Carvana Co. (CVNA) Q2 2017 Earnings Report, Transcript and Summary

Good afternoon, and welcome to the Carvana Second Quarter Earnings Conference Call. [Operator Instructions] Please note, this event is being recorded. I would now like to turn the conference over to Alex Wellins from The Blueshirt Group. Please go ahead.

Thank you, operator. Good afternoon, ladies and gentlemen. Thank you for joining us on Carvana's Second Quarter 2017 Earnings Conference Call. Please note that this call will be simultaneously webcast on the Investor Relations section of the company's corporate website at www.investors.carvana.com. The second quarter shareholder letter is also posted on the IR website. Joining me on the call today are Ernie Garcia, Chief Executive Officer; and Mark Jenkins, Chief Financial Officer. Before we start, I'd like to remind you that the following discussion contains forward-looking statements within the meaning of the federal securities laws, including, but not limited to, Carvana's market opportunities and future financial results that involve risks and uncertainties that may cause actual results to differ materially from those discussed here. Today's discussion may include forward-looking statements within the meaning of the federal securities laws, which are subject to risks and uncertainties that may cause our actual results to differ materially from such statements. A detailed discussion of the material factors that cause actual results to differ from forward-looking statements can be found in the Risk Factors section of Carvana's prospectus related to its initial public offering filed pursuant to Rule 424(b) under the Securities Act with the SEC on April 28, 2017. The forward-looking statements and risks in this conference call are based on current expectations as of today, and Carvana assumes no obligation to update or revise them whether as a result of new developments or otherwise. And now with that said, I'd like to turn the call over to Ernie Garcia. Ernie?

