Credit Acceptance Corp (CACC) 2017 Q4 法說會逐字稿

Good day, everyone, and welcome to the Credit Acceptance Corporation Fourth Quarter 2017 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 Senior Vice President and Treasurer, Doug Busk.

Thank you, Andrew. Good afternoon, and welcome to the Credit Acceptance Corporation Fourth Quarter 2017 Earnings Call. As you read our news release posted on the Investor Relations section of our website at 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.

At this time, Brett Roberts, our Chief Executive Officer; Ken Booth, our Chief Financial Officer; and I will take your questions.

(Operator Instructions) Our first question comes from Moshe Orenbuch with Crédit Suisse.

Great. I guess I was sort of hoping that you could expand a little bit on the discussion in the press release about the changes kind of in the timing of the receipt of the cash flows in your new estimates? And maybe if you could, just relate that to the fact that now you've got kind of negative variances on 3 years relative to your initial forecast and how we should kind of think about that?

Okay. Sure. So 2 different things that we have. As you know, a number of models that -- statistical models that we use for forecasting. We use the models not only to run the business but to support our financial statements. One of the models, the timing model, which predicts or forecasts the timing of future net cash flows was revised this quarter. So a couple of things. We described kind of the reasons for the revision and the impact in the release. So I won't necessarily repeat that, answer any specific questions you have on it, but just want to make a couple of points. The first is it's a timing change. So it doesn't change the amount of revenue that we'll recognize in the future for adjusted earnings. And for GAAP, it's the same concept. You just have to include the provision for credit losses in as well. So if you look at kind of net revenue or revenue minus the provision and our GAAP statements, including the fourth quarter and going forward, the total amount that we'll recognize is unchanged. The only thing that changes is the timing of when that revenue will be recognized. The second thing I'll highlight is it's a large estimate, $6.3 billion of undiscounted net cash flows spanning 120 months. So it's a large estimate. It covers a long period of time. To forecast the timing, we have to forecast the customer payments including any prepayments, the amount -- timing of repossession proceeds, deficiency balance collections and then the outflows for portfolio profit and portfolio profit express. So not only is it a large estimate, but it's got a lot of moving pieces. Now even though it doesn't impact the amount of revenue on the loan portfolio going forward, it does impact the economics of the loans, particularly the longer-term loan. So based on our revised model, the longer-term loans are less attractive than they were before. The shorter-term loans are the opposite. Now the longer-term loans are still very attractive, given the price that we paid for them. They're just less attractive than they were before.

I guess I'm going to have to come back to that because when you talk about -- this is level yield accounting, and so determining that yield, the factors to determine that yield, the 2 factors, are the total amount of cash and the timing. So if the time is longer then the yield is lower and that is an ongoing effect, correct?

The yield is lower, you're correct, but -- that's true. But it's also true that the total amount of revenue in our adjusted statements or the net revenue, as I referred to it in our GAAP statements, is on effect. It's just the timing of when it's recorded.

I guess I'm missing how the net revenue or GAAP statements can't be lower that....

Well, the revenue, if you think about it, is just the outflows. The inflows that we expect in the future minus the outflows that occurred as a result of us acquiring the loan. So the difference between those 2 numbers is our revenue. We didn't change the amount that we're expecting on it...

Right. But at any given year, it would be lower. Wouldn't it, Doug?

Well, it would be either lower or higher. That's what we're saying is the timing change, so it could work both ways. What you're thinking about -- sorry, go ahead.

This change would mean that in any given year, that you're now going to be applying those new standards to it would be lower than it would have been before, correct? Under your current models.

(inaudible)

Our next question comes from David Scharf with JMP Securities.

Maybe switching away from the accounting for a moment. You, obviously, saw a pretty material surge in volume this quarter and I think an acceleration in the year-over-year growth in active dealers. Can you give us a little bit of a sense of what you're seeing from competitors that's perhaps driving this surge? Any sense that they're easing up? Or by the same token, it looks like you're going to -- continuing to extend the loan term, it was 56 months an average this quarter. Is the volume more a function of actions you're taking as opposed to what competitors are doing?

