Good day, everyone, and welcome to the Credit Acceptance Corporation Second Quarter 2020 Earnings Call. Today's call is being recorded. A webcast and transcript of today's earnings call will be made available on Credit Acceptance's website. At this time, I would like to turn the call over to Credit Acceptance's Chief Treasury Officer, Doug Busk.

Thank you. Good afternoon, and welcome to the Credit Acceptance Corporation's second quarter 2020 earnings call. As you read our news release posted on the Investor Relations section of our website at ir.creditacceptance.com and as you listen to this conference call, please recognize that both contain forward-looking statements within the meaning of federal securities law. These forward-looking statements are subject to a number of risks and uncertainties, many of which are beyond our control and which could cause actual results to differ materially from such statements. These risks and uncertainties include those spelled out in the cautionary statement regarding forward-looking information included in the news release. Consider all forward-looking statements in light of those and other risks and uncertainties. Additionally, I should mention that to comply with the SEC's Regulation G, please refer to the financial results section of our news release, which provides tables showing how non-GAAP measures reconcile to GAAP measures. We understand that our 10-Q is not yet available. Consistent with past practice, we submitted the 10-Q to the SEC at the same time we released earnings. Its release is delayed due to technical difficulties at the SEC that are preventing it from accepting filings from any company. At this time, Brett Roberts, our Chief Executive Officer; Ken Booth, our Chief Financial Officer, and I will take your questions.

Our first question is from Moshe Orenbuch from Credit Suisse. Your line is open.

Great, thanks. I guess, could you talk a little bit about when you made the cash flow assumptions this quarter, how did you take into account what's likely to happen given that as it stands today, stimulus ends tomorrow, could be renewed, could be renewed at the same, different level? Like how did you kind of take that into account as you look forward?

So in the first quarter, if you remember, we had a $44 million unfavorable change to our forecast that was— that reflected just the forecasting model reacting to what happened in the first quarter. And then since the forecasting model doesn't know that there's a pandemic or that we're in a period of economic disruption, we took an additional roughly $160 million write-down to account for what we expected would be a difficult period from a loan performance perspective. This quarter, we didn't make any additional subjective adjustment. You could think of the $44 million as a mechanical result and $160 million was largely subjective. This quarter, we didn't make any additional subjective adjustment. We gave you some collection numbers you can look at. Generally seeing favorable trends there. We just let the forecasting model fit out the result, and that's what we book to.

Right, Doug. But I guess I was asking the forecasting model books— I mean, what are the inputs? Does it look at— it just looks at— I mean, what are the inputs?

Well, the forecasting model's been in place for many, many years. It's an empirical scorecard-based approach, where it looks at many attributes of a loan and then forecasts future cash flows based on the historical performance of similar loans. And so that part hasn't changed. But it doesn't have macro factors in there, like when are they going to pass another stimulus bill or to what degree, or unemployment. It just looks at the historical performance of similar loans based on variables that we capture either at origination or throughout the servicing process.

I just want to make this clear. You said it doesn't take into account unemployment rates or the likelihood that that borrower will continue to pay. I mean, how do you assess the likelihood that that borrower will continue to make payments?

I think it's just what I said.

Look, it's based on the performance of loans with similar attributes and characteristics.

But it has nothing to do with the current state of the economy?

Correct. That's correct.

The subjective adjustment, as I described, is intended to take that into account.

Right. And you reversed part of that subjective adjustment, right? Now would you tick that back into income?

No, I'm not sure if you're listening. So we didn't make any additional subjective adjustment in the second quarter.

Right. And I'm looking at the front page of the press release. There was a remaining reversal of provision for credit losses of $14.8 million in the quarter, that's what I was referring to.

Yes, so that's the CECL adjustment. So we're just talking about the change in the net cash flows here.

Got it. Okay. I see. And so— and how was that number, that $14.8 million computed?

