Bill Nash - CEO Tom Reedy - CFO Katharine Kenny - VP, IR

Sharon Zackfia - William Blair Brian Nagel - Oppenheimer Craig Kennison - Robert W. Baird Matthew Fassler - Goldman Sachs Mike Levin - Deutsche Bank Scot Ciccarelli - RBC Capital Markets John Murphy - Banc of America Merrill Lynch Michael Montani - Evercore ISI Rick Nelson - Stephens Inc. James Albertine - Consumer Edge Research Bill Armstrong - CL King & Associates Seth Basham - Wedbush Securities David Whiston - Morningstar Paresh Jain - Morgan Stanley Chris Bottiglieri - Wolfe Research

Good morning. My name is Victoria and I will be your conference operator today. At this time, I would like to welcome everyone to the CarMax Fiscal 2017 Third Quarter Earnings Conference Call. [Operator Instructions] Thank you. I would now like to turn the call over to Katharine Kenny, Vice President, Investor Relations.

Thank you, and good morning. Thank you all for joining our fiscal third quarter earnings conference call. On the call with me today is Bill Nash, our President and Chief Executive Officer, who by the way is very proud, along with Cliff Wood, and our many CarMax JMU alumni, of the Dukes, who are playing in the FCS National Championship. And of course with us is Tom Reedy, our Executive Vice President and CFO. Before we begin, let me remind you that our statements today regarding the company’s future business plans, prospects and financial performance are forward-looking statements that we make pursuant to the Safe Harbor provisions of the Private Securities Litigation Reform Act of 1995. These statements are based on management’s current knowledge and assumptions about future events that involve risks and uncertainties that could cause actual results to differ materially from our expectations. In providing projections and other forward-looking statements, the company disclaims any intent or obligation to update them. For additional information on important factors that could affect these expectations, please see the company’s Annual Report on Form 10-K for the fiscal year ended February 29, 2016 filed with the SEC. Please remember to ask only one question and a follow up before getting back in the queue. Thank you so much. I’ll turn the call over to Bill.

Good morning everyone. Katharine, I was not going to mention JMU since we’re a little bit more team agnostic now that Folliard has left, but since you did mention it, I’m excited about it and I wish them well. For today’s call, I will first review the key highlights of the quarter, then I’ll turn the call over to Tom Reedy to cover financing, and then I’ll close the call before Q&A with a brief update on some of our new initiatives. As you read in our press release this morning, our total used unit comps for the third quarter increased by 5.4% and total used units grew by 9.1%. Total used units were driven by strong improvement conversion as well as by a small increase in store traffic. We believe the increase in used units was due to a variety of factors. These include improved store execution, our website redesign and related online capabilities which have made it easier for our customers to submit leads. We continue to be very pleased with our core business as we again saw headwinds from tier three sales. The estimate used unit comps for our non-tier three customers were again significantly stronger this quarter at 9.8%. Our total web traffic was flat compared with the prior year quarter, while web leads performed well in the quarter. We continue to see a positive response to the website, with increased activity such as engagement with the search tool and car pages, as well as submitting leads. Our goal is to drive leads and convert more customers from the website to the store and finally to a sale. Our wholesale units declined by approximately 2% in the third quarter. This was driven by a decrease in appraisal traffic, partially offset by the growth in our store base and a higher buy rate. The decline in industry new vehicle sales during the recession is affecting the supply of older vehicles. Remember, we saw this dynamic in our core retail business following the recession. We continue to maintain a consistent gross profit per used unit at $21.55 compared to $21.60 in the third quarter of last year. Gross profit per wholesale unit decreased to $900 compared to $949 in last year’s third quarter. Again, as we reported in the second quarter, last year’s wholesale gross profit per unit was the highest recorded in any third quarter. So like last quarter, it was again a tough comparison. A few other topics before I turn the call over to Tom. As a percentage of our sales mix, zero to four year old vehicles fell this quarter to 785 compared to 81% last year, once again as a result of higher customer demand for older and less expensive vehicles. SG&A expenses for the third quarter increased 5.7% to $357 million. This growth partially reflects the 12% or 18 store increase in our base since the beginning of the third quarter of last year. While we reported leverage of approximately $74 per unit, remember that our advertising expense in last year’s third quarter was significantly higher due to the production and rollout of our new brand launch. Absent this increased spend in the prior year, leverage was modestly positive. Other overhead costs were up year-over-year, reflecting our continued investment in strategic initiatives. Now I’ll turn the call over to Tom.