Yes. I think it's always difficult to separate those 2 factors. The best number you can look at in the release, I think, if you want to try to get a sense for the competitive environment is the change in volume per dealer. That was up 2.1%. That's a positive change in the trend line there. We've been running negative numbers there for, I think, 7 consecutive quarters. So that's a positive data point. It's just 1 quarter, so I think it's probably a little bit early to draw any conclusions about the competitive environment based on one data point, but it is a positive one.

Got it. And this is a very hypothetical question, but are you willing to venture sort of maybe what some of those volume metrics would have looked like had, let's say, the -- either the average size of the loan remained maybe at levels that we saw earlier in the year or that the average term had remained the same? I'm trying to get a sense whether those 2 metrics, number of months and average loan size, are more a function of the type of customer, the type of vehicle or if it's in response to competitive issues?

Yes. I can't give you a figure. But I think we've changed our program to make it more attractive, offering longer terms. No question that's had a positive impact on volume.

Got it. And maybe one follow-up. Can you expand a little more explaining the language in the press release regarding, I guess, the reclassification of some dealer loans into purchased loans? Number one, is this something that you've historically done in the past but maybe not as greater degree? And for our purposes, are there any implications about how the yields, going forward, on those loans are going to be treated?

Yes. I don't think I would worry about that too much with -- the reason we put that in the release, I mean, the answer is we do reclassify loans from dealer loans to purchase loans if the dealer don't meet their performance obligations. We're entitled to do that under the terms of the agreement. It's typically small dollars. What happened this quarter is we had a catch-up. We haven't been reclassifying those loans in a timely basis, and so we caught that up this quarter. The reason it's in the release is it impacts some of the numbers in the collection and spread tables, so we changed we way we do those tables. At this point forward, the tables will reflect the way the loan was initially recorded. And that will just allow us to keep those numbers consistent over time. So we highlighted it in case you had a record somewhere up. What was in prior releases you wouldn't be wondering why they didn't tie out.

Got it. Got it. And then just lastly, anything in the fourth quarter to report from the hurricanes? Any residual impacts and so forth that we might take note of?

No.

Our next question comes from John Rowan with Janney.

So I'm trying to tie together a few comments that you made. So duration's up 56 months from 55 last quarter, but you're also talking about the economics of the loan being substantially less attractive on the longer-term loans versus the shorter-term loans, but you're still increasing duration from, I guess, from a competitive standpoint. Is there any impact on the dealer partners as far as your advance rate or hold back that's impacting their desire to push loans under the Credit Acceptance's platform based on the changes that you're making to the collection patterns?

I don't think the changes are big enough that the dealers would notice. I mean if our new forecast is accurate, it means that cash flows will come in a bit later. But I don't think an individual dealer would notice the difference.

But with that -- I mean just for -- if you took loan A, give us an idea of how much longer the collection period is? I mean, is it 10 months additional, 5 months, 20 months? I'm just curious to know how much longer the collection period lasts to get you to the same forecasted collections?

Yes. You're talking about a very small difference there when you're thinking about it in those terms. So it's a 1% to 2% type adjustment, not 20%, 30%, 40%.

Okay. And then just to switch gears a little bit. You talked about the holdback forfeiture in the press release. Was there any gain reported from holdback forfeitures?

No. It's taken over time. We just account for that loan as a purchase loan going forward.

Okay. Any impacts from the changes in the sales force for you guys have made over the past year? Any -- is that driving a higher dealer partner productivity numbers?