Thanks, Alex, and welcome, everyone, to our second quarter conference call. Hopefully, you've all had the chance to look over our shareholder letter, which is available on our Investor Relations site. As I go through my prepared remarks, I will reference some graphs and figures in the letter, so it may be helpful to have that with you. We are extremely excited about the results we had in the quarter and about the trajectory of the business in general. As we laid out in our Q1 call, we are focused on 3 key objectives for the business: growing retail units and revenue, increasing total gross profit per unit and demonstrating operating leverage. First, let's hit objective #1, growing retail units and revenue. In the quarter, we grew unit sales 145% year-over-year, added 7 markets, our fastest quarterly market opening rate to date, and opened one vending machine in Dallas, Texas. This accelerating growth was driven by the market share ramp trends we have seen historically. Our most mature markets continue to grow in market share. New markets continue to ramp sales faster on average than older markets, and vending machines continued to contribute meaningfully to our market share ramp when opened. In addition, we upsized and extended our inventory line with Ally Bank on August 4 to support our anticipated inventory and sales growth over the next year. Ally has been a great partner for us for a long time and we are excited to be extending that partnership further. Next, we'll hit objective #2, increasing total gross profit. We continue to make significant progress toward our midterm goal of $3,000 of total GPU in the second quarter, increasing total GPU sequentially from Q1 by over $300. Importantly, this improvement was achieved despite the headwind provided by average days to sale increasing room 93 to 105 days in the quarter, as anticipated and described in our Q1 call. Days to sale is generally driven by 2 factors, the speed at which we are selling cars relative to our inventory size; and the timing of when we purchase the cars that are in our inventory. In this case, I think it is important to decompose those factors, so let's hit sales relative to inventory first. As you can see in the graph on Page 7 of our shareholder letter, inventory has been relatively flat since we've hit an inflection point in the fourth quarter of 2016, and over that 6-month period, we have grown sales by about 90%. There's a clear evidence that we are able to grow sales significantly without growing inventory by adding new markets, allowing younger markets to catch up in market share to older markets and continuing to grow market share in our more mature markets. We think this is a really big deal because it makes clear the leverage inherent in our pooled inventory model that will power our decrease in average days to sale over the next several quarters, and because this improvement in average days to sale is the single biggest contributor to our walk from $1,500 in total GPU today to our midterm goal of $3,000. The remaining effect that drives turn time is when sold cars were purchased. This can be seen in the same graph in Page 7 of the shareholder letter. We built up significant inventory in the fourth quarter of 2016 in anticipation of the sales growth in the first half of the year and have held inventory relatively flat since then. This vehicle acquisition timing was a significant driver of average days to sale going up in the second quarter as many of the cars we sold were purchased in that fourth quarter inventory build period. We believe this is good news and that the first effect is the bigger driver over time, and we are expecting average days to sales to start to reduce from here, providing a significant tailwind toward our midterm total GPU goal. As we did in Q4 last year, we do expect to start growing our inventory in the fourth quarter when prices are historically lower and in anticipation of the significant sales increases we expect from the first half of 2018. That said, this increase will take place at a rate that allows us to keep reducing our average days to sale in the fourth quarter and into 2018. During the second quarter, we acquired and joined forces with Carlypso, a fellow automotive disruptor. Carlypso's business had been focused on making vehicles at auction available to customers by merchandising and listing those cars on its site prior to purchasing the cars. This focus caused them to build the tool set that is powerful and highly complementary to our own. Getting high-quality vehicle data describing which features, options and packages are on any given car and valuing those features and options is a notoriously difficult problem in the automotive industry. Carlypso built a number of tools that collect and clean vehicle data from many disparate resources and then analyze that data to properly understand the value of each car. The uses for us of these tools are clear and take many forms: automated bidding optimization and vehicle acquisitions across channels including direct-from-consumer, merchandising inventory with better vehicle descriptions and price optimization of vehicles available for sale on our site. Many of these integrations are already well underway and we expect some of the early benefits to begin to show later this quarter, with more meaningful benefits rolling out over the next 12 months. Now on to objective #3, demonstrating operating leverage. This is another area where we made significant progress, cutting our EBITDA margin losses by 25% in the quarter sequentially. This number did benefit a bit from the timing of certain expenses being pushed into Q3. But even when taking that into account, the progress significantly exceeded our own expectations. This improvement resulted from leverage showing up in many areas of the business and from a lot of hard work from our operations and technical teams that deserve special recognition. I'd also like to touch on our national marketing initiative. As discussed in our previous call, we have now reached the point that several cable channels are less expensive to buy nationally than they are to buy in some of our local markets. As a result, we began rolling out national advertising in the quarter and the initial results, while highly preliminary, look strong. Specifically, we have seen early funnel activity in markets where we do not offer local delivery or utilize local marketing increase by about 80% quarter-over-quarter. There's evidence we are building awareness and demand in markets that we do not yet deliver in but believe a vast majority of the sales won't begin to show up until we open those markets for free local delivery. Accordingly, we expect the benefits of national advertising to be slow building and don't anticipate meaningful contribution to our unit sales volume through the end of this year. Lastly, I'd like to briefly hit on our guidance for the third quarter. Our range for units and revenue for the third quarter is a bit wider than we would typically provide. Historically, we have seen sales bumps in the third quarter that we believe are driven by OEM advertising for their model year-end sales. Given that we sell a lot of late-model cars, we generally benefit from this advertising surge. Looking back to the last 3 years, we have seen significant variation how large that transitory bump is, and therefore, our guidance range of 129% to 159% year-over-year unit growth in the third quarter reflects our uncertainty around the size of this year's impact. In summary, this is an exceptionally strong quarter and the underlying fundamentals of the business are strong. Something we are very focused on as a management team is executing across each of our objectives while also scaling at very high rates of growth and maintaining the great customer expenses that got us here in the first place. The national cadence of the business, given our seasonal patterns, is that we experience our most explosive growth in the first half of the year, which took the form of 90% growth from Q4 2016 through Q2 2017. The first half of the year is the time that execution risks are most pronounced as a result of that growth, and we are very proud of our various teams for the jobs they've done over the last 2 quarters to power us through. The second half of the year is about continuing that momentum and building our operational and technical capabilities to be able to support that expected growth in the first half of 2018, and we are well on our way. Mark, with that, I'll hand it off to you.