Thanks Bill. Good morning everybody. In recent quarters we’ve discussed the trend of year-over-year increases in credit applications from customers at the higher end of the credit spectrum and less application volume at the lower end and we saw this continue in the third quarter. We recently purchased credit bureau data related to auto credit inquiries in the marketplace and our analysis of this data suggests that this trend is not unique to CarMax. This application mix growth in CAF’s net penetration which increased to 45% compared with 43.3% in last year’s third quarter. Net loans originated in the quarter rose more than 9% year-over-year to $1.3 billion due to a combination of CarMax sales growth and the higher penetration, partially offset by our lower average selling prices. In addition, as you would expect, the higher credit mix applications drove growth in the portion of sales where customers paid cash or brought their own financing. Tier two penetration fell year-over-year in the quarter to 15.7% from 16.9%. But just as we said last quarter, our tier two lenders continue to provide strong offers based on what we’ve observed in customer conversion. Tier three sales mix was 10.2% of used unit sales compared to 13.8% for the same period last year. This decrease in penetration continues to be due to the same factors, credit tightening early in the year and lower application volumes. Performance by our tier three partners however has been consistent since the middle of Q1. CAF income fell 3.2% to about $89 million compared to the third quarter of fiscal 2016, while average managed receivables grew by 11% to $10.3 billion. The provision for loan losses increased and the portfolio interest margin decreased very modestly. During the third quarter, we continue to see higher losses experience. Our ending allowance for loan losses at $115 million was 1.1% of any managed receivables compared to 1.08% last quarter and 0.97% in last year’s third quarter. This level of losses is a departure from our experience in recent years. However, losses in the last several years have been quite favorable. The current level of loan loss reserve is consistent with our range of expectations given our origination strategy and portfolio mix. For loans originated during the quarter, the weighted average contract rate charged to customers was flat year-over-year at 7.3%. Total interest margin declined slightly to 5.8% of average managed receivables. This compared to 6% in the third quarter of last year and 5.9% sequentially from Q2. During the quarter, we repurchased 3.8 million shares for $199 million. As of the end of Q3 we had $1.7 billion remaining in our stock repurchase authorization. Now I’ll turn the call back over to Bill.

Thanks Tom. During the third quarter, we opened six stores, including two in new markets Boise and Grand Rapids. We also opened four stores in existing markets, Philadelphia, Daytona, and two in San Francisco. Late in the current fourth quarter, we expect to open four more stores; two in our Los Angeles market, one in Murrieta will include a centralized production and auction facility to support our growth in Southern California. In addition, we plan to open two small format stores in the fourth quarter, both in new markets, Mobile, Alabama and Albany, New York. Next, I’d like to give an update on our new online offerings and the tests we are conducting to continue to advance the customer experience. First, as we discussed last quarter, we’ve been testing a new online financing capability that has helped our customers get prequalified for a loan. When we talked at the end of the second quarter, we were piloting this product in 13 stores. Towards the end of Q3, we successfully rolled this new capability out to all stores nationwide and we’re pleased with the early results, which includes increased leads. Second, we continue to test our home delivery offering in Charlotte, North Carolina. As a reminder, home delivery allows the customer to shop for and buy a car at their convenience without coming into the store. We continue to learn about the customer demands and operational scalability, while also ensuring it meets our standards for an exceptional experience. The last update is [regarding] a one store test we’ll be conducting in the fourth quarter. We plan to test a new digital solution for customers who are interested in getting an appraisal value for their vehicle by submitting their vehicle information online without having to come into the store. These initiatives position CarMax to deliver more of the car buying experience online and we remain committed to giving the customers the ability to go between the digital and in-store experience whenever and however they would like. While we are constantly working to innovate and improve the business, we are also very focused on being cost conscious and eliminating waste across the entire organization. We continue to look for new opportunities to ensure we are growing in a way that is in the best interest of our customers, associates and shareholders. Like last quarter, I remain confident that we’ll continue to lead the industry. We have 23 years of experience to build upon to help us excel into the future. With our great associates, our national footprint, our unparalleled inventory, brand strength and continued focus on improving the customer experience, both online and in-store, no one is in a better position than CarMax to fully deliver the best car buying experience. With that I will open up the call for questions, so Victoria?

[Operator instructions]. Your first question comes from the line of Sharon Zackfia with William Blair

Hi, good morning. A question on the search engine optimization you’ve been working on, I think, earnestly this year. I guess I'm a little surprised that web traffic is flattish, given that. So, if you could give us an update on where you stand on search engine optimization and how much of a driver that might really be going forward.

Sure. Thanks Sharon. So SEO is a big focus for us. The things that we’ve done so far this year to really focus on that, first of all the website which we’ve talked in the past. In addition to the website, recently we’ve been working on some of the design of the website to make sure that search engines can get the information in a quick, time sensitive fashion so that they can pull all the relevant data down. So that’s something that we’ve been working on here more recently. We’ve also been publishing some content that our customers want and expect, but this is an area now that we’ve got the website up, we’ve made some tweaks to it. This is an area that we think still has a lot of opportunity going forward.

