Before you put your money at risk in a Vegas casino, you’d want to know if the dealer was using hidden tactics to rig the game. The same should be said for prudent auto lending and investing – market participants should want to know that the auto dealer hasn’t rigged the game by deceiving the purchaser in ways that make default more likely. Yet, recent government investigations and private actions against auto dealers for misleading and deceptive practices indicate that there could be a lot of latent risk lurking in the system related to dealer fraud. It’s often thought of as a consumer issue, but it’s actually much more than that – often when the consumer is defrauded, loan performance suffers.
With this in mind, we spoke to Joshua Wortman of General Forensics. Josh is an A.I. and Information System professional with nearly two decades of experience creating risk and analytics products for auto lenders and consumer finance companies. Given the potential for increased losses in loan and credit markets if fraudulent practices remain hidden, we asked Josh to share his familiarity with the analytics to advise on the largest areas of fraud that could impact loan and securitization performance.
Industry Professional’s Advice to Guard against Dealer Fraud for Better Portfolio Performance
How significant an issue is dealer fraud and should investors in securitizations be concerned?
Not all kinds of dealer fraud require investor attention. Investors must only be concerned with the type of dealer fraud that has a relationship to portfolio performance. This kind of dealer fraud is referred to as “consumer abuse” by law enforcement and the regulators trying to stop it. It is subtle. Consumers don’t even realize it is happening. The loans are getting funded, securitized and they even perform ok in the short run but down the line, they start to default at a higher rate than models predict. I’m talking 10% to 30% of loans are affected depending on the originator. Even when it is possible to stop, lenders have not prioritized stopping it, consumers don’t realize it is occurring and investors are left exposed. I think that’s a huge issue and the whole dynamic lays the groundwork for regulator intervention.
What are the main types of dealer fraud that could impact loan performance?
There are several types of “consumer abuse” dealer fraud that investors need to worry about.
One type is deceitful sales tactics, such as lying to a consumer. For example, some dealerships falsely warrant that faulty cars are safe, repaired and in good condition to be sold. But when this is a lie, it sets up a consumer for substantial repair bills, leading to unaffordability and, ultimately, loan failure at a higher rate than a credit score predicts. This form of fraud seems very seductive for dealers, as even big name dealers, such as CarMax and JD Byrider, have found themselves in trouble for this in the last few years.
A second type of consumer abuse is lying about consumer income or down payment in a loan application. The consumer is unaware it is happening because a dealer employee is entering the numbers into the computer loan application. The fraud makes it appear the consumer has more skin in the game, and makes loan terms seem more affordable to the consumer than they are in reality. Investors need to know this leads to worse than predicted loan performance. Others, including dealers such as Hallman Chevrolet, have gotten into trouble.
A third type of consumer abuse is selling a car for more than the advertised price. This is illegal in some states. For example, California vehicle law code states that “it is a violation for the holder of a dealer license to sell a vehicle over the advertised price regardless if the consumer has knowledge of that price.” But it happens anyway, showing up in over 10% of subprime loans and leading to 40% more defaults. It’s like power booking, but they don’t need to fudge the options, they could just make up a new sale price. This is easy to test for in a portfolio of loans and I recommend doing so.
How can investors protect themselves from bad dealers?
I think investors would benefit, in their portfolio predictions, by trying to understand more about the dealers involved in those portfolios. To do it systematically, you need to acquire risk data about dealerships and figure out how to use it in your pricing models, just like you use consumer risk data in your pricing models. There are various data sources available for this. Investors also probably want to measure what portion of loans in a portfolio involve cars sold for a legal price, which is to say “as advertised.” There is data for that too.
In the end, car dealers are sales organizations. Their job is to move product, which isn’t easy. In order to do this, they get “creative” and – call it fraud or not – that creativity leads to worse than predicted loan performance. The issue has become a reality and we ought to think about ways to watch out for it.
At General Forensics, we help manage this diligence for investors – whether at the whole loan level or the securitization level.
Looking Ahead
Going forward, and especially in any downturn, it is likely we will learn the true extent of misleading practices by certain dealers. But there is no reason that market participants should be surprised when the truth comes out. Given dealer fraud has shown to be correlated with an increased risk of loss, the business and legal diligence on any transaction should recognize that the how and why of an auto sale transaction is more important than the sale itself. To Josh’s points, analytics can provide the necessary insight into the behavior that is driving dealer sales.
