Earnings Call: Q3 2019

Nov 6, 2019

Good afternoon, and welcome to the Carvana Third Quarter 2019 Earnings Conference Call. All participants will be in listen only mode. Please note, this event is being recorded. I would now like to turn the conference over to Mike Levin, Vice President, Investor Relations. Please go ahead. Thank you, Gary. Good afternoon, ladies and gentlemen, and thank you for joining us on Carvana's Q3 2019 earnings conference call. Please note that this call will be simultaneously webcast on the Investor Relations section of the company's corporate website at investors. Carvana.com. The 3rd 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 would 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 future financial results that involve risks and uncertainties and may cause actual results to differ materially from those discussed here. 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 most recent Form 10 ks and Form 10 Q. 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. Our commentary today will include non GAAP financial measures, including, but not limited to, ex gift measures that exclude the impact of the 100,000 milestone gift to our employees. Reconciliations between GAAP and non GAAP metrics for our reported results can be found in our shareholder letter issued today, a copy of which can be found on our Investor Relations website. Please note that all gross profit, SG and A and EBITDA metrics mentioned by us on the call today are on an ex gift basis. And now with that said, I'd like to turn over the call to Ernie Garcia. Ernie? Thank you, Mike, and thanks, everyone, for joining the call. Q3 was another great quarter for us on our mission to change the way people buy cars. It was our 23rd straight quarter of triple digit revenue growth. We also saw nearly 2 50 percent growth in the number of cars we bought from our customers, 2 50%. As a result of that growth, we bought 70% as many cars from our customers as we sold to them, and we sourced 31% of our retail cars from other Carvana customers. Our rapid growth in both cars bought and sold led to total transaction growth of 143% in the quarter, which is the fastest rate of growth we have seen since late 2017. That may not sound like that long ago, but at that time we were a business roughly 1 fourth the size the business we are today. It's pretty exciting to still be growing as fast as we are at the scale of over $1,000,000,000 in revenue per quarter. Our offering of buying cars from customers was a standout this quarter, so I want to take a little extra time discussing it. The annual growth is a remarkable figure, but I think the quarterly growth is even more telling. We grew Carspot from customers by 40% quarter over quarter. That is pretty exceptional and has unsurprisingly put some pressures on the business to quickly adapt. Those adaptations are well underway and include investments in several areas as well as additional preparations for another big growth year in 2020. We view the progress we are seeing in buying Carsmart customers as a significant improvement to our platform. When reducing used car sales to what they fundamentally are, they're simply swaps between different customers through the mechanism of all the middleman institutions that make up automotive retail. The more of that chain that we can integrate and improve, the more value we can pass on to our customers and the better business we can build. Now I'd like to turn to the current state of the business. When we launched Carvana, we felt like we had 3 simple questions that separated us from achieving our goals. Number 1, could we build an offering compelling enough that customers would buy a car in a whole new way? Number 2, could we do that with strong unit economics? And number 3, could we execute against that enormous opportunity? Revisiting these questions is a useful way to assess our progress. We believe the first question has been addressed. The quality of our customer offering, which drives our growth, answers it. The fact that in less than 7 years, we have become the 3rd largest retailer of used cars in the U. S. With a completely new offering answers it resoundingly. We've built something that our customers love. We believe the second question has been answered as well. We are not yet a profitable company and we remain intently focused on this goal. But on the question of unit economics, the data is pretty clear. In the Q3, 80% of our markets, which made up 97% of our sales, were contribution positive and 14 markets, which made up 35% of our sales, were EBITDA positive after fully allocating all logistics and corporate expenses. The company level gains are every bit as powerful. In just 3 years, we've taken GPU from about $1,000 to about $3,000 and we've improved our EBITDA margins by nearly 20%. All that progress and leverage has come despite the investments required to grow retail transactions roughly 10 times and total transactions approximately fifteen times over that same period. We've built a business that already has strong unit economics and there's clear visibility to our long term model. This leaves us with the 3rd question. Can we execute against this incredible opportunity? Our execution so far gives us confidence. In about 6.5 years, we've gone from 0 to 70,000 transactions per quarter. That said, this is a question that is never fully answered. It just suddenly changes to can we continue to execute. I believe we will continue. That belief comes from the quality of the passionate people we've assembled and the quality of experiences those passionate people deliver to our customers. Thank you to all of those passionate people. Our goals are ambitious and clear. We want to change the way people buy cars and become the largest and most profitable automotive retailer. And we're still just getting started. Mark? Thank you, Ernie, and thank you all for joining us today. Unless otherwise noted, all comparisons are on a year over year basis. We are pleased to report another quarter of exceptional growth in