Great. Thank you, Ernie, and thank you all for joining us today. Full details on our second quarter financial results are available in our shareholder letter. We've put a lot of financial detail into the letter. So similar to last quarter, I will not repeat many of the numbers. Instead, I will highlight a few key aspects of our financial model, discuss a few key results and give you more color on our expectations for the second half of 2017. Unless otherwise noted, all comparisons will be on a year-over-year basis. We're excited to report another quarter of rapid growth in both retail unit sales and revenue, combined with continued margin improvement. We view retail units sold as the most important driver of our financial model. Retail units sold totaled 10,682 in Q2, an increase of 145%. And total revenue was $209.4 million, an increase of 142%. We are reaffirming our full year retail unit sales outlook to 44,000 to 46,000, reflecting a growth rate of 135% to 145%. We remain focused on increasing gross profit across all parts of the vehicle transaction. Total gross profit in Q2 was $16 million, an increase of 166%. Importantly, we improved total gross profit per unit sequentially by $332 to a total of $1,501 in the second quarter. We have several levers to drive GPU and we saw increases across a number of them during Q2, including efficiencies in our reconditioning centers and improvements in finance gross profit following our transition phase to our new financing partner. We see GPU increasing another $125 to $225 sequentially in the third quarter and are reaffirming our full year guidance of $1,475 to $1,575 as we make progress toward our midterm goal of $3,000 per unit. EBITDA margin was negative 16.1% in Q2, approximately 2 points ahead of our prior outlook and a significant improvement from the prior quarter. Our EBITDA margin outperformance was driven by several factors, including higher-than-expected revenue growth, higher-than-expected total GPU and lower-than-expected operating expenses, the latter reflecting some expenses budgeted for Q2 that will instead take place in Q3. We see EBITDA margin improving sequentially in the third and fourth quarters. We believe that our second quarter results and outlook demonstrate further progress toward our commitment to improving EBITDA margin, an evidence of our clear path to profitability. As we move into the second half of 2017, we expect to continue to grow retail unit sales, open new markets, drive GPU and improve EBITDA margin. We expect OpEx to increase in dollar terms as we expand but to decrease as a percent of revenue in the second half of 2017. OpEx includes anticipated stock-based compensation of approximately $3 million for the second half of the year. We expect very modest increases in share count in the second half. You should use approximately 137 million shares on a fully exchanged basis for Q3 and Q4. We ended Q2 with $144 million in cash and equivalents. Following quarter-end, we upsized our inventory financing facility with Ally Bank to $275 million through the end up 2017, and $350 million through the end of 2018. We also completed our first sale leaseback transaction on our Austin vending machine in Q2 and are working on several additional deals to ensure we are efficiently funding the assets on our balance sheet. We believe asset-based financing combined with the leverage in our business model, which pairs increasing sales in GPU with fixed cost leverage, will significantly reduce our cash investment over the next several quarters and provide sufficient financial flexibility to execute our plan. Our strong first half performance and our expectations for the remainder of the year demonstrate the progress we are making on the top and bottom line. We are executing against our key priorities, leveraging our fixed cost structure, growing our market presence and optimizing our model to realize efficiencies across the business. Thank you for your attention and we're ready for questions.

[Operator Instructions] Our first question comes from David Lim with Wells Fargo.

Several questions. Can you give us more texture on gross profit per unit. If you could bucket that a little on a sequential basis. Just talking about retail growth without the F&I side.

Sure. I'm happy to hit that. So we saw gains in vehicle gross profit in the second quarter over the first quarter, particularly on the recon side, so on the cost side of the vehicle gross profit equation. Some of the initiatives that we have had underway in the inspection centers, including bringing more services in-house and using the infrastructure that we have there in the 3 ICs, paid dividends and we saw reduced costs on the COGS side of the equation. So that was the biggest single driver of improvement in vehicle gross profit in Q2. We also saw gains in optimizing, purchasing and pricing, which contributed to some of the increase in vehicle gross profit in Q2. Both of those, of course, were sufficient to overcome the increase in average days to sale that Ernie referenced in his remarks.