Your next question comes from the line of Brian Nagel with Oppenheimer

Good morning. Nice quarter. My question is on used car unit comps, two parts. I'm going to shove it together. First off, a number of retailers really across the board have discussed some sales disruptions, like the change around the recent presidential election. So I know it's not typical for CarMax to talk about intra-quarter trends, but I thought maybe you could discuss that. And then the second question also relating to car sales, you talked a lot in the prepared comments about some of the internal initiatives. We've clearly seen a sequential improvement in used car unit comps over the past few quarters. From a market standpoint, has CarMax started to benefit yet from the supply of vehicles improving within the industry? Thank you

Yes, Brian. So on the first question, I really don’t have a comment on that and the Trump effect. On the second part of the question, as it relates to the supply, I absolutely think that we are starting to see the supply come back in the auctions and I think that’s a good thing for CarMax. As that supply comes back, it helps to drive the difference in price between a new car and a late model used car. So while I didn’t highlight that as one of the factors in the prepared remarks, I absolutely think that that’s also a contributing factor.

Thank you.

Your next question comes from the line of Craig Kennison with Baird.

Good morning. Thanks a lot for taking my question. It's on the wholesale business, which fell a little bit. Do you think you are seeing your fair share of appraisal traffic or are sellers starting to use some of these alternate channels, which, again, you may have access to with your new strategy? And then on a related note, how does the increase in lease activity, we've got a lot more leased cars today, impact your opportunity for appraisal given those lessees are not selling their car. Thanks.

Okay. On the wholesale piece, I think that’s more indicative of what I talked about in the early remarks, that as the new car SAAR works its way through from the recession, the shortage of new cars, for us that’s right into seven to nine year old vehicles and you know that’s right in the sweet spot of wholesale. And we saw the same phenomenon earlier on when those cars were zero to four. We saw less of them coming through as appraisal traffic we expect and we’re seeing the same thing come through in the seven to nine year old vehicles. So we think that’s more of a factor of what’s at play there. And I’m sorry the second question, what was it in regards to lease activity?

With the increase in overall lease activity, there are a lot of people who have cars that are leased and not therefore for sale. Does that impact your ability in the wholesale business to attract those cars?

No, I don’t think so at this point because first of all the leases are just starting to come back at this point and our wholesale cars are generally older. I do think that as the leases come back, that’s a good thing for the core business because that’s a lot more selection of younger vehicles. But I don’t think that’s having a big impact on the wholesale business.

Thanks.

Your next question comes from the line of Matt Fassler with Goldman Sachs

Thanks a lot and good morning. Within the SG&A, other than the advertising, which was down for reasons that you stated, your compensation expense was also quite well controlled. It looks like on an aggregate basis it was up only 3% year on year, this comp and benefits, and up really, barely up on a two-year basis in aggregate dollars despite a big increase in stores. Was there anything related to stock comp, which I know you have sometimes called out, or any other factor, incentive comp, that would have led this number to be so subdued here in the third quarter?

Hey Matt. Yes, stock comp actually this quarter is the first time it’s been at a level so small that it’s not worth talking about on a year-over-year basis. But there are some good guys embedded in that comp and benefit number, both some unfavorability last year and some unfavorability this year. But despite that, we still did leverage that line relative to growth in unit sales. You could probably double the percentage increase and it would be a rough idea of what we -- kind of run rate it would be.

So, in other words, the run rate would have been something like 6% in dollars, normalized for everything, which still would have led you to good leverage.

Yes, that’s fair to say.

Got it. That's very helpful. I'll count that as my follow-up and come back if we have more. Thank you.

Your next question comes from the line of Mike Levin with Deutsche Bank.

Morning everybody. Just wanted to focus on the loss provision rates at CAF for a second. Is there a way that you can talk about, is this factoring in just what you are currently experiencing or does this take into account additional expected declines in used vehicle values moving forward? If it's just now, how much do you think this could continue to move down with used values coming down further from here?

Let me hit on a couple of things. One is used values definitely has a impact on loss rates, but historically customer behavior has been a much bigger driver and on an ongoing basis there’s many things moving. So it’s hard to say that that one individual factor could have a dramatic impact on a go-forward basis. As to the determination, every quarter we determine based on what we know we to date rather than based on any kind of speculation, a 12 month look-forward on loss expectation. So as I said, it’s based on historical information that we have and then how we are performing to date relative to that historical performance. So in a quarter where things are moving, you’re going to see us move that loss expectation. We have not in the past when things were going good taken a view to try to get ahead of things and we haven’t done that here as well.

Got it. Maybe to put that in a little better perspective, these current rates versus history, I know the accounting was different at the time, but could you compare where these loss rates are versus where we were, say, through the crisis?

Yeah, I think well through the crisis, we typically try to manage our portfolio to kind of a 2 to 2.5% cumulative net loss rate. And the average life for these loans has been roughly two years as a proxy. So during the recession, we saw loss rates go to double that in the portfolio. But one thing I think I’d point out, after one point, if you look at the size of the loan loss allowance today which is a 12 month look, it’s 1.10%. So if you double that, you can see they were still well within the range of expectation in the zone that we are looking to originate. And something to remember also is that our tier three CAF is also part of that overall loan loss provision and represents almost 10% of it.

10% of the 1.1%?

10% of the dollar volume, the $114 million, $115 million of loan loss allowance.