both retail units and revenue. Retail units totaled 46,413 in Q3, an increase of 83%. Total revenue was $1,095,000,000 an increase of 105%. This marks our 23rd consecutive quarter of triple digit revenue growth and the first in company history with more than $1,000,000,000 in revenue. Total gross profit per unit was $29.96 in Q3, an increase of $6.94 Retail GPU increased by $169 reflecting gains in acquiring cars from customers. Wholesale GPU increased by $61 driven by 165% growth in wholesale units sold. Finally, other GPU increased by $4.64 reflecting gains in finance monetization and increased attachment of ancillary products. EBITDA margin was minus 5.1 percent in Q3, an improvement of 3.2%. SG and A levered by 1.6 percent despite investments made in the quarter to relieve pinch points, support accelerating growth in buying cars from customers and prepare for growth in the first half of twenty twenty. We ended the quarter with $578,000,000 in committed liquidity resources and held an incremental and $10,000,000 in real estate and securities on our balance sheet. Following quarter end, we also upsized our floor plan with Ally, increased our credit line to $950,000,000 and adding flexibility to expand our inventory selection and buy more vehicles from customers. As of September 30, this upsize would have unlocked an additional $73,000,000 in liquidity based on inventory on our balance sheet, bringing our total liquidity resources adjusting for the upside to over $760,000,000 dollars With our 9 new markets opening in Q3, we now serve 66.9% of the total U. S. Population, up from 58.6% at the end of 2018. For the remainder of the year, we plan to turn our focus to preparing for growth in the first half of twenty twenty and for buying more vehicles from customers. We plan to resume a rapid pace of market openings in 2020. Beginning next quarter, we plan to guide on growth in population coverage rather than growth in number of markets as this metric will be more relevant as we move into smaller markets and fill in existing regions. In Q3, we began construction on our 8th inspection and reconditioning center, a four line facility in North Carolina that we expect to add 67,000 units of annual production capacity to our existing footprint of 350,000 units at full utilization. In addition, we have identified 5 IRC sites that we expect to become 4 line facilities over time. Continue to view IRCs as a long term competitive advantage as we further expand our as soon as next day delivery infrastructure. Q3 also marked another successful quarter for our finance platform. On September 27, we closed our 3rd auto loan securitization, selling $600,000,000 of principal balances and further diversifying our investor base. The finance GPU was $10.78 in Q3, an increase of $3.73 We are excited about what this progress means for our finance platform and expect to recognize additional gains over time on our way toward our long term financial model. In terms of outlook, we are raising our full year guidance retail units sold to 174,000 to 176,000 and total revenue to $3,850,000,000 to 3.9 $5,000,000,000 based on another strong quarter of results. We are also raising our guidance for total GPU and fine tuning our guidance for EBITDA margin, reflecting incremental investments in our business of buying cars from customers and scaling retail unit volume both this year and in 2020. As we look toward the end of 2019, we are excited about our progress toward our long term financial goals. Thank you for your attention. We will now take questions. To one question and one follow-up, then you can go back into the queue. The first question comes from Rajat Gupta with JPMorgan. Please go ahead. Hey, thanks for taking my question. Just wanted to follow-up a little bit on the retail GPU number in 3Q, which was down from 2Q levels, a little more than what we saw last year despite a higher mix of retail sourcing. Is that is there a lag of the benefit that we could expect from retail sourcing? Or is that just normal seasonality that we should be expecting going forward? And I have a follow-up. Thanks. Sure. So we certainly benefited from retail sourcing from consumers in the Q3 relative to the Q2 and also year over year. That said, there were a number of offsets if we look sequentially. Those offsets included high wholesale prices in Q2 and the early part of Q3, followed by relatively high depreciation rates in the latter part of Q3, which definitely had an impact on our vehicle margin. Moreover, another sequential change was a reduction in delivery revenues on a per unit basis that came along with us scaling inventory on the eastern half of the U. S. With our Indianapolis, Cleveland and Nashville IRCs coming online. And so those are some of the offsets in sequential retail GPU. Obviously, we're very excited about our progress overall in retail GPU. It's up about $170 year over year with buying cars from customers definitely contributing to that. I think as we look forward, we see a lot of upside as we continue to source more retail cars from customers and continue to optimize our bidding and pricing algorithms, where I'd say we're at a very early stage in doing that so far and look forward to a lot of upside in the future. Got it. That makes sense. And just on the other costs within the SG and A, that did not lever up as we just expected. And it's probably also the biggest bucket of opportunity longer term for your SG and A unit targets. You talked about some technology investments in the release. Was that more of a one time step up that should not repeat going forward? Or was it related to the retail sourcing initiatives? Could you just give us some color on that? And how should we expect that to lever going forward? That's it. Thank you. Let me open and then I think Mark will come in with a little more detail. I think the simplest way to think about this is just that we really have grown this business of buying cars from customers, largely over the last 12 months. And I do think that doesn't show up in our revenue as