Home Slider The Analytic Advantage: Stay One Step Ahead of Cyber-fraud

The Analytic Advantage: Stay One Step Ahead of Cyber-fraud

by internationalbanker

andrew-davies-2016By Andrew Davies, VP, Global Market Strategy, Financial Crime Risk Management, Fiserv





Commercial and retail customers of financial-services companies demand convenient access to their money through more and more channels and devices. They expect money movement to be immediate and secure. Also, the proliferation of non-traditional payment-service providers is eroding the revenues that traditional finance institutions (FIs) such as banks derive from the provision of payment services. Customer retention is a high priority for all FIs. Fraudsters, using social engineering for phishing and malicious software for cyberattacks, pose ever-increasing threats to FIs and their customers; the theft of money and assets through cyber-fraud is at an all-time high. The combination of all of these factors is greater than the sum of the individual parts and requires FIs to deploy defence mechanisms to ensure the security of their own and their customers’ assets. Now more than ever, FIs need to use technology to protect individual FIs and the integrity of the financial systems in which we are all invested.

Bearing all of the pressures in mind, the challenge that now faces FIs is how they can stay competitive while also meeting their customers’ demands for security in this brave, new world of faster and even instant payments. Of course, technology has developed to assist organisations with collecting and utilising data to offer the best service possible, yet this evolution in security technology has also resulted in increased sophistication of techniques and tools used by fraudsters. This is happening more frequently as criminals search for the paths of least resistance in technology that will result in the largest gains.

With stories about cyber-fraud hitting the headlines on a regular basis, FIs must be able to react rapidly to fraudulent attempts, and, in an ideal world, pre-empt them. Once an organisation’s security is breached, the stability of financial institutions and the global financial system is compromised. Following recent warnings from SWIFT, their member institutions have been cautioned and advised to demonstrate greater security awareness, as the spotlight is currently fixed firmly on high-value cybercrime.

The success of a financial institution if dealing with cyberattacks relies on it being able to pre-empt and prevent high-value fraud and to successfully take action against it. That way, customers experience the least amount of disruption to their banking experiences and continue to trust that institution with their money. Implementing profiling engines, predictive models and leveraging consortium data can all help to facilitate the high level of security required by financial institutions.

Leveraging the data.

Due to new payment infrastructures and services such as real-time payment settlement, traditional methods of identifying suspicious activity are no longer advanced enough to adequately protect institutions and more importantly their customers. Despite these advanced payment technologies streamlining financial processes, they also open the door for sophisticated fraudulent activity to take place—something that can cause a huge amount of damage to an organisation. One way to mitigate the threat of sophisticated fraud attempts is to deploy models to anticipate and recognise fraud before it happens.

In order to protect themselves better, FIs and corporations can leverage the value of data in order to pre-empt the attacks on their systems, and thus reduce the chance of money being fraudulently stolen. Predictive analytic models that are specifically designed to detect payment fraud can be deployed to ensure that attacks are recognised before money leaves the FI, meaning that ultimately fraud can be combatted before the loss. Reputation is a huge factor in customer retention, and stopping attacks before they happen will result in customers being far less inconvenienced and reduce the defection of customers due to security concerns.

In order to be able to get the most out of predictive fraud-prevention models, institutions need to understand customer behaviour and their typical transaction activities. Incorporating this data with previous fraud patterns and the outcome of past fraud investigations allows organisations to create a more detailed picture of customer habits. This also makes it easier to identify real-time anomalies when a transaction falls outside of what is deemed normal behaviour, and as a result the fraud analytics team has the opportunity to carry out any further investigations immediately. Transaction frequency, beneficiary history, time of day and size of payments all contribute to the overall picture of an individual’s behaviour, and collating them into real-time scoring, case management and industry-level predictive models allows issues caused by false positives to be mitigated.

Detecting fraud, and recognising fraudulent attempts, for electronic funds transfers (EFTs) across payment types including, for example SEPA, SWIFT and Faster Payments, is essential. FIs have a responsibility to identify fraud in a sea of legitimate transactions, while also adhering to the highest level of security for their customers. Ensuring that predictive models and banks of data are in place can set clients’ expectations and ultimately result in customer loyalty and retention.

The power of consortiums

Customer data is vital to the success of identifying fraud patterns, and the richer the pool of data, the more informed analytic models will become. Consortiums gather information on customers from a variety of payment organisations that are part of the conglomerate, which means that a detailed, near-complete picture of customer behaviour can be curated. Consequently unusual behaviour can be flagged and fraud can be detected earlier than would be possible without the consortium.

Pattern-recognition modelling can be implemented based on the consortium data, resulting in the most advanced protection for FIs and organisations. For the success of these consortiums to be realised, FIs should pool their data on both legitimate and fraudulent payment transactions to help determine fraud patterns. Predictive models are able to flag possible fraudulent transactions should characteristics of attempted transactions match up with ones that have already occurred, adding an additional layer of security to anomaly detection.

Although consortium data collects information from organisations on every aspect of their customers, modelling processes need to be successfully deployed to ensure that the relevant data is used for the correct analysis. Fraud indicators derived from investigations of commercial customers may differ from those of retail customers, and being able to align the right information from a wealth of data across all industries can enhance the level of detail and accuracy in fraud detection.

As the amount of non-cash payments continues to rise in the face of digital transformation, real-time fraud-prevention and detection solutions are vital. The vast amount of data collected allows nuances and overlays in fraud to be tracked and analysed, and consortiums are able to gather data from across geographies and industries, further increasing the richness of the overall data.

Preparation is key.

Predictive and pre-emptive fraud-prediction models are effective methods to ensure that customers’ security expectations are upheld. FIs need to provide customers with an efficient multichannel service that fits into their fast-paced lifestyles, whilst at the same time delivering them the highest level of fraud protection. Staying one step ahead of cybercrime is essential, and they can achieve this by leveraging analytics and predictive models.

Given the growth in technology that has led to the subsequent increase in fraud sophistication, financial institutions need to prepare themselves for, and ensure they are empowered to deal with, any possible situation. Leveraging historical transactional and non-transactional data, alongside consortium data, can—when combined with the outcomes of previous investigations that inform an analytic model—considerably help and lower the risk of fraud in organisations globally. This means customers continue to receive a seamless experience that leaves them satisfied and safe from possible cybercrime.


Related Articles

1 comment

QUEEN EMILY PILEO December 9, 2016 - 3:29 am



Leave a Comment

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.