By Todd Roberston
The Fed, FDIC and OCC recently issued a joint statement requesting public input on how to combat rising levels of payment and check fraud. With fraud often spanning multiple institutions and payment rails, no single entity, public or private, can take a lone wolf approach to tackling this issue.
This type of coordinated response from three regulatory bodies reveals the growing urgency and complexity of payment and check fraud. Check fraud accounts for more than $1 billion in losses for U.S. financial institutions and remains a persistent threat with no end in sight.
The call for public comments ends on September 14th, so any forthcoming guidance, new rules, or broader industry shifts will remain to be seen. In the meantime, financial institutions can take proactive steps to bolster their fraud mitigation strategies, protect their customers and reduce risk across all channels.
Check Fraud Is Alive and Thriving
Paper checks usage has steadily declined over the past few years as digital payments and ACH transactions become more widely adopted. This doesn’t mean check fraud has followed the same downward trajectory – in fact, it’s done the opposite.
An annual survey conducted by the Association for Financial Professions (AFP®) identified that among all payment methods, checks were the most susceptible to fraud, with 63% of organizations experiencing attempted or actual check fraud.
Fraudsters are constantly evolving alongside financial institutions, becoming more sophisticated in their attempts to swindle banks and their customers. They view checks as soft targets that are easy to exploit and yield large payouts, especially as checks are often used for big-ticket expenses (i.e., rent, utilities and income payments).
A popular method of accessing checks involves fraudsters stealing paper checks placed in U.S. Postal Service (USPS) blue collection boxes or even robbing USPS employees out on deliveries. The issue has gotten so severe that the FBI and USPIS issued a public service announcement at the beginning of the year warning consumers.
Once fraudsters get their hands on legitimate checks, they either alter them, or in other instances leave them unaltered and deposit them with forged endorsements. Check washing is a popular alteration method among fraudsters. Using chemicals, they can “wash” ink from a legitimate check and alter the original payee and financial amount.
Another method called check cooking takes this a step further. By digitally manipulating a stolen check image using photo editing software, fraudsters can create multiple fraudulent checks. These checks are typically written for small amounts to avoid detection and prolonging the scheme.
It’s common for fraudsters to conduct repetitive, small deposit account transaction fraud attempts. Due to their small size and frequency, it’s challenging for banks to prevent these fraudulent transactions. This issue is exacerbated for banks that are reliant on legacy systems that can’t keep up with more advanced types of check fraud.
Be Equipped to Combat Fraud
Fraudsters are constantly looking for vulnerabilities in a bank’s defenses. Staying a step ahead of evolving fraud tactics requires banks to deploy preventative measures early, ideally at the point of disbursement, and continuously monitor activity throughout the entirety of the clearing process.
Account and check validation services that support image-based verification and check authentication can help close gaps in a bank’s armor. Through the utilization of advanced algorithms, such as unsupervised machine learning, modern solutions elevate fraud detection and minimize financial loss. By applying machine learning constructs, institutions accelerate continuous business improvement.
Machine learning enables computers to detect their own weaknesses and provide feedback when they encounter things they do not understand based on low-confidence computations. Using this data, the solution improves the machine learning training for set populations, permitting a low-confidence pattern detection to become a known pattern within the algorithm.
Here are five structural pillars to consider:
- Use multiple machine learning methods to improve fraud pattern recognition. New machine learning innovation significantly improves both pattern recognition and predictability.
- Leverage ‘sensory monitors’ for key event, outlier and threshold recognition.
- Consider a larger relevant contextual data universe in machine learning predictive decisioning.
- Deploy a combination of transactional data and image analysis data in the scoring mechanism.
- Improve disposition electronic workflow efficiency.
The Bottom Line
The reality is check fraud isn’t going away anytime soon. Banks must implement modern fraud prevention strategies that use real-time data and automation. Staying diligent and responsive to new threats is critical for stopping fraud before it impacts customers and causes irreparable reputational damage to institutions.
About the Author
Todd Roberston serves as the SVP at ARGO. Founded in 1980, ARGO develops, installs, and supports high-value mission-critical software for the financial services and healthcare industries. ARGO currently works with nearly 500 banking customers in all financial services sectors, including six of the top 10 banks and non-bank financial services lenders. ARGO also provides solutions for a leading-edge healthcare information exchange and major healthcare providers.