Yes. We're starting to see some positive momentum there. It's early but we're seeing a nice progression. We started expanding the sales force in mid-2016. The group that we hired sort of the last 6 months of 2016 grew faster than our overall average this quarter. The group we hired during the first 6 months of 2017 also grew faster than the overall average. They didn't grow as fast as the first group, but still higher than average so we're starting to see a nice progression. And then the ones we hired in the last 6 months of 2017 grew slower than our average, so some positive change there. The average size of the sales force in Q4 was up, I think, 27% over the prior year same period. Unit volume was up just short of 11%. You'd like to see unit volume growing faster than the size of the sales force. We're not quite there yet, but we're seeing, like I said, some small movement in the right direction.

Okay. Two more quick questions. Repurchases, any plans there? It didn't seem like you repurchased any stock this quarter?

We didn't buy back any stock this quarter, and we'll continue to think about it the same way going forward. We bought back a lot of stock over time so the repurchase patterns are a bit lumpy and they may will likely be lumpy in the future.

Okay. Last question. We don't have a K yet, so I just want to know if there's going to be any changes to the regulation or contingent liability section?

I mean, we'll release the 10-K in another week, 10 days or so, so we can talk about it at that time.

Our next question comes from Vincent Caintic with Stephens.

Just wanted to get maybe the practical drivers of maybe some of the questions that were asked. So on the provisioning and the timing adjustment before the provisioning, is there maybe something that would drive that from the customer perspective? So I'm just thinking of potential example if maybe you're doing more workouts with your customers where you might restructure some of the loans and that would drive a timing change. I'm just trying to think of examples of what would cause timing to be different than to the -- for the provision to be higher?

As I mentioned it, it's a large complex estimate covering a long period of time, lots of moving parts. It was very small adjustment in the scheme of things to a large number. We highlighted the most important thing that changed, which was our estimate for the longer-term loans when we built the model. We had less than a full data set for those longer-term loans. We now have more data, which we used up data estimate and attempt to make it more precise.

Okay. Got it. And then for the forfeiture of the dealer holdback. Could you maybe just give us a background in the types of performance obligations the dealers have to meet and so what would cause the forfeiture to hold back?

Yes. The simplest one is that they have to do 100 loans in order to be eligible for the dealer holdback. So if they close their first pool of loans, which is they have to get to 100 to close it, if they don't make it to that 100 milestone, then we're entitled to retain the dealer holdback so that's what that is.

Okay. Got it. So would there be a mix of new, relatively new dealers that will be under this -- the forfeiture?

It would be all dealers that hadn't got to 100 would be included in that.

Okay. Okay. Got it. Just one more question, any updates on competition that you're seeing in the environment? We're hearing of maybe some folks pulling back or maybe other folks getting more competitive. And just what's your sense?

There's not a lot to what we said already. I don't see a big change in the environment at this point. As I said, we had one positive data point for the quarter, which was the increase in volume per dealer, but I think it's just one data point.

Our next question comes from Jack Micenko with SIG.

Looking at the positive variance that continues in '17 and then, I guess, in particular in the purchase loan portfolio. I guess my question is, is that -- are you seeing better underwriting and better quality in the purchase book? Or is that more a function of going in your estimate might have been maybe more severe than actual or some combination of both? Just trying to get a sense of why the purchase numbers keep driving some better positive variance, overall, in the recent quarters?

I think the best way to think about loan performance is the number that we provide that show you the change in our net cash flow forecast for the quarter. It's $13.7 million. So it's a positive number, which is nice, but that's again on $6.3 billion of undiscounted net cash flow. So $13.7 million divided by $6.3 billion is a number that's pretty close to 0. I think the most reasonable conclusion to draw from that is that loan performance was stable for the quarter, and it's been stable for the last several quarters. But -- that's good news, but I don't think there's really a lot more that you can say about changes in the forecasted collection rate because it just didn't change that much.

Okay. And then I think back in June there was a -- the CFPB reached out with some inquiries. Is there any update on the status of that?

No update at this time. We'll file our K, as Doug said, in a short period of time and you can look for any updates in there.

Our next question comes from Leslie Vandergrift with Raymond James.

A quick question on the change in that cash flow timing. I know you discussed it earlier in the call, but does any of that have to do with the upcoming accounting standard changes, specifically with the ASU 2016-13 change?