That’s really helpful. Thank you.

So when we step back and look at this, yes we’ve seen deterioration relative to recent years, but still not to the point where we were pre-recession.

Got it. Thanks.

Your next question comes from the line of Scot Ciccarelli with RBC Capital Markets.

Good morning, guys. You guys have always talked about tier 3 sales as really being incremental to the core run rate. So my question is, when CAF increases penetration, do you have an estimate what percent of those CAF sales are incremental? So in other words, is CAF financing some people that wouldn't have received credit elsewhere?

I think the answer to that is yes and not necessary because we are more aggressive than other folks, but every lender has a way that they look at credit. We think we have ways to look at our applicants and find pockets that work for us that may convert at a better rate because they got a CAF offer than if they get a tier 2 or tier 3 offer. So if you look at the conversion of customers based on a CAF offer versus based on a tier 2 or 3 offer, it’s much more significant. So you would expect that some of the CAF offers would be incremental just because of that phenomenon.

But, Tom, if you had, let's call it 7,600 extra unit sales because of the increase in CAF penetration, is there a percentage you would suggest that was incremental based on the way you guys look at your models?

Like I said, there would be some incremental there, but we haven’t shared anything as far as fractions. I can tell you that that percentage would not be expected to be up or down relative to recent years.

Right. Understood. And then related to that, because it has to do with on the credit side, given the rise in interest rates that we’ve seen, I would assume that your funding costs have actually increased. Just how should we generally think about how you guys plan to manage the spread versus moving interest rates too quickly on customers with the potential for negative unit elasticity? Thanks.

Well the nice thing about our business is we can test quickly and we can test pretty accurately in the CAF space. So we have seen an increase in benchmarks in recent weeks I guess if you look at it. And we have taken action to raise rates in certain pockets selectively to offset it to some extent. Now what we’ll need to do is as we look at it going forward to see if the market will allow us to keep those rates up because I’ve said many times we’re going to be a market lender. We’ve got to provide competitive rate to the customers so nobody is getting sour on CarMax over the finance offer they got. And we’ll closely monitor our three day payoffs. We’ll closely monitor conversion as we ramp up rates and we’ll do what the market allows us. But we are - this is something that is just ongoing business as usual for CarMax Auto Finance. We are constantly testing ways to maximize the profit and sales coming from that portfolio.

Okay. Got it. I'll follow-up later. Thank you.

Your next question comes from the line of John Murphy with Banc of America Merrill Lynch

Good morning, guys. Just a question on supply and the impact potentially on grosses. It seems like there's a lot of the new vehicle dealers that are getting off the stop sale wagon and retailing a lot of used vehicles they had stored on their lots in the short run. So, I'm just curious what you think about supply there. And then, also, as you look at this leasing bubble that's coming over the next three years, where you think we are in that increase in supply. And as you think about those two legs of supply increases, what you think you can do as far as holding your grosses. Are you going to stay committed to this $2,100 to $2,200 range as we see that increase in supply really pick up?

Yes, John. So on the first part of the question about the stop sales, I don’t really think that’s much of an impact. Lots of dealers have been selling vehicles that have open recalls. While you’ve got the franchise dealers that cannot - by the manufacturer standard, you can’t sell it for that franchise, they still sell other non-brand franchise at those stores. So for me that’s not really - I don’t really see that as either a positive or a negative. I think it’s just continuing on. As far as the lease thing, I said earlier we’re seeing more vehicles come back from the leasing channels and we think it will continue through next year and even into the year after that. And I think we’ve proven over time that we’ve been able to hold our margin throughout many different cycles. Now the only caveat that I would give there is you know we constantly do testing and if we think that moving that margin down would more than pay for itself in units, then we would look at that. And so we will continue to test, but we feel confident that we can continue to maintain those gross profit margins.

And if I could just follow up, we simply think that if you dropped your gross by 10% you would expect to get at least a 10% pick up in volume or else you wouldn't do it. Is that correct? Or 11% increase in volume to offset that drop in gross? How would you think about that?

Yes, we would want to see total more than offset the drop than what we see. So it would have to be a win on total gross.

Okay. Great. Thank you.

Your next question comes from the line of Mike Montani with Evercore.

Hi guys. Good morning. Thanks for taking the question. Just wanted to ask if I could and apologize if I missed it, but what are you seeing now in terms of SUV truck penetration? I think last year it was 23% by your measures and I was looking for it to be maybe low 30s. Could you talk about that?

Yeah. So our large SUV mix for the quarter was - excuse me. Our large SUV mix for the quarter was up year-over-year. It’s about 25.7% versus last year’s third quarter was 23.3% so it’s about 250 basis points. It’s up about a point from the second quarter of this year. So the mix is up a little bit, which would inherently drive our ASPs up a little bit, but then because we had a shift in some of the older vehicles, that brought the ASPs down as well acquisition process. Lower acquisition prices brought it down. So SUV mix was up a little bit.