Understood. And then on the guidance at the midpoint and relative to consensus, it looks like Q3 and Q4 revenue guidance is a shade light versus expectations. What are the factors that can make you feel more positive about the outlook? And as a follow-up, how has your new markets trended? And one thing that during the prepared comments was -- what I found interesting is as OEMs spend more on advertising, that could actually drive used vehicle flow-through. Can you sort of parse that out? I would think that because of new vehicle unit incentives and advertising that it could possibly be detrimental to use. But if you could clarify that, that would be very helpful.

Yes. So let me try to handle all that. So first, I'll talk about Q2 and I'll talk about new markets and we'll roll into the OEM effects and how that impacts guidance. So I think in Q2, everything feels really, really good. All the trends, as we said in the prepared remarks, are intact. Our mature markets continue to ramp quickly. Our new markets continue to ramp more quickly. The vending machine impacts are clear. So I think everything there is kind of very good news and we've got clear tailwinds at our back and we're feeling very good about the remainder of the year. I think this OEM effect is a complicated one. I think it's highly variable across different years. If we look back 3 years, we didn't see much of an impact at all from OEM advertising in the third quarter. If we look back last year, we saw an impact that, at our current scale, would probably be worth about 1,200 units in the quarter give or take. And I think you appropriately alluded to kind of the things that can drive that. It depends a lot on if they tend to pour more money into advertising on TV, I think that will generally drum up the interest that we benefit from. If they tend to put their money more into incentives to compete with used cars, it can go the other direction. And I think what we've seen so far in the quarter is it's been a little bit delayed versus what we saw last year, and so that has us providing a little bit wider range for this quarter since we think that transitory effect is a little bit harder to forecast. And then we want to wait until we see how that plays out before we lean too far forward into the full year. But in general, in terms of all the things that we think drive the underlying business, everything that we have control over and visibility into, the trends are very strong.

Got you. And then just 2 more. On the advertising for retail, it came down about 10% in the quarter. Can you give us a little bit more color on that? Are you -- I mean, during the prepared remarks, it looks like you are going through national advertising. And I know that you mentioned that the benefits there haven't really come through, but what is the -- what are the factors that caused the improvement? And then finally, on the Calypso acquisition, can you give us some idea of how much you paid for that acquisition?

Yes. So on advertising, first, I think, definitely, the biggest impacts in the quarter are just a function of kind of market age and new market openings. So in general, the trend that we've seen historically and trends that we continue to see are as we open new markets, the customer acquisition cost drop pretty dramatically over time as the market shares ramp up. The offset to that is that is as we're adding more and more markets more and more quickly, we tend to see higher customer acquisition costs in new markets. And so I think what you see for the last several quarters at least and you should probably expect to continue to see going forward is, on average, at a company level, customer acquisition costs are dropping and that basically is comprised of rapidly dropping customer acquisition costs in markets that are open and then adding new markets as well. And in this quarter, we added more markets than we've ever added historically. So I think those are the major drivers. As it relates to national advertising, national advertising is, as we said, just starting to get efficient. And so right at this precipice of efficiency, it's not really a big economic pickup. It's not like we're able to advertise in all the markets for less money than we spent before. What it really is, is we pick up advertising to more customers in markets that we don't currently serve, and so that's kind of what the initial pick up looks like. We're seeing really good early funnel activity there, much like we do in new markets when we launch them. That, we expect that kind of activity to be very low converting as we don't have free delivery in those markets. But we think it's a very good sign and has us hopeful for what national advertising could do in the future. And then I think as we move kind of into the mid-part of next year and further, that's where the economic benefits of national advertising really starts to take hold because now you're spending kind of the same amount of money to go into more and more markets. And as you add those new markets, you don't have to increase your budget as much. So that has a concrete and kind of mechanical impact on decreasing customer acquisition cost. As it relates to the Carlypso acquisition, the terms of that deal were not disclosed but they are contained in our Q2 financials. And so you can kind of get a sense of how that worked out by looking at the financials.

Our next question is from Nat Schindler with Bank of America Merrill Lynch.

Hey, can you help me out on just understanding -- you added -- well, I guess 25% of your markets were added in this past quarter, big, huge change. I assume that most of those were doing virtually nothing in the quarter in revenue, but they cost you a lot in OpEx, particularly in sales and marketing. Can you walk back how much you probably spent on those launches and what that would have resulted had you not launched those in your EBITDA margins?

