Real-time Fraud Prevention

Payments Leader

Posted on January 7, 2014

Financial fraud is a pervasive problem that payment card issuers go to great lengths and great expense to minimize. Fraud prevention has become an industry in and of itself, which isn’t surprising when you realize that annual losses attributed to card fraud total $8.6 billion. In this environment, finding effective fraud prevention solutions isn’t a luxury; it’s a necessity.

Like many things in business, success often comes down to the tools you’re using and fraud prevention is no different. The better your anti-fraud strategies and solutions, the better you’ll be at protecting your institution and your cardholders. These pressures are driving the industry toward more and more real-time solutions that attempt to stop fraud before it starts. Generally speaking, the myriad of ever-evolving real-time fraud prevention tools fall into one of three major categories:

  • Consumer-enabled controls. An upcoming trend is providing consumers with specialized smartphone-enabled technology tools that allow them to set their controls and alerts based on their preferences rather than the FI’s. These preference types might include specific merchants, transaction types, and locations. Additionally, consumers would even be empowered to set their own threshold levels. With these card-enabled technologies, consumers could set limits as to when and where cards can be used and received interactive alerts when these restraints are breached. Ideally, a set of common functions could be piggybacked on the anti-fraud technologies, allowing users to make balance inquiries, transfer funds determine when and where their payment cards are used. As a company interested in providing advanced anti-fraud tools and technologies to our customers, FIS Global is entering into a partnership with OnDot Systems. A Silicon Valley technology company at the forefront of developing value-added mobile card service platforms, OnDot helps deliver user preferences and real-time communication to payment processing. Expect to hear more about this initiative in the months ahead.
  • FI-enabled controls. Stopping flash fraud is a 24/7 priority because by its very nature is unexpected and therefore unpredictable, it strikes at any time. Flash fraud can even hit an FI’s most precious resource, its own portfolio. Effectively combatting flash fraud only happens when real-time authorization rules are invoked through the use of flash fraud rules. Flash fraud rules provide FIs with defined parameters that are highly effective in catching and blocking transactions that become suspect because they fall into pre-determined risk categories. Real-time is highly effective in reducing flash fraud losses because the rules work when flash fraud strikes. It is a proactive strategy rather than a reactive strategy.
  • Real-time neural network protection. This tool relies on processing collected historical transaction-related data. A neural network is patterned after how the human brain learns behaviors and processes information. Within the historical transaction data, the FI flags those transactions that were confirmed instances of fraudulent activities. Data modelers then take this volume of information and ‘train’ the neural network to recognize the differences between normal, authorized transactions, and potentially fraudulent transactions. As new transaction data is passed through the neural network in real-time, it is compared to the historical data. Based on what it ‘knows,’ the network scores the transaction, with a high-scoring meaning that is likely a fraudulent transaction and declines it in real-time. While not perfect, a neural network is the most efficient and effective real-time antifraud tool available today. It is particularly well suited for companies with large volumes of historical data; in fact, the more data, the better it is able to ‘learn.’ It is the neural network’s ability to issue a ‘real-time’ decline that is an effective deterrent against fraud. We know that the longer it takes for FIs and consumers to recognize when fraud takes place, the more costly it becomes.

For customers, the anti-fraud experience goes beyond their costs. With the FI controls discussed above, there is always the threat of interruption. This can come in two forms: the interruption of daily activities to deal with an incoming call, text or email fraud alert; and the potential disruption of the shopping/buying activity that accompanies a hard decline of an authorization request. Consumers also sometimes suffer the pain and humiliation of false positives. False positives indicate the number of transactions that are flagged as fraudulent when they are not. False positives can be triggered by too many transactions occurring within a defined geographic area, transactions accruing in defined ‘high fraud risk’ areas or transactions happening in different geographical locations, such as someone flying from New York to Denver, let’s say, but who has a layover in Chicago and decides to buy something there.

Neural networks are particularly adept at helping significantly reduce the ratio of false positives. Of course, fraud strategies and models must be refined and ‘taught’ to identify the latest fraud patterns as they spread across geographical markets, transaction types, card products, and cardholder segments.

Switching to real-time fraud prevention solutions can vastly reduce annual losses attributed to card fraud. Now is the time to start implementation these solutions, not only for your best interest but the interest of your card holders.

Leave a Reply

Payments Leader

Payments Leader from FIS provides insights on credit, loyalty, fraud and emerging payments strategies through blog posts from our industry experienced authors.