Okay. That’s helpful. Can you talk a little bit about SAAR setting and how its trending in terms of buys that you would do and appraisal weighing versus auction in other areas? Just thinking about big picture, it would seem that you would be in a better position to sort and merchandise given that acquisition cost has come down. I just wanted to see how that’s playing out.

Is this in particular to the large SUVs or just in general?

No, just in general.

Yeah. I think what we are seeing out there is what we typically would see this time of year with fall depreciation. And I think with the fall depreciation obviously is going to lower the acquisition prices. Again, we feel like that’s a good thing for us because as the acquisition prices go down, we just can pass those lower prices onto the customers and there is a bigger delta between new cars and late model used cars, so we think it’s good.

Thank you.

Your next question comes from the line of Rick Nelson with Stephens.

Thanks. Good morning. I wanted to ask about third-party lead generators, where you stand with those. And did you do anything different this quarter with the lead generators?

Yes. We have all of our inventory listed on - for the quarter it was on Car Gurus and that was the only site that we had our inventory listed on during the third quarter and we’ll constantly test some different things, but that was the only one during the quarter.

Okay. Thank you for that.

Your next question comes from the line of James Albertine with Consumer Edge

Great. Thanks for taking the question and good morning, everyone. I wanted to ask, and I know it's early, but certainly we're getting a lot of questions. We hosted a call yesterday on the subject matter in fact. Wanted to understand your thinking and maybe what you're hearing from your advisors as it relates to tax planning for next year and maybe some of the pushes and pulls as you think about the potential deductibility of CapEx or PP&E and maybe the reversal of deducting interest expense, and so forth, as well as the reduction in corporate tax rate. So just any strategy thoughts there would be helpful in helping us frame how you guys are thinking about it relative to peers, even. Thanks.

Well James, since you had a call yesterday, maybe you should brief us on it because you probably have more information that we do. We really don’t have any thoughts at this point. It’s all so new and there’s so much up in the air. It’s just it’s too early to really talk about.

Okay, great. Thanks. I'll get back in queue.

Your next question comes from the line of Bill Armstrong with CL King & Associates.

Good morning, everyone. So you mentioned a small increase in overall traffic into the stores, but you also had a reduction in appraisal traffic. Are you seeing any change in consumer behavior in terms of fewer people coming in with trade-ins? What do you think explains for that apparent disparity?

I think like you said, we saw a small increase in store traffic and I talked a little bit about store traffic last quarter in that you really can’t look at just store traffic by itself. You really have to look at web traffic and especially for us, leads and store traffic to understand the overall big picture. Consumers are a lot more educated. We know that of the consumers that buy vehicles from us, nine out of 10 of them have already done a lot of research on CarMax.com. So they’re already much better educated. I think on the difference between getting a little bit more store traffic in and get appraisal traffic down. Again, a lot of it is tied to what we’re seeing as the SAAR rate works its way through the bubble. And I think that’s a main driver of why you see some of the appraisal traffic down.

Because they are getting less money for the …

Just because there aren’t as many. So like I talked about earlier, the seven to nine year old vehicles, there just aren’t as many as there were because seven to nine years ago the new car sales rates were going down. So if there aren’t as many customers that have those cars, they’re obviously not coming in to get them appraised.

Got it. Thank you.

Your next question comes from the line of Seth Basham with Wedbush Securities.

Thanks a lot and good morning. You guys called out increase in the mix of sales to higher credit customers and their supply of late-model vehicles, but at the same time you are selling a higher mix of older vehicles. Can you help square away that dynamic, which has occurred for the last two quarters now?

Yes. I think contrary to what a lot of people think is that your tier 3 customers are going to be the ones buying your late model or your older, less expensive vehicles and for us, that’s just not the case. Majority of the vehicles that we sell of the older, less expensive vehicles are to non-tier 3 customers. So I think what it says is that the consumers are still out there looking for a great deal and they’ve opted to go a little bit older and a little bit cheaper. So I think that explains the dynamic. It’s not just tier 3 customers that would be buying older, less expensive vehicles.

Okay, that's helpful. And just as a follow-up, considering the increased supply of late-model vehicles, would you expect your mix to shift back to that direction over the next few quarters?

Seth, we’ll sell whatever the consumers want. So if the consumers want that late model vehicle, then we’ll absolutely adjust our inventory accordingly. If they want to continue to buy older vehicles, then you won’t see that change. So it just depends. It’s all driven by consumer demand.

Thank you.

Your next question comes from the line of David Whiston with Morningstar.

Thanks. Good morning. Just wanted to talk about the future in e-commerce with the pilot programs you are doing. Obviously there are consumers who want this process to be very easy, but at the same time it's still roughly a $20,000 purchase for people. So how willing do you think people are to buy a vehicle without even coming into the store to do a test drive? Is it a small portion of the American public? Is it ultimately going to be a lot of people? I'd love to hear your thoughts on that.