Sure. So I think the way we think about the expense of adding markets from a marketing perspective, and we've talked about this concept before, but as an approximation, the way we approach marketing is we go in to an individual market and we start marketing in a way that's sort of correlated with the population in that market. And so when you think about the markets we added in Q2, there were some -- actually, there were some big markets and some small markets mixed in there and they launched at different times in the quarter. But as a general matter, our approach to marketing in those markets was similar to what we've seen in the past, where we go in and market more or less pro rata to population relative to what we were marketing in other markets. So that's the easiest way to think about it.

Yes, and just to jump in really quick. So I do think if you were able to just kind of look at what the financials would have looked like if we didn't open the markets, the answer is a lot of those numbers would have gotten better in the quarter, that's definitely true. I think it's also worth noting, though, that in general, the cash drag that new markets create is rapidly decreasing as our gross margin increases, right? Because effectively, the way you kind of think about that in any given market, the major kind of costs and revenues, the revenues are total gross profit across the sum of all the cars sold in the market. And because of our shared inventory pool, new markets benefit from the higher GPU across the entire company and the major costs are basically turning on marketing in that new market. And those costs, while not diminishing yet due to national marketing, we anticipate continuing to diminish given national marketing. So I think it would look better if we weren't opening markets, that is a drag on many of these different metrics, but we expect that to continue to be a smaller and smaller drag over time.

Okay. And can you guys walk up a little bit your GPU increases from -- that you have seen? Especially considering that you're not getting it off what would be the largest differential between you and your competitors, at least the obvious differential being the time you have the car. And that's actually walking the wrong way. So you made up that plus additional, as you got better in GPUs this quarter, where are you getting that right now?

Sure, yes. So let me hit that again. So the way we can think about vehicle gross profit, there's the revenue side and the cost side. The revenue side benefits from lower average days to sale. We obviously didn't get that in this quarter. And then the cost side benefits from a number of initiatives that we've talked about over time, including getting more efficient in the recon centers, getting more efficient in our inbound transport logistics and then even you could throw purchasing better into that mix as well. The areas where we saw the kind of most significant -- the single most significant benefits in Q2 were in recon efficiencies. So we were able to utilize our infrastructure to take some recon services in-house that we previously outsourced. So we're getting -- utilizing our space to do that. And then we also -- we had some benefits just in terms of kind of lower overhead and other costs. And so big one there is the recon centers, we also feel like in the quarter, we made some gains in just optimizing our purchasing and pricing algorithms and so picked up some there as well.

And when you say optimize your pricing algorithms, are you still thinking that you're running roughly $1,400 below comparable for the consumer? Or has that been -- any movement on that side?

Yes, let me try to walk you how we think about pricing. I think we think about pricing as having had 2 forms. The first is kind of the initial price that we put up on the website and kind of the write-downs they think they're after, and the second form is how any given car is priced relative to another car. I think, in general, when we use the term optimize, we're talking about tweaking the way that cars are priced relative to one another. So we can identify a car that has more demand, we may want to write that car down less than the car next to it. If we identify a car that has less demand, we may want to write down that car a little bit more. I think those optimizations are driving the majority of the sales. The initial price -- or excuse me, the majority of the gains in GPU, the initial price continues to be fairly consistent across borders.

Our next question comes from Mike Levin with Deutsche Bank.

I wanted to follow up a little bit on the kind of cash balance coming in a little bit lower than I think we were thinking at this point. I believe it's just because you're kind of progressing faster. Could you sort of walk through how you're thinking about liquidity from here getting to free cash flow breakeven and whether or not you'd be kind of considering additional capital raises to give you a little bit more cushion and maybe grow a little faster?

