Plastic Card Fraud Detection

Speakers

Registration

Abstracts

The fight against fraud is a constantly-evolving activity across all facets of the financial services industry.
As new fraud prevention technologies are developed and implemented, so the fraudsters will simultaneously be developing ever more sophisticated methods of gaining access to other peoples' money.
This presentation addresses the fundamentals of fraud detection in the payment cards arena:

• the current payment cards landscape
• the underlying risks giving rise to the need for fraud detection solutions
• the operational considerations of implementing and successfully running a detection solution
• the fit of detection solutions in the wider security infrastructure
• an insight into what the future holds for fraud detection

Credit Card Transaction Fraud: Modelling Methodology for Small/Medium European Issuers and Processors

In recent years, the plastic card industry has invested heavily in fraud prevention measures. Chip and PIN technology has had a demonstrable effect on the incidence of credit card fraud in the countries where it has been introduced. However, while Lost/Stolen and Counterfeit fraud have decreased significantly, fraudsters are versatile and technologically advanced and have simply moved their focus to other channels, for example, where Chip and PIN provide no protection - mail order, telephone order, internet transactions: Card Not Present in general. Given that complete fraud prevention is, currently, not feasible there remains a need for detection systems. In the US, the huge credit card market generally requires and expects sophisticated systems for scoring large volumes of transactions in real time. However, smaller European issuers and card processors have a need for solutions that are no less effective but smaller in scale and less expensive. We present a methodology for building predictive transaction fraud models that attempts to fit this requirement, combining robust regression techniques with next generation neural network technology. The talk covers areas from Sample Requirements and Data preparation to Monitoring and Optimisation and suggests appropriate performance measures.

Plastic Card Fraud Detection using Peer Group Analysis

Peer group analysis is an unsupervised method for monitoring behaviour over time. For each credit card account a 'Peer Group' of accounts is determined; These are accounts that exhibit similar behaviour. As time evolves, it is assumed the behaviour of an account is tracked by those accounts in its Peer Group. An account whose subsequent behaviour deviates strongly from its Peer Group is considered to have behaved anomalously and is flagged as potentially fraudulent. Since it is unlikely that an account will be tracked indefinitely by its Peer Group, we investigate how tightly accounts are tracked by peer groups. The real-time practical issues of both summarizing the behaviour of an account and comparing accounts are also described.