clearly. Many of those cars end up going through wholesale and we basically only see the gross profit portion of those in revenue. And then many of those cars go to retail and they basically just show up as lower COGS. So when you're looking at kind of leverage on a percentage basis, I think it's important to keep in mind that the business has grown a lot in terms of the total number of customers that we're servicing. We actually grew transactions year over year by about 42,000, which compares to the 46,000 total cars that we sold. So that's a lot of growth that isn't flowing through revenue. It isn't going to show up as percentage leverage. Despite all of that growth, all that extra work we had to do in the business, we still did lever by 3.2 percent. So I think overall, we're pretty excited about that and Mark can give you a little more detail. Yes, sure. So I think Ernie hit on some of the key points, but I think the main things that we're focused on from an SG and A perspective are 1, alleviating pinch points and 2, supporting the business of buying cars from customers, which is growing at levels that certainly exceeded our expectations. And also, we now expect to have more of an impact on the business in 2020 than we previously anticipated. And so we're investing along a number of dimensions for those two reasons. First, we're investing in staffing to help alleviate pinch points both in the short term and to make sure that we don't see further pinch points in the first half of twenty twenty with the substantial transaction growth that we've been seeing. We're also investing in technology really with the same purposes in mind, working to make sure both in the short term and as we move into 2020 that we're making the investments to ensure we're providing a great customer experience and have the back end process efficiencies to support this level of growth. And then 3rd point that I would raise, and this relates almost more to Q2 Q4 rather than Q3. I think we are planning to advertise a bit more than previously expected to further support building our brand as we now have this fledgling but rapidly growing and now actually sizable business in SDC. Got it. That's great color. I'll pass on. Thanks. The next question is from Chris Bottiglieri with Wolfe Research. Please go ahead. Hey guys, thanks for taking the questions. Just first one, is there any elaborate on the bottlenecks at all? Does that have any impact on kind of retail unit growth or any of the market level or is it more on the back end side of the equation? Sure. Yes. So the primary pinch points that we hit in the quarter were at the market level. I think if you think about the impacts that the business of buying cars from customers has on the business, It certainly impacts last mile delivery, where we provide a great experience to the customers by going to pick up the car at their house after we've appraised it and they've agreed to accept our offer. That takes delivery slots that can then impact the retail business and extend delivery times. There are other impacts throughout the business. We're taking more calls. We're getting more visits to the website. Yes, I think there's it's running through everything that we do. But I think the biggest pinch points that we felt acutely in the quarter were related to last mile delivery. Got you. Okay, that's helpful. And then on the credit business, you called out reaching more investors on the platform. Hoping maybe elaborate a little bit more there. Did see that in the equity residual side as well? And then maybe if you can just kind of give us like how you're thinking about the direction of better monetizing your side of the equation on the equity residual and the excess notes, that'd be helpful. Thank you. Sure. Yes, so I think we think about better monetizing our loans through the securitization program by through a number of channels. I would say definitely continuing to educate the market about the quality of our loans and we've talked at length about this in the past, but our loans performed significantly better than other sort of auto market loans conditional on credit factors because of our online model, our deal structure and a number of other factors. And so I think educating the market about the quality of our loans, continuing to work with existing investors, bringing new investors across the entire capital structure throughout the entire securitization structure. Those are all certainly things that we're going to stay focused on. We feel like we've made great progress so far here with 1078 GPU this quarter, and we'll look to make further gains in the future. All right. Thank you. The next question is from Tom Champion with Cowen. Curious if you could just elaborate a little bit more on the rapid rise in customer source ratio within retail sales, that's 31%. I mean, it's just a very dramatic rise. And what changes had to be made, need to be made to support this new higher level of direct customer sourcing? Thank you. Sure. So I would start with the customer experience that we deliver to customers when they buy a car from us. So customers go to our website, in seconds they get a value for their car, and then they schedule a time and we come pick it up, take it away and put money in their account. So it's a very, very simple experience. And it's an experience that we're uniquely positioned to provide because we've built a logistics network that enables that across all of our markets. And we've built a logistics network with enough flow to do that very quickly. And so I think that's where it starts. And then once we start to advertise that offering, customers become aware of it, we start to drive a lot more volume. That's where we often talk about our growth in buying cars from customers generally before kind of separating which cars are going wholesale and which are going retail. That's the business that just continues to grow very, very quickly. Again, this quarter, that was 2 50% roughly and about 40% quarter over quarter. I think that's just the strength of the