Yes. So there's a lot in there. So I think, first of all, the way we're thinking about capital, I would break it down into 2 buckets. One bucket is operating capital. So how much cash are we spending to run the business. And as we discussed in our previous call, the way we think about that is we're rapidly growing revenue, we're also bringing down EBITDA losses or EBITDA margin losses pretty quickly, and we expect that to kind of take the form of relatively flat dollar losses through the end of this year, and then we expect it to improve pretty quickly thereafter heading into next year. So that's kind of what's happening with operating cash. Then on the CapEx side of the business, we continue to invest in CapEx at a very similar rate to how we've been invested in the past. The vast majority of that CapEx is financeable assets or generate financeable assets. So as Mark alluded to, we did our first sale leaseback on our vending machine in the quarter. We upsized this line with Ally. So we anticipate doing additional lines to reduce our cash consumption based on CapEx. We also, I think it's worth noting, have significant assets on the balance sheet that we've built up over time through that accumulated CapEx that we have the ability to finance as well. So those are probably somewhere between $50 million and close to $100 million of financeable assets on the balance sheet as well that we can go and access over time. So that's how we're thinking about it. In general, we told you last time that feel like we've got sufficient capital to run through our plan. That continues to be the case. We ran ahead of our plan last quarter and we're feeling good.

Great. And given your current growth trajectory, when do you believe that you guys will have kind of full occupancy of the first 3 IRCs and you'll sort of effectively be ending any appearance of support from DriveTime?

That's a hard question to ask because I think implicit in it, it is multiyear guidance. So I think what I'll say we've been growing at 145% year-over-year in this quarter. We expect to continue to grow at very fast rates. We said before that those 3 inspection centers would support about 150,000 units of annual capacity. I think that gives you kind of a sense of where we expect that to happen. And then just to reiterate, I think you stated it well, but I think the appearance of kind of the financials not being completely clear because of the DriveTime partnership is the only thing that will change there. We've worked very hard to make sure that the economics are at arm's length and are fully representative of kind of how we're spending money in those inspection centers.

Got it. And just maybe lastly, we haven't really seen an update on that market penetration curve really since the end of 2016. Is it possible to maybe just sort of say kind of about where Atlanta or Nashville might be running currently or maybe kind of what those sequential market penetrations are kind of moving at, at this point just so we have a way of thinking about how that curve kind of moves forward?

Yes. So just in terms of reporting in general, we plan to report on kind of the cohort curves in our market shares every year in our case. So that's what we plan to do. I think we want to be careful about giving too much guidance at a market-by-market level. I'll just go back and reiterate, in general, the trends, they continue to be the same as they've been in the past. Our mature markets are still ramping very quickly. Our newer markets are ramping faster. The vending machines continue to have a positive impact on market share, and that's why we were able to exceed our goals in Q2.

Our next question is from Mark Kelley with Citigroup.

The first one is just a follow-up on the last one on the GPU side. Some nice improvement this quarter, obviously, and you talked about a lot of that was due to IRC utilization improving. Could you just talk about the right way for us to think about utilization and how it impacts the P&L once the Phoenix facility is up and running. I know you don't want to give multiyear guidance, but just the right way for us to think about that. And then second, on Carlypso, are there any other areas that you think you can add to your own in-house tech stack? I know you've built a lot of it yourself. Just curious if you have any other kind of obvious areas where you might want to buy versus build.

Great. So Mark, I'll hit your question about the Phoenix IC. So our plan for rolling out the Phoenix IC, it'll basically happen one line at a time. So we've spent some time talking about how these ICs function. Each IC has 3 production lines, and then each of those production lines can run at least 2 shifts per day at full utilization. As we've also discussed historically, we haven't really implemented double shifts in any of the facilities with the exception of half line or so in one of them. And so in Phoenix, when we think about how that will come online, we'll just do it one line at a time. So we'll start with one line. We'll staff that line, start producing units out of that line, and then at that point, that facility will be somewhat under-utilized, and we'll then add a second line, a third, et cetera, as sales demands. So that's the way we think about that.

Yes, and on your point about technology, I don't say anything too definitive here because I think we will opportunistically kind of look at opportunities as they become available. But in general, we don't anticipate doing a lot of acquisitions. We don't think that's a big part of our strategy. I think we'd be highly unlikely to do any operating acquisitions. We're trying to integrate operations. I think an acquisition like this one that is really kind of folding technology into the business in ways that we think are supportive of our core objectives, I think those sorts of acquisitions could occur in the future, but we don't have anything that we're looking at today.

