
Last summer, a $1.5 million bank/wire fraud case garnered quite a few news headlines. A bank account of a California-based escrow firm was hacked, and three payments totaling $1.5 million were wired to accounts in China and Russia. Only $432,215 was recovered, and the loss forced the escrow company to shut down.
Experts say that such online fraud attacks have been significantly spiking since 2012. Corporate bank account takeover fraud loss is estimated to cost $347 million to businesses in North America in 2014, and this loss is expected to grow to $371 million by 2015. While small to midsize businesses are taking the biggest hit, consumer banking customers could be next in the bull’s eye.
The impact of online fraud, as most banks know, is not just financial losses from the fraud itself. Banks must also deal with high reputation risks and erosion of customer trust, leading to revenue loss spillover in other products as well.
How can banks ensure secure transactions, while still providing customers the flexibility of 24×7 online banking at a click of a button? The growing sophistication of digital fraud threats has challenged traditional fraud management controls. Fortunately, retail banking offers many new opportunities for banks to assess consumer behavior and predict fraud.
To fight online fraud, banks should monitor not only financial behavior but also how their customers are navigating on their portal. The various behaviors prior to completing a financial transaction order are quite revealing. In fact, FICO online banking models use such non-monetary information to improve risk detection.
Essentially, it is difficult for fraudsters to re-create the typical non-monetary footprint of individual consumers, and our models can assess these behavior differences. Non-monetary information used in the models may include activities like balance inquiry, statement retrieval, downloads of account information, adding a new payee account (recipient), request for overdraft protection and adding additional user access.
FICO has used such non-monetary variables in two recent online banking models for international banks, leading to significant added fraud detection. For example, at 1% Account False Positive Percent (AFPP), the Account Detection Rate (ADR) increased by 10%. This performance boost can be
