Ways to mitigate fraud risk during a recession

As macro-economic storm clouds continue to gather, fraud, and payment professionals should be on high alert. Why? Because economic downturns often lead to a spike in fraudulent online activity. As the hunt for monetizable customer data and accounts grows in intensity, merchants, financial institutions and other organizations should investigate whether their fraud management tools are up to scratch.

They may find machine learning (ML)-powered solutions an increasingly useful way to mitigate fraud risk.

Emerging types of fraud

In the last few years many consumers have began shopping online for the first time, putting them at the mercy of cyber-crooks. Three quarters of global online merchants reported a net increase in fraud attempts in 2021 compared to prior years. ¹

Now the world is facing another period of prolonged economic uncertainty driven by inflation, continued supply chain issues and rising energy bills. Advanced economies are forecast to grow by just 1.2% in 2023, with recession likely to stalk some countries all year.² Unfortunately, this might also create the conditions for fraud to flourish – whether it’s:

• Payment fraud, using stolen card details and/or access to compromised accounts
• Account takeover (ATO), to steal saved payment and personal information in order to sell it on the dark web
• New account fraud (NAF) where stolen and synthetic identities are used to fraudulently open new lines of credit
• Friendly fraud, when legitimate customers file chargeback claims saying they have not received orders that, in fact, have been delivered. In 2021, nearly 40% of online merchants worldwide reported experiencing this type of attack, making it the most common type of e-commerce fraud¹

According to some predictions, e-commerce losses to online payment fraud alone is set to soar by 17% between 2022 and 2023 to reach $43bn this year.¹

In a downturn, individuals may be more likely to lose their jobs. That, coupled with rising living costs and rampant inflation, could cause some to suffer severe financial distress. This in turn may tempt some individuals to engage in first-party fraud (“friendly fraud”). It may also make them more likely to fall for a phishing attack in which fraudsters pretend that they will be forced to pay a fine if they don’t reply to the message, sharing their personal/financial details. It might also increase the number of people who would be prepared to share such details in return for filling out a fake survey or similar. And it could persuade others to become money mules, as a way to make extra cash.

Machine learning algorithms and solutions for fraud detection

The cumulative impact of these trends could be to increase the risk of first- and third-party fraud for organizations operating online this year. The question is what they should do about it. From a fraud prevention perspective, the combination of ML and rules-based solutions is increasingly compelling.

ML algorithms can process large datasets with multiple variables, to uncover complex patterns that human eyes might miss, indicative of sophisticated fraud. They’re also able to continually adapt to keep pace with changing patterns. IDC predicts that by 2027, 10% of attempted identity fraud will be reduced because of more sophisticated AI/ML and deep learning algorithms.³

Rules-based systems form a solid foundation for fraud prevention, but must be written by analysts and can become overwhelmingly complex and cumbersome to maintain as they expand. That said, they can be written to mitigate fraud risk for organizations whilst ML engines are being trained. Teams could even use ML algorithms to suggest new rules to write.

A two-sided payment network serving up comprehensive, high-quality, fresh data from consumers and businesses is the perfect fuel for intelligent ML engines. This is the value of PayPal’s platform, which now serves over 400 million active accounts across the globe and more than 25 million businesses.

Whether fraud attempts spike during the coming downturn or not, it’s time for a smarter approach to enterprise fraud management.

Learn more about how PayPal is helping merchants address changing fraud.