Our next question is from Colin Sebastian with Robert Baird.

A couple from me. First off, just looking at the second half ramp, it continues to imply that units per market should grow fairly significantly, in particular, in the fourth quarter. And if you could update us on the visibility you have behind that increase, and if there's any stabilization in ASPs implied with that? And then secondly, with respect to new market openings, just hoping to gauge your comfort level with being able to maintain this sort of -- this pace of market expansion from an operational and management bandwidth point of view at least as we look out over the next 3 or 4 quarters.

Yes. So I mean as far as second half ramp, I think, again, I would just reiterate, the markets are performing as we would have expected and we feel very good about our ability to continue to hit our numbers and for those markets to continue to perform. Moving over to the new market openings, I think those markets also, just to start, continue to ramp faster than preexisting markets, which we think is a great trend that we're excited about. As far as our ability to continue to open markets in this pace, I think this year we're opening markets at a significantly faster pace than we ever have. I feel like where we are right now, we're probably opening those markets more comfortably than we were a year ago because we build out -- or built out more operations to support that market growth. So we feel like what we're doing right now is a sustainable pace that we can keep up.

And then maybe lastly, Mark, if you could quantify the amount of expense that was pushed into Q3 from Q2, that will be helpful.

Yes. So we think that was -- it was between 50 and 100 basis points of margins, so $1 million to $2 million.

Our next question comes from Sharon Zackfia with William Blair.

Just 2 quick questions. I guess first on the sale leaseback in Austin, could you give us what the terms were for that? And then secondarily, obviously, Chicago, I think, is your biggest market. I know it's early, but is there anything that you're seeing in Chicago that would influence the way you attack larger markets in the future?

So first, on the sale leaseback, I think our policy here in general is going to be to not disclose terms on many of these smaller financing deals, but we do believe that we will be able to access these markets again for other vending machines in the future. Chicago, it's really early to be giving any feedback there. So I don't want to lean into that question too much. What I would say, again, new markets tend to ramp faster than older markets. That's true in general, and I think that's true in the very early data so far in Chicago as well. And then in general, we tend to grow market share a little faster in smaller markets than we grow it in larger markets. I think Chicago, in particular, may be a little bit of an exception to that rule so far with very, very limited data. So I think there's lots of reason to be very excited about our kind of 2017 opening cohort, but I think all of that is very, very early for us to be talking too much about.

Our next question comes from Ron Josey with JMP Securities.

This is Andrew Boone on for Ron. There's a concern that the new and used car market is slowing. Can you guys just talk about how Carvana is positioned there should there been a slowdown in the used car market? And then with the Phoenix IRC set to come online, can you guys give us any details around your West Coast rollout?

So on the first point, I think we probably should break that into 2 points though. First order, I think as the market moves, we would expect the first order impact to be similar to us. So if the market goes up by 5% or down by 5%, I think we would probably expect the impact on us to be similar. I don't think we're immune from industry moves in that way. So I think that's how you should kind of think about that. I think second order, we're growing at -- in the last 12 months, we've grown by 145% year-over-year. I think that's really, really large compared to any reasonably expected moves in the industry more broadly. And so we are much more focused on our execution, opening new markets, continuing to give customers great experiences and watching those market share curves grow in the markets as we think that's a much, much greater driver of our success over the next several years. So I think that's how we would think about the new and used car market movement. And then as far as Phoenix IRC goes, we're on pace there. That's tracking well. We're looking to open that the second half of the year. Opening up Phoenix was a really big deal. That meant that we connected our logistics network from Texas, which is our furthest west markets out all the way to Phoenix. That sets us up really well to go into the West Coast and the Southwest through the second half of the year and we're very excited about what that means for the business.

Our next question comes from Dan Salmon with BMO Capital Markets.

Ernie, could you talk a little bit more about the trade-in program, in particular, sort of traction you've seen on it? I think you do get a little better margins with what you see there. And how you think about marketing that a little bit more within the context of your broader advertising and marketing plans?