I mean, the $14.8 million is the change in the present value of the expected future net cash flows. It's the amount— let me say it differently. It's the amount of expense required to reduce the loan asset to the present value of the expected future net cash flows: collections less dealer holdback. In this case, it wasn't an expense. It was a reversal.

Our next question is from John Rowan from Janney. Your line is open.

Good afternoon, guys. Did you have to contribute any cash to any of the ABS facilities this quarter? Or not cash, any collateral?

No.

And then was there any discernible shift in the credit profile of your borrowers given that your vintages out another month and the loans this quarter were up to 59 months in duration?

No.

Okay. I mean, why go out further in duration when, obviously, the outlook for the next couple— at least next year is certainly a little uneven given the pandemic? And that's it for me. Thank you.

We didn't change our program during the quarter. So we offer a variety of loan terms. And it's the customer that selects the vehicle along with the dealer and the loan term that matches what their desired results. So we don't necessarily control the average loan term.

Okay. Thank you very much.

The next question is from Giuliano Bologna from BTIG. Your line is open.

To just kind of dig back into the model question a little bit. One of the things I was curious about was you were saying that it kind of goes back and looks at the statistical model that you use. And I was kind of curious from an accounting perspective, because you elected CECL, is there any requirement to use an economic forecast in the estimates that you use on a forward basis? Or are you using a little bit of a different framework there?

No, there's no requirement. It's supposed to be your best estimate. And if you go back, we've been forecasting for many years. We've taken a number of runs at including unemployment data in the actual forecasting model, and it just doesn't help the accuracy of the model, is what we've found historically. We went actually to even local unemployment data to see if we could figure out how to use that information to forecast more accurately, and it just didn't work. So if it was possible, we'd do it if it made it more accurate. But since it doesn't, we've just treated that like we did in the first quarter as a subjective adjustment, very similar to what we did during the financial crisis. We made an adjustment of a similar magnitude. And actually, in retrospect, it turned out to be pretty accurate. It doesn't mean this one will. I think I probably should emphasize like we did last quarter: we're in the early stages of this. There's a lot of uncertainty. I don't think you should look at our forecast either at the end of the first quarter or at the end of the second quarter as a hardened number. It's an estimate, and I think we've stressed that there's a lot of uncertainty around that.

That actually makes sense. What I'd be curious about is, how are you kind of planning or adjusting your business for the potential roll-off of some of the stimulus? And are you thinking about putting in place different programs to provide support to customers if they run into more trouble? Or how do you think about how to operate your business, given the uncertainty around, yes, unemployment— additional unemployment benefits and other things of that nature?

It's really what we've always done, so we've tried to run our business in a conservative fashion. We know that predicting the future is difficult. We don't expect it to always turn out like we planned. So we've been through lots of challenges over the years and the conservative approach is the only one we know, and that's the one we're going to continue.

That sounds good. Thank you for taking my questions. I will jump back in queue.

Hello? Hello?

I can hear you, Dave, but I don't hear anything else.

Looks like there's one more person in the queue, we lost our operator. Let's just give a second and hope he comes back.

Yes.

Let's be patient. In fact, the operator is responding via the chat. So he's trying to get back in.

Yes.

Apologies for the delay. Your next question is from David Scharf from JMP Securities. Your line is open.

Good afternoon. Still here, perseverance paid off. A couple of things. One is, I know in the past, you've mentioned that used car values have not had a material impact on kind of your all-in returns. Just given that the Manheim— that used car values are at such high levels at least at the end of June, can you just give me a 30-second tutorial again, why that shouldn't impact the net returns on the portfolio, particularly given the degree to which you repossess?