It’s a good question, David and I think you hit the nail on the head. It’s not an easy transaction. It’s a complicated transaction. That being said, I do think there’s a small subset of customers that want to try that experience. I think for us, we certainly are going to meet those customers wherever and however they want to do business, but I think more importantly is that I think consumers will want to do more of the transaction online in the near term and we will have the ability to not only do more of the transaction online, but it will be a seamless integration into the store. And I think that’s really what we’re focused on. In addition to giving the customers the ability to do everything, it really is okay, you want to do part of it online. Let’s make sure it’s a great experience and then when you walk into the store we just pick up from where you left off and that’s really what we’re focused on. I think that’s really where the consumer demand is going to be in the near term.

Okay, that's helpful. And for my follow-up, I wanted to ask a comment about buybacks. Stock has been near a 52-week high. Are those still full steam ahead?

Pardon me. I didn’t hear the tail end of your question.

The stock is near a 52-week high. Are buybacks still full steam ahead or are you looking to be a bit more cautious on that going forward?

As we’ve discussed before, we take an enlightened averaging in methodology to repurchases. So at any given time, we’re trying to average into the market and maintain our capital structure in the zone that we’ve targeted. You saw that in this quarter we bought back $199 million versus $125 million in Q2. That did represent a little bit of a step up in aggressiveness because we saw some weakness in the stock price, but the parameters we put in place allow for a little bit of that, allows us to be a little more aggressive when the price is weak and a little more conservative when the volume and price is looking strong. But we intend to be in the market on a regular basis.

Okay. Thank you.

Your next question comes from the line of Paresh Jain with Morgan Stanley.

Morning everyone. Just a quick question on inventory resourcing. Would it be fair to say that you currently source about 40% of your inventory from auction? How has that number changed in the last two years, if that's changed at all?

So when we think about self-sufficiency which is opposite of the auctions, but actually getting it through our own appraisal and we’ve given the range of 40% to 50%, we’re at the lower end of that range, but we’re higher than we were a year ago and we’re very similar to where we are last quarter.

And a follow-up on the profitability of the vehicles sourced from auction, has the gap in profitability with those sourced through trade-ins changed at all in the last two years? I think the gap is about hundreds of dollars between the two.

So the vehicles that we source through our appraisaling are still more profitable than the ones that we source offsite. As you can imagine, you don’t have transportation fees. You don’t have folks going out and incurring expenses, that kind of thing. So there’s still about the same amount of profitability over what they were before.

And no change in the last two years in that gap?

I can’t speak really off the top of my head on the last two years, but it’s been very similar.

Okay. Thank you.

Your next question comes from the line of Chris Bottiglieri with Wolfe Research.

Hi, thanks for taking my question. Did you guys give out the zero to four mix this quarter, like you historically have, and how that compared versus last year?

Zero to four we were down slightly 81 to 78.

Got you. Okay. And then just a follow-up to that, given the scarcity of those cars given supply, I'm wondering if you're holding back more of those vehicles from wholesale to sell it used, if the two are related at all.

Actually I think that supply is starting to come back on those so we’re seeing more out there. But as far as the selectivity, we’re always looking at selectivity when we purchase a car, whether we designate it wholesale or we put it through. I would say at this point when you’re talking about zero to four year old cars, we really haven’t changed anything recently as far as selectivity goes.

I'm sorry. I meant holding back the seven to nines where you had tighter supply. Have you held back some of those from wholesale to retail at this point?

I don’t think it’s considerably different than what we’ve been doing before. It really depends on the vehicles as they roll through. If they’re good quality vehicles and we can turn them into CarMax cars, then we’re going to do that. But we look at selectivity every single quarter and make the decisions that are appropriate at that point. I wouldn’t say that this one is necessarily - there hasn’t been a big change in this quarter versus other quarters.

Okay, great. Then one quick follow-up. Two questions have been asked on auction supply. It seems to me just looking at the auction data that there seems to be a mix of improvement in SUV and truck availability at the auction lanes. Has that been a driver for your business in terms of increased SUV and truck penetration?

I think it’s, when you see that increased inventory, that will drive the prices down and that probably has a factor. It is a factor into where the customers want to buy the car, depending on how attractive it is from a price standpoint. But again, we really go off of customer demand and if they’re looking for large SUVs and we can get them at prices that they’ll pay, then we’ll put them on our front lot. Just this quarter folks were looking more for older, cheaper cars and so that’s what we did.

Okay, that's helpful. Thank you for the commentary.

We do have a follow up from the line of Matt Fassler with Goldman Sachs.

Thanks a lot. Good morning again. If we were to be able to accurately calculate how new space productivity looked versus a year ago just in terms of raw numbers, what do you think we would see? And how did new store performance compare to your expectations?

We’re pleased with new store performance. As we’ve talked about in the past, it’s difficult to talk about store productivity when you look at the size of stores that we’re building, when you throw in the small formats, when you look at going into new metro markets where awareness is at different levels. You’re going to have different ramping up. All the new stores are reaching profitability quicker than the old stores because of the GPU and they’re also meeting all of our internal rate of return. So we’re pleased with new store productivity.

