A contributory database in which lenders share their knowledge helps the entire industry to reduce fraud. Recent fraud trends highlight the importance of lenders working together to proactively address fraud before we have all been swindled of our money. Utilizing a feedback loop and confirming fraud across multiple channels increases Informed’s accuracy rates to stop fraud in its tracks. Fraud is not lender specific and has a significant impact across the banking industry.
As prices go up and loan originations shift to digital, income and employment fraud is rising, costing lenders billions of dollars each year. In just one sector – employment and income fraud – cost auto lenders $4.7B in 2021.
Financial Services Information Sharing and Analysis Center (FS-ISAC) is a cyber intelligence collaborative focused on financial services. The organization leverages its intelligence platform, resiliency resources, and a trusted peer-to-peer expert network to anticipate, mitigate, and respond to cyber threats. FS-IAC is one of the few contributory communities to leverage a robust peer-to-peer network to identify and resolve cyber threats. Networks like these help fight fraud as a community. Police departments report that identifying fraud within financial services leads to arrests of individuals who often have extensive criminal histories in multiple states.
Within the auto industry, police departments launched grassroots efforts aimed at stemming fraudulent activities such as organizing vehicle fraud units and auto theft divisions to support and train local dealers and lenders on ways to combat fraud. Houston Police Department’s Auto Theft Division/ Vehicle Fraud Unit is one such unit that has helped prevent $740,000 in fraudulent vehicle purchases. New I.D. scanning technology plays critical role in helping HPD catch fraud suspects in the act. One of their main focuses is on ID scanning devices and identity theft, which is a crime that can result in felony charges and arrests. However, because the focus is on identity theft, there is a layer of fraud that sometimes goes unaddressed and is critical to a lender’s ability to ensure a consumer can repay a loan: paystub fraud. People report fake income and support it using fake documents like paystubs, bank statements, etc., which can affect credit risk and portfolio performance.
What is paystub “fraud,” “scam,” and “alert”?
All these words put dealers, lenders and banks on edge and make you think of the “darknet.” When a paystub is falsified, this paystub fraud. There are many types of paystub fraud, and as a defender of fraud, one needs multiple ways to defend against criminals.
Paystub fraud is on the rise, and the same tactics that fraudsters use in phishing emails are also used by cyber-criminals, who use technology and tools to commit fraud. As referenced above for ID cards, there are many solutions and solution providers because IDs are standard and government-issued; however, paystubs can be issued by any registered company. This makes this problem exponentially harder. You can’t just call a government entity to verify a paystub! They are no defined standards.
You have to keep up with a growing list of fraudulent paystub sites ensure you can catch fake documents without causing inconvenience to borrowers. The last thing you want to do is blame an innocent consumer for faking an income document. But lenders also have to weigh a 90% increase in delinquency within the first 60 days if income is overstated, and analysis shows that income is overstated on approximately 20% of loan applications. This is the dilemma.
Sending a fake/ look-alike template-based paystub:
Hundreds of websites have sprung up in the past few years which allow you to make a fake document within seconds. Some of them charge you up to $10! A quick google search of “Fake paystub generator” shows you the number of such websites.
Here are some examples of fraud we see very often:
How to spot fake Paystubs: 3 Signs that a Paystub isn’t Real
Paystub fraud detection can’t just be a whitelist logic; it has to be the other way. For example, currency and driver’s license fraud detection can utilize whitelist logic. An analyst can verify a 50-point checklist for accuracy and completeness, such as watermarks, codes, etc. For paystubs, since there are no standard verifications, an analyst has to identify random discrepancies out of an endless list of possibilities. Identifying discrepancies in format is very difficult without a point of reference. Some criminals try to generate paystubs that look like well-known providers like ADP.
1. Missing Data:
Seeing missing data is a sure way to flag a pay stub as questionable. There’s a new trend of selling paystubs on e-commerce sites such as etsy.com vs. paystub generators as consumers think these are authentic or more valid but are usually the same template-based forms found on those websites, as seen in the below comparison.
2. Font variances, incorrect logos, logos incorrectly positioned.
Subtler variations without a reference point is a larger challenge for manual underwriting and funding. Leveraging adaptive behavioral technology is key for banks to successfully combat financial crime and fraud. The intersection of technology and innovation provides tactics to combat rapidly evolving fraud trends.
3. Same details with different demographics.
These are also more difficult for humans to identify as they need to review many fraudulent paystubs and be able to recognize that the paystub has slightly different demographics but was repurposed from another application. When we see this issue occurring at the same dealership, this can be an issue with an employee. This would be considered an Insider Threat – a person who intentionally misuses authorized access to negatively impact the confidentiality, integrity, or availability of an organization’s information or information systems. Prepared protocols should be in place to help mitigate risk and respond appropriately.
When this issue occurs across multiple dealerships, it is even more difficult to address and is most likely a fraud ring or a bust-out where several individuals are purposefully purchasing vehicles with no intention of paying back the loans. When there is a bust-out, the individuals are purchasing multiple vehicles from different dealerships and hoping the lenders and dealers do not communicate. In these instances, it is critical to have AI/ Machine learning in place to detect fraud in these overlapping data points.
Remember this is only one type of fraud, and the effort and cost to train employees to manually reference fraud flags is time-consuming and ineffective as fraud scams change constantly. Fraud templates are being created as you read this, and fraudulent paystub websites are being created to assist this billion-dollar market. Fraud detection is a complex undertaking and depends on timely reporting, consistent identification, and implementation of lessons learned.
How common is paystub fraud?
Across our portfolios, we see a 2.25% average fraud rate; however, when you review this by dealer, there can be significant outliers with fraud rates from 10-12%; when Informed implements additional fraud controls – automated, fraudulent paystub checking, and paystub accuracy – fraud is significantly reduced as well as its potential business impact.