Here's how I would think about it. So if you originate a loan today, maybe it's a 60-month loan or a 48-month loan. Some of the loans that you originate, you're going to end up repossessing. And on the day you sell that collateral, it matters what the market is. In a good market, you'll get a little bit more money; in a bad market you'll get a little bit less. But what happens over the course of the 60-month loan is you'll go through ups and downs, probably three or four of them. And at some point in that spectrum, you'll get lucky and the market will be good. At other points, the opposite will happen. And so typically, if you just look at one month's originations, it sort of evens out. And so the way we look at it is we kind of track the values and look at the depreciation curve. If you look at it over an 18-, 24-, 30-month period, there's not nearly as much volatility as the way people look at the Manheim Index, which obviously changes every month. So that's one factor: you tend to have ups and downs and it evens out over a period of time. And then the other factor is repossession proceeds just aren't a huge portion of the total cash flows. Most of the cash that we collect comes from customers making their payments. So the combination of those two things mean that for us, used car values are a critical variable. Yes, sure, it matters to some extent. But because it evens out and because it's not a huge percentage of the cash flows, it's not as important as people perceive it to be.

Got it. No, that's helpful.

I guess one other thing I would add there is on the portfolio program, 80% of the increase or decrease in the value of the vehicle either increases or decreases the amount of dealer holdback. So the fact that we're splitting the collections on the loan with the dealer provides some insulation for us in terms of variation in vehicle prices and other factors that impact loan performance.

Got it. Got it. It sounds like level yield accounting allows for the short-term fluctuations in values to get kind of smoothed out over time or it just doesn't impact it as much. Understood. That's helpful. I'm wondering— I'm not sure if it was addressed in the last question. Just in terms of forbearance and deferral trends, obviously, for pretty much every consumer lending asset class, the percentage of accounts in some sort of deferral program peaked sometime in April or May. Can you give us a sense for how that's trended throughout the quarter? And maybe some context of maybe a year ago, sort of what the typical percentage of accounts are that are in some type of forbearance or deferral program?

Sure. So we— like we always do, work with customers that are having difficulty making their payments. The objective is to keep the customer in the vehicle. It's good for the customer, it's good for us. The reason we disclose the year-over-year change in cash collections and break out the front-end collections is it's really a simple way to see how the pandemic is impacting our results on a monthly basis. If you look at those numbers, it tells a pretty clear story. We started out in January, which is kind of your baseline, January and February. The results suffered in March and April, started to improve pretty dramatically in May and that's continued through the last date we gave you, which is July 28.

Okay. And I know I don't have those numbers right in front of me, but are they getting close to sort of pre-pandemic levels, percentage wise?

Well, I think, right now, they're better than pre-pandemic levels. If you can look at January and February as your baseline, we're doing better than that. I think the opposite side of that is we're— it's very early. And as other people pointed out, there's been federal stimulus and enhanced unemployment. There is uncertainty whether that's going to continue or how long it will continue even if it does. So early innings, things look pretty good. I think if you'd asked me back in March if I thought we'd be in this good a position, I would have said no. So we're pleased about that. But also keep in mind that it's very, very early, and this is going to play out over a long period of time.

Got it. And then lastly, just in terms of maybe— well, pricing and overall returns, stop me if this is a too simplistic way of looking at things. But instead of looking at all these different vintages every quarter, I just divide finance charges by your average balance for the quarter to get a gross yield, which is obviously that level yield that is kind of the average of all of the different static pools that you forecasted. At least this quarter, it looks like that I calculated the yield on a simple average, it is 22.6%. And it's actually— that's the highest in about five or six quarters. Did you see anything in the pricing environment or just the demand environment or how much independent dealers were willing to accept in terms of holdback? Is that an accurate calculation directionally that you're forecasting your portfolio to generate a higher yield, whether it's...

No, it's not. So I think what you probably did is you took the GAAP number. You have to remember with CECL, the GAAP finance charges are higher than what our economic yield really is. So you have to start with the adjusted numbers. What you find there is just like you would expect, when we reduced our cash flow forecast in March, the yield dropped. We got a portion of that back during the second quarter, but it's still lower than it was pre-pandemic. As the cash flow forecast recovers, the yield will recover over time. And if it doesn't, then the opposite. But no, the yield only increased in the world of GAAP; it's difficult to figure out the economics if you're looking at it through that lens.