No.

Okay. So then any update on those changes coming forward?

No. We're continuing to work on it. But at this point, we don't expect to reach a conclusion until late '18 or into '19. So when we determine the methodology and quantify the impact, we'll disclose it.

Okay. And then for this quarter, on the adjustments to the floating yield, is it a positive $37.9 million adjustment and in the breakout, about $12 million of that was due to those onetime estimate changes. What about the other $27 million approximately? What was built into that?

I don't think we follow the question. Can you -- I'm sorry, can you repeat that? I didn't really understand it.

Okay, sorry. The floating yield adjustment for the quarter was $37.9 million. When you look at adjusted revenue from GAAP, about $12 million of it seemed to be -- to do with the timing of the changing -- change in estimates due to timing of net cash flows. What about the other $27 million? Like, what else was in that?

I don't know, you stomped us on that one.

Okay. All right. And then on -- sorry, go ahead.

[Well that's just the] difference because the GAAP provision for the adjustment is taken onetime.

Maybe you can call Doug tomorrow, and he'll have an answer for you.

Yes.

Okay. That's fine. I'm going to follow up on that. And then last question. You discussed earlier in the call that the longer-term loans are becoming less attractive out there. So does that mean in the near term we could see those terms start to pull back? So 56 months average term in fourth quarter. Could we be heading back the other direction?

I think we'll just have to wait and see. I mean, that's certainly one factor. Again, I think one other questioners said I said it was a substantial revision -- well, I don't think it's a substantial revision. They're just less attractive than they were before, but still very attractive. So we can continue to write longer-term loans and we find the pricing attractive, we'll continue to do that.

Okay. And then, I guess, a follow-up on that is as you've seen -- obviously, you got the dealer holdback back this quarter for the ones that didn't meet the pool requirements. Have you seen the ones who were close? Are we seeing those dealers who get close to 100 with the pool doing more of the longer term? So maybe they're less attractive than the quarter before, but they're still willing to do them in order to ensure that holdback is paid to them?

No, I don't think those 2 things are related.

Our next question comes from Kyle Joseph with Jefferies.

Sorry, I had to hop on a little late. So apologies if these are already asked. But in terms of the tax rate going forward, have you guys -- do you -- where do you anticipate that 37% moving towards in terms -- for adjusted EPS?

We say 23% in the release.

Okay. And then just as a follow-up on the provision and the changes there. Is this just a onetime item you see? Or do you anticipate there being a future impact as well?

Well, I mean, the revision in the quarter was it was fully reflected in the quarter.

Our next question comes from Jason Hahn with Principal Global Investors.

I guess, maybe to explore that last question just a bit further. As we think about the performance of the business going forward, how should we think about provision expense as a percentage of the loan book or net revenues? Is it -- it was -- the fourth quarter provision a bit of a catch up? Or is this something we should be viewing as more of a normalized level of provision for the future?

Yes. I think the best way to think about the provision for credit losses is to focus on the adjusted results. Where there is no provision for credit losses, that's the way that we think about it internally. And that provision, we talked about it in prior calls, we described why we think that makes it difficult to assess the economics of the business and there's lots that's been written on that, so I will just refer you to that.

We do say in the press release on Page 3, we talk about the impact of the revision on both the GAAP and adjusted yields. The impact on the GAAP yield was 90 basis points. On the adjusted yield, it was 140 basis points.

Our next question comes from Daniel Smith with Teton Capital.

Do you feel like loan durations have kind of reached a limit? Or is it just that you think that you should be paying like a little bit lower advanced rate against the long duration stuff?

No. I mean, the advance rate we paid during the quarter is the advance rate we thought was optimal. Otherwise, we write loan terms from 24 months out to 72. And the average initial loan term that's in the press release is just a function of the mix. So we're happy to write 72-month loans. We're happy to write 24-month loans. It's whatever the dealer and the customer desires.

