…continued from part 1 of AutoZoom interview with Scott Carlson
Kenny: …Can AutoZoom predict which customers will pay slow or default completely and when? Can it do that?
Scott: We get that question a lot and when some of them ask,
- Can you tell me when a specific customer is going to default?
- When they’re going to repo?
- When they’re going to pay slow?
No. In fact I don’t know a technology that exists that can do that…not even in the banking environment. However Autozoom is more actuarial and the way that it works is let’s say for example you originated 100 loans over a period of time and those 100 deals fall into different ranges of strength or weaknesses whether applying a number to it or not. At AutoZoom we do apply a number to the various strengths and weaknesses of customers. By looking at all the analytics of over 1,300,000 deals being scored through our system, and over several hundred thousand repossessions following that, we can tell you certain groups with a score of “X”, your repo rate will be at 30% with a standard deviation of 2.5. We can’t make those kinds of predictions on a specific customer, but we can give you a range of probability based on various factors and scores, including the Customer Score.
For example, you might get a low scoring customer designated in a failure rate. Even though you probably shouldn’t even do business with them, you go ahead and score up a deal structure with a lower car cost and a bigger down payment. Maybe you don’t even do that and you just do a standard deal. Maybe that customer ultimately comes through for you and never misses even a single payment. Unfortunately, that’s just not very likely. There is a very high probability that they will pay slowly, which will lead to difficulty and trouble keeping up and, ultimately, that will lead to repo. Not only that, they will repo much earlier than a customer who would score out much higher because of their strengths.
You also might get a customer who scores in the 400 range. (The ideal AutoZoom Customer Score is somewhere around 375.) Well, you’d expect the higher-scoring customer wouldn’t default but they might repossess in 45 days and just shock you to death! The system cannot predict live, unexpected events that may or may not come up. But, what it can do is measure the strengths and weaknesses of customers, their habits, and their ability to keep a job, to keep a residence, and things like that. This is totally different than a score from a credit bureau because a credit bureau re-measures entirely different things. Again, that high score doesn’t guarantee you that a certain customer is going to make every payment perfectly. It doesn’t guarantee they won’t miss payments. It doesn’t guarantee a thing.
What AutoZoom does do is serve as an actuary indicator to suggest which customers may or may not default and when, during the life of the contract.
In other words, AutoZoom can help you to understand when you can afford to take a smaller down payment on a higher-end car to be able to capture those higher-end customers – customers who previously might have walked off the lot if you weren’t willing to sell with low enough down payments or loan for a higher-end car. That’s a problem because, frankly, Americans are kind of spoiled; we’re all kind of spoiled. In the buy-here-pay-here industry, anyone or any dealer who can afford to sell vehicles in the highest car-cost range (i.e., vehicles with the lowest miles, with the most equipment, for the longest term, for the best terms, best interest rates, to the right customer for the least down), that’s the dealer who will get the majority of the business.
Kenny: Right. They want everything. So, another question for you is if our company has a low repo rate would I want to be a Zoomer and why?
Scott: Well, a lot of guys believe having a low repo rate is really the name of the game; the goal is to be somewhat focused on keeping your repo rate low. In our 40-year history in this business, and with all the analytics we have performed, we found there is actually a relative optimum repo rate to maximize sales and collectability – relative for different dealers for different parts of the country. A 30- to 40-percent repo rate might mean that out of the 100 deals you put on the books, you could expect 30 to 40 percent of those 100 deals to eventually default and repossess. Of course, if 30 to 40 percent of the deals default within the first 90 days, that wouldn’t be a good thing. But, if they are defaulting much later in the life of the contract, let’s say during the peak of time when they’re repossessing maybe seven to 11 months out into the future of the contract, then that’s a very strong business model where you’re only writing off maybe 15 percent of your dollars in debt. A lot of dealers may or may not know what that means, but it’s a very good business model. A low repo rate is not necessarily a good thing at the expense of grooming too many deals off the lot.
Kenny: Sure. The goal should be overall profit and if that means a higher or lower repo rate, then so be it; which brings us to the next question . . .
Scott: Retail dealers typically have less focus on repo rate and more focus on collectability and selling all the cars you can without compromising that collectability.
Kenny: Right, which is a good segue to the next question: What about the other side? What if the repo rate is too high, can AutoZoom reduce the repo rate without slowing down the sales?
Scott: Well, that’s usually the concern most dealers have. If you’re going to work on that bottom tier or the low-scoring customer and try to eliminate those, will you slow your sales down? The answer to that question is: It will eliminate some of those lower-end customers to the degree they are not willing or able to make the higher down payment needed, consequently moving them to a lower-end car. You may lose a number of those customers because that’s what the model is telling you to do. But, you’ll make up for those lost sales by being more focused on and capturing the higher-end customer, even though you might have to accept a lower down payment for a higher-end car. Most dealers will tell you they know they are missing the higher-end customers and they’re not sure what to do about it. It’s hard to know how to move up into that higher-end car bracket with a lower down payment if you have not been there before. It’s difficult to change your business model. I would say that with the higher repo rate you should be more concerned about losing some sales of that bottom. It’s not a bad thing to lose some sales; but, you must be aware when the economics of those are not working. You’re just kind of fooling yourself because even with a lot of activity, you’re probably killing your overall profits by doing too many of those. Like I said, you can try to make up for it by capturing the higher-end customers with lower down payment requirements on higher-end cars.
Kenny: Right. Scott correct me if I am wrong, but with all the adjustments you probably made over the years to your technology in your system after analyzing the analytics, would you say that you have gotten more and more accurate simply because in addition to striving for constant improvement you have a very large sample size. Would that be safe to say?
Scott: I’ll tell you what’s more important.