Right. Okay. No, thanks for the clarification because it was obviously counterintuitive. Thank you and stay well.

Your next question is from David Feller from Wells Fargo. Your line is open.

Hey, good afternoon. I wanted to follow up on the prior question about just used car prices overall. You've obviously seen those bounce around. What are you guys thinking internally about used car prices for the year? Just seeing a number of different retailers and players in the auto space give various different estimates, so I wanted to hear your view.

We don't really have a view.

Okay. And then I appreciate all the color on the month-by-month changes in collections, both front end as well as total collections. As you look at the March and April data and the decline there, is there any way to think through how much of that was more driven by the auctions being closed versus just overall kind of collections on payments from customers? In other words—

The front-end collections would reflect payments from customers. Total collections reflect all cash flows. That's why we broke it out into two columns so you could assess the different trends between those two numbers. Clearly, we weren't selling cars at auction so total collections are going to be impacted. That doesn't impact front-end collections, though. We broke that out separately so you could see that.

Got it. And the prior question where you talked about it getting back to pre-pandemic levels. What are kind of the major drivers there of the year-over-year growth?

So in January and February, those months were unaffected because that was pre-pandemic. You go through March, April and May and you saw the impact of the pandemic and then June and July you see better results. So we're really starting in May. May, June and July you see better results, which likely reflects to some extent the impact of the stimulus money and the enhanced unemployment benefits.

Okay. Could you talk about regional performance at all? And maybe talk about where you have higher exposures in terms of geographical exposure?

I mean, we really aren't going to talk about anything beyond what we've already disclosed. We have disclosed in our 10-Ks where our top five states are. So you can take a look at that and draw some conclusions about what's going on from an economic perspective in all of those states. But we're not going to disclose anything beyond what's in the release and what will be in the 10-Q when it gets filed.

Okay. And then just last question for me. The decline in active dealers, can you just talk about major drivers there? And are you guys expecting any major overall changes in the industry in terms of dealer closures through the end of this year or maybe into 2021? Or do you expect it to be more muted?

I think you saw the active dealer count decline year-over-year this quarter. Attrition was, if you look at it in terms of the number of dealers, about average, maybe a little bit better than average. The difference from prior quarters is we just didn't have as many new actives as we have had historically. Part of that might be that our sales force was largely working from home. But the trends in new actives and new actives per person haven't been great. They were declining before the pandemic and you saw a continuation or a worsening of that in the second quarter, probably partially related to the fact that the reps were working out of their homes.

Okay. And then any thoughts on the industry overall, whether you'd expect more significant dealership closures?

No, I think it's hard to forecast that.

Okay, thank you for taking the questions.

We have a follow-up question from Moshe Orenbuch from Credit Suisse. Your line is open.

Thanks. Maybe can you guys just talk a little bit about the competitive environment and kind of how you see it at this stage in terms of the ability to originate in what you're doing?

Sure. It was difficult, as you know, before the pandemic started. There was maybe some thought that capital might dry up for the industry; that didn't happen. So you have lots of capital. All the same competitors are out there that were before the pandemic, so I think you can expect it to return to a fairly competitive state.

Got it. And you had mentioned before that unemployment doesn't factor into the models. But does unemployment factor into your ability to originate a loan? Like will you originate a loan to someone who is not employed or is employed and furloughed?

Historically, when we go into an economic downturn, we've done very, very well from a demand perspective. Whether that will happen again, who knows. But typically, that hasn't been the concern. This is a little bit different in that you have an economic downturn but lots of capital available. So we'll just have to see how it plays out.

Okay.

Thank you. With no further questions in the queue, I would like to turn the conference back over to Mr. Busk for any additional or closing remarks.

We'd like to thank everyone for their support and for joining us on our conference call today. If you have any additional follow-up questions, please direct them to our Investor Relations mailbox at [email protected]. We look forward to talking to you again next quarter. Thank you.

Once again, this concludes today's conference. We thank you for your participation.