So the Q4 -- the loans that you purchased in Q4 was that with the knowledge of the later collection timing?

It won't happen at the same time. We bought the loans during Q4, and we revised our timing during Q4. So there was some overlap there, but it was a period where the -- our knowledge was somewhat fluid.

Okay. So going forward, you don't feel any differently towards 72-month loans or potentially longer duration of the future? It's just that, all else equal, your advance rate should be a little lower and probably volumes a little bit lower?

Yes. Correct. We'll take what we learned from this -- from the revision and we'll apply that to future pricing.

Okay. But I mean going out -- I mean, looking out a couple of years, do you see duration continuing to increase? Or do you feel like it's kind of reached a limit?

I think it's impossible to forecast what the average duration would be 2 years from now. But like I said, we're happy to write 72-month loans. We're happy to write 24. And whatever the mix is just really depends on what the opportunities are in the market and what the dealers and customers desire.

But like in the next year or so, do you see yourself going out beyond 72 months on any loans? Or I don't know how you normally test these things.

Yes. We don't have any plans right now to go beyond 72.

Do you have like in the next year, do you have a plan to like increase the mix of like 72 months loans?

I think I covered that. I don't know how to answer it any other way than the way I just did. So let's move on.

Our next question comes from Clifford Sosin with CAS Investment Partners.

So -- sorry to be back on the timing of cash flows question. But my question is, I recognized that the change in your timing model adjusted the yield on an adjusted basis by 1.4% in the quarter. If I were able to look at just the forecasted levelized yield of new originations, by how much would this change in forecasted timing of longer duration loans impact that yield?

Pretty close to the same number.

Okay. So there's no sense that a loan that was originated a couple of years ago, where you've been collecting revenue at the old sort of -- or forecasting revenue the old schedule, so to speak, would have a larger variance if the change in timing was done only in the back half of the loan life?

You lost me there.

So if you originated a $100 loan 2 years ago and you've been recognizing revenue according to the levelized yield that you forecast at that time, and let's say things aren't exactly according to plan, there'd be some smaller principal balance outstanding today but the timing difference would then be added. And so in theory, the changes in timing would be sort of recognized over a shorter amount of remaining time and a smaller principal balance and so in theory...

I follow you, Cliff. So I think the piece there to fill in the blank is the old timing model, it adjusts over time. So as the loan ages, if our timing is off, the yield gradually adjusts over time. So it's not like everything just fell off a cliff. The loans were gradually adjusting over time. And what this revision does is sort of anticipate that going forward rather than letting it occur over time. So I think that's why it's a different (inaudible).

Got it. And then what -- roughly, when you talk about longer-term loans having a slightly longer duration and I realized that in all likelihood all the loans had some adjustment in the duration. But roughly, what portion of your balances are you thinking of as the longer-term portion which -- who's levelized yields are going to be lower and accounts the 1.4% overall change?

So the 72-month loans is where we have the least amount of data. 72-month loans are about 12% of our forecast at the end of '17. 66-month loans, we have quite a bit of data on those. We don't quite have a full 66 months of data, but we're close. And 66-month loans are about 26% of the overall forecast. So you have those 2 components together are where we have the largest changes.

(Operator Instructions) We do have a follow-up question from David Scharf with JMP Securities.

I think my question was asked in one fashion or another. I guess, Doug, I was trying to get it how to think about forecasting the level yield going forward. And on a GAAP basis, given the 90-basis point correction, I was just trying to get a sense for how much of that is arguably captured in the Q3 to the Q4 kind of simple average yield that's calculated? Or should we be taking the finance charge in Q4 divided by the average balance as a yield and then reducing that by a full 90 basis points going forward?

No. Take the actual yield for Q4 and use that as your starting point going forward.

Okay. So the allowance charge that hit the provision is captured in that full yield adjustment in the quarter. Got it.

With no further questions in the queue, I would now 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 ir@creditacceptance.com. We look forward to talking to you again next quarter. Thank you.

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