Got it. And then, secondly, I know you are far more focused on your customers than on your competition, but we did see one piece of competitive fallout with one of the higher-profile, albeit small, online startups essentially ceasing operations in its prior form and merging with another company. What’s your sense about the competitive environment online? Is it intensifying? Are you starting to see any signs of more rationalization in terms of the kind of offers that your customers can pursue away from CarMax?

I think first of all just a comment on your first part about the online competitor ceasing operation. I think it goes back to something I talked about earlier which this is not a simple business. This is a hard business. It’s not easy to go out, source inventory for the right amount. It’s not easy to go out and put a cash offer on every single car. And I think that’s just an indication of why some of these folks are struggling. I think competition is hot in both the online space and the traditional and I think our challenge is to make sure that we continue to leverage the equities that have made it successful and them even better and leverage our data and our processes and our systems and our people to continue to widen the gap between us and our competition. And like I said earlier, we’ve got 23 years’ worth of experience and 23 years’ worth of data and we’re leveraging that data and tapping into it in ways that we haven’t been able to and haven’t done in the past. So although the competition is hot and heavy, we feel really good about the positon that we’re in and I think ultimately to be as successful as we can be, we do need to make sure that we can be in the online space, but equally important is we’ve got to be in the brick and mortar space. I think you have to do both very, very well and I think we’re positioning ourselves to do just that.

Got it. Thanks so much.

Your next follow up comes from the line of Scot Ciccarelli with RBC Capital Markets.

Hey, guys. I guess this is a bit of a follow up on Matt's question. But I guess what I'm trying to figure out is, on some of these delivery tests, are there any key differences to what you are trying to do versus the guys like Vroom and Beepi? And also, how is it that you are allowed to close a transaction outside of the dealership? Because I guess I was always under the impression just legally you had to close the transaction at a dealership. Thanks?

So on the second part of your question, there are some states where you can’t close it outside of the dealership. Our delivery, where we’re doing our tests, that’s obviously not the case. I think as far as differences, one of the difference in our home delivery test is you don’t have to buy the vehicle before we bring it out to you. We’ll bring it to you. You can test drive it at your home and if you decide to purchase it, we can do the transaction right there. That’s different than any other tests that are out there where they make you purchase the vehicle before it’s brought out to you. Again, it’s a small test. We’re trying to understand exactly what the customers want and the good thing is we can adjust the test as we move forward.

Got it. Okay. Thanks, guys.

Hey Scott, I think Tom was just saying, he thinks maybe one online competitors might allow you to …

Yes. There’s one that allows you to test drive, but they also make you give up your car as they’re trying to sell it.

That’s right.

I'm sorry, can you repeat that? It was broken up.

There is one competitor that allows you to test drive as a purchaser, but they’re a C2C enabler and the other part of their business offer is that they take possession of your car while you’re attempting to sell it so you don’t have use of your vehicle.

Yes. So there’s nobody doing it exactly the way we are from a business to consumer, but to Tom’s point, there’s a C2C that will allow you to test drive without purchasing it beforehand.

I think it’s important. I think when we’re looking at this, we’re trying to make sure we’re not doing anything that inhibits the customer from wanting to transact with us.

Yes.

Got you. Thanks, gentlemen.

Your next follow up comes from the line of James Albertine with Consumer Edge.

Great. Thanks for taking the follow up here. And then on the same theme, on the online versus your traditional sourcing methods, I was curious, from a qualitative perspective, given you are now pretty far along in terms of your investments digitally, how do you think about operating margin? Are there any differences in the final operating income contribution whether you have sourced the customer through an online channel versus a traditional mechanism? And if you can maybe at a high level talk about what are the differences and pushes and pulls from an expense perspective that we should think about that drive that delta, if there is one?

So James, I think it’s a little early to be talking about the differences. While we’re committed to expanding our online capability, we really we have such a small test going on home delivery and outside of that, all the other online capabilities customers are utilizing them and then transitioning into the store. So it’s a little early to talk about the economics of one versus the other and the delta.

Okay. Then maybe a quick follow-up. You’ve done a great job of laying out milestones for us historically to think about pilot programs. Are there any additional measures that we should look for in terms of calendar 2017? Any other major investments coming or looming that you thought about and are willing to articulate, at least from a qualitative perspective?

No, I think the one that I highlight was one that we were going to be doing in the fourth which is the online appraisals and then in the fourth quarter, we’ll highlight any new ones for the upcoming year.

Okay. Great. Thanks again.

You do have another follow up question from the line of Michael Montani with Evercore.

Hey, guys, I just wanted to follow up on two things. One is for Tom, if you could just quantify what recovery rates were like in the quarter and contextualize that for us. And then, secondly, what are you seeing in terms of credit availability from some of your third-party lenders? If memory serves, I think it was last March or April that you saw a noteworthy pullback from one of the largest ones.