Yes. So the trade-in program has been growing roughly at the same rate as the business for probably the last year or so, give or take. I think it is an enormous opportunity for us to grow that business even faster than the core underlying business, and that's something that we're putting a lot of thought and effort into on the product side of the business. I think, in general, those product improvements and advertising improvements tend to be a little bit chunkier, so I'm not sure that's something that you should expect to see rolled out quarter-after-quarter. I think as we roll out product initiatives, the impact of the trade-in program or VSC or any of our other revenue items, we'll make sure that we inform you about that. But I do think that's a big opportunity for us and something that we'll look to advertise and invest in more from a product perspective over time.

Our last question is a follow-up from David Lim with Wells Fargo.

Just a follow-up. If you think about you guys maintaining that 90-day supply in Q2, would that have been like $150 or $180 add to GPU in the quarter? And also can you give us some color on how your operations are in Texas? I mean, obviously, AutoNation and Group 1 have cited some difficulties in Texas. Your model is obviously different, but we'd like to get some color there as well.

Great, David. I'll hit the first point. So we've historically seen on the order of $10 per day in sticker price depreciation for the cars that we hold in inventory. That's still true today and so do think that had we not seen increased average days to sale, that our sticker price -- our sticker prices would have been higher, gross profit would have been higher. So that's just to hit that point.

And then on...

Yes, I'll jump in on Texas. So Texas and Florida are 2 markets that we've heard a decent amount about potentially softening across the industry. I think if you look at our performance in the last quarter, those were the 2 states we actually grew the most. They outgrew the rest of the country fairly significantly. So we're very pleased with that performance. Now I do think we would have expected that for 2 reasons. In the case of Texas, we launched several vending machines there over the last 6, 12 months, and so we would have expected that to impact things positively, but it is -- that is coming to fruition. And in the case of Florida, that's a younger market, and generally, younger markets tend to grow a little bit faster as a percentage. So we would have expected that as well. When controlling for those 2 factors, I think debatably, we've seen a little bit of softness in Texas and Florida relative to the rest of the country, but certainly nothing super pronounced. And I think that's just kind of more evidence that what matters more is us executing our plan, not really what's happening in the macro environment because those affects, while we're not immune to them, they are just small compared to the impact of us running our business.

And we do have one final follow-up from Mike Levin with Deutsche Bank.

Hopefully, just 2 quick ones. You noted getting to 80 days to sale by Q4. Do you still see getting to 45 days by the first half of 2018? And then just also do you see any need to upsize the finance receivable portion of the facility with Ally? Or does the $400 million kind of carry you through to the end of next year?

Sure. Yes. Let me hit both of those. So David, we do expect to see declining average days to sale through Q3 and Q4 and are targeting 80 days in Q4. We also have historically seen significant sales growth from Q4 going into Q1 because we get seasonal tailwinds that boost sales after we spend a couple of quarters fighting through seasonal headwinds. And so we've given that 80 number for Q4 and do think that it will fall further in Q1 associated with that increase in sales growth. So I'll hit that. And then the second point on Ally, so as it relates to the ally receivables facility, so everything there is going well. As we've discussed previously, we're very happy with that arrangement and we have still significant capacity to sell receivables over into that facility. Because we are growing so quickly, we will, at some point, need to either upsize that facility or replace it, but everything right now is going very well.

Yes. And Mike, just to add a little more color on kind of the turn time. I do think it's worth noting that since Q4, we've grown sales by 91% while holding inventory flat. I think if we do that again and you're kind of acquiring inventory at the same speed you historically have acquired inventory, you would almost cut turn times in half there. Obviously, the dynamics are a little more complicated than that. But roughly speaking, that's approximately what will happen. So there's clearly a lot of power in that, and then as Mark alluded to, we generally see the biggest growth in sales in the first half of the year. And so that's the time when we will benefit most from holding inventory relatively flat to see that coming down. So we do expect that to continue to come down pretty quickly.

This concludes our question-and-answer session. I would like to turn the conference back over to management for any closing remarks.

Thank you all for joining our conference call. We look forward to seeing you out at the conferences, we'll talk to you again soon.

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.