So as far as recovery rates, it’s down a couple of points year over year in the quarter, but as I mentioned, consumer behavior in payments and recovery, them recovering themselves out of trouble tends to be more impactful. And I’m sorry, I didn’t quite get the tail end of your last question.

I guess the last question was just, you did the 10% comp if you back out third-party subprime. I guess the question is, do you start to cycle the pullback in March, April, if memory serves? And then, also, what are you seeing in terms of lender behavior from third parties? Does it appear stable at this level or is there further pullback that you can see now?

Sure. As far as cycling, I think in the current period right now we’ll begin to cycle over the trend we began observing in traffic. So remember we’ve been talking for several quarters about seeing fewer application volume at the lower end of the credit spectrum. That started somewhere around this time last year. So that portion of the decrease we’ll start to see some cycling. As I mentioned in my comments, the most recent tightening we saw was in mid Q1 from one of our tier 3 lenders. Since that time they have been behaving very consistently. If we look at conversion of customers in the tier 3 space, since the middle of Q1 it’s been very consistent. It’s below last year, but it’s been consistent. So yes, all else equal, in April you start to see us rolling over that on a year over year basis. In the tier 2 space, it’s almost exactly the opposite phenomenon. Conversion has been consistently better than last year since the first quarter and as I mentioned, they’re doing a great job of converting the volume that they see. A lot of the mix they’re now foreseeing, a lot of it is from the customer traffic coming through the door. That’s why CAF is up. That’s why other is up and that’s why that tier 2 and 3 are down.

So, are you guys happy with the third-party subprime where it's at now at around 10%, or could we consider new lenders that might help to grow that again?

I think it’s going to be a matter of - obviously it’s going to be a matter what customers come through. We’re happy today with the amount we’re investing in the tier 3 space, which is about $1,000 a car of margin that we give away in order to enable those transactions. The tier 3 space is a lot more fragmented than the more prime space and other lending. We’re happy with the performance of our partners. We test other partners occasionally and to the extent one can add value, we consider bringing them in the mix. But I think if you step back and think about our current base as a platform for growing, it’s a much more profitable dynamic to grow at this mix than to have the mix migrate back to something higher.

Thank you.

Your next follow up comes from the line of Seth Basham with Wedbush Securities.

Thanks for taking the follow-up question. Just thinking about the fact that most of your customers are shopping online before visiting the stores, as you consider the store model of the future what kind of changes are you thinking about? We notice, for example, in the Seattle area you're opening a pretty large megastore, which is likely a reconditioning focused facility in a lower cost area. Is that something you'll consider in other markets too?

Yes. I think what you’re seeing in Seattle is more driven by the market dynamics and what’s available land wise. Obviously we’re looking at our store prototypes. We’ve got the small format store out there. You’ve seen we’ve opened more of those here more recently. I think it’s something that we’ll stay close to as we better understand the customer desire for transacting more and more parts of the transaction online. But I tell you, it’s not like we have 1,000 stores around the country. We have 169 stores and I think they’re all strategically located and the product we sell, you need space and you need places to build those cars. So I feel very comfortable with our growth strategy, what we’ve done so far and I feel very comfortable with what we’ve got in the pipeline currently.

All right. Very good. Thank you.

Your next follow up comes from the line of Mike Levin with Deutsche Bank.

Thanks, guys. I wanted to follow up on your comments about increasing applications at the high end of the credit spectrum. And you said you also bought some market data that you were taking a look at. Just wondering if you could maybe put a few numbers behind what you are seeing at CarMax, how that compares to what you are seeing within the market. Are you outperforming those increases? And then how we might think about higher credits being a larger percentage of your mix impacting loss rates going forward.

I’ll start with the last one. We’ll see how it impacts loss rates going forward depending on the environment. It’s way too early to tell. We’re not in a positon to give you any numbers regarding this, but I can give you a little color around what we looked at. We bought data related to auto inquiries, meaning folks that are hitting the bureaus in conjunction with an application for credit for an automobile for the past several years, both through submissions to captives and submissions or inquiries through the major dealer groups. And while we likely started feeling the phenomenon a little earlier, the data is telling us that the lower and lower middle part of the credit spectrum has seen a year over year decline in recent quarters for everybody. As far as the higher credit folks and the growth there, our data appears to show that we are growing a little bit stronger than the rest of them, but there could be a lot of factors between new car mix, used car mix et cetera that drives that. I don’t know if there’s anything really to take away from that.

Okay. I appreciate the color.

That concludes the Q&A portion of the call. I would now like to turn the conference back over to the presenters for any closing remarks.

Thanks, Victoria. In closing, I just want to thank you for joining us today. I want to thank you for your continued support. I also want to thank our 23,000 associates that are working for us across the country. Our success is absolutely due to them and the exceptional customer service they provide on a daily basis. And I’m really proud of everything they’ve accomplished this year. I wish everyone a wonderful holiday and happy new year and we will talk to you next quarter. Thank you.

Again, thank you for your participation. This concludes today's call. You may now disconnect.