offering speaking. And then once we buy those cars, we have to determine which of those cars fit our retail standards and which don't. We would love for all of them to fit our retail standards, but we've got fairly tight standards on that side and so they don't all fit that. So some of them do go and get wholesaled, and we monetize that as a wholesale line item. And then those that do fit our retail standards, we obviously retail. And that's been growing very, very quickly. I think that's a number that tends to have a little bit of momentum in it relative to the first number, the number being percentage of cars that we buy for customers relative to those that we sell because it takes us more time to turn a car retail and then go sell it to a consumer than it does to buy it and then go wholesale it. So we had to make a bunch of adjustments across our entire pipe to be able to handle that. As Mark talked to earlier, you get more customers calling in with questions about how they pay off their loan balance on our previous car. You have to handle the title transfer. We have to increase our wholesale ops. We also tend to buy a slightly different kind of car. The cars you buy from customers tend to have a bit of a broader distribution and cover more kind of make, model, year trend space than the cars that you get at auction. And so that also has impacts our pricing algorithms. We're just starting to really get a look at a lot of these cars and it's already smarter how we price them, both on the buy side and on the sell side. So I think there's some changes there. So across the entire business, I think there's just a lot of changes that we have to make to continue to get smart here. And I think there's a lot of upside remaining. We're very, very excited about it. But that number being 31% is pretty exciting. In early 2018, that number was approximately 7%. And that's when we really started to put effort behind this. And then it's obviously grown very quickly from 17% last quarter to 31% this quarter. We put out our long term financial model. Our goal range there was 38% to 52%. We've made a ton of progress against that. And I do think that all just comes back down to the simple customer offering that we provide. And then we're going to turn as many of those cars retail as we possibly can and adjust the business to handle that volume. Thank you. The next question is from Seth Basham with Wedbush Securities. Please go ahead. Thanks a lot and good afternoon. My question is related to buying cars from consumers. How much did you dedicate to advertising to promote that business this quarter on a year over year basis relative to the incremental $10,000,000 if I recall last quarter? So I think last quarter we outlined that it was approximately 1 5th of our ad spend. And this quarter, I would say, it was in a similar range. And until we update you, you can probably anticipate it being in a similar range going forward, but that's roughly where it's been. Got it. And if you think about the GAAP in retail profitability in cars you source from consumers versus from other sources, how has that trended over the last couple of quarters? So historically, the incremental profits that we get from the car that we buy from the customer relative to a car that we buy at auction, it has been similar in magnitude to the wholesale profits that we get when we buy a car from a customer and we sell it at auction. So I think that's a pretty good mental model for approximately what the incremental profits are. Obviously, there's little fluctuations quarter to quarter, but that's roughly a good way to think about it. All right, fair enough. Thank you. Thank you. The next question is from Nat Schindler with Bank of America Merrill Lynch. Please go ahead. Yes. Hi, just a couple of quick questions. 1, as I look at and I know these are biased because they only tend to report when there's a complaint. But if I find complaint sites, one of the things that I really saw as a stark new change was a lot of complaints on title transfer and that happened recently. Is that related to the fact that you're suddenly buying so many new cars from consumers and making it more challenging in some cases to do it. And I don't believe this is happening at a significant rate. It's just those are the complaints that I'm hearing that I didn't see before. Secondly, on a different question, but as you look at 63% market coverage of the U. S. And you look at your oldest markets as they continue to gain share in those markets, do you compare yourself to, let's say, CarMax, who sits at around 5% in markets where it has stores and say that across, call it, 63% of the U. S, you could realistically get to that kind of share within those markets? Sure. So let's start with the first question. So I think the first question is a fair question. I do think it's very valuable in general to drill into all the areas that you can improve. But just to start, we continue to deliver very high quality customer experience and are very proud of the offering that we provide to customers. But there's no doubt there's areas where we can always get better. And I think the area you called out is probably a reasonable one. When we have these pinch points anywhere in the business, that just creates a little extra friction, a little extra time for things to go wrong and people are a little busier across the board. And so I think those things can emerge. I think specifically the title cases, I'm not aware of the specific complaints you're pointing to. But I do think it's probably pretty reasonable to assume that, that is coming from our growth in the business of buying cars from customers as we do need to transfer those titles and handle loan payoffs and all the other things you have to do when you buy a car from a customer. So I think that's probably the source of it. And then I think that's where some of those investments are going to hire more people to make sure that we're in a good position to be able to resolve any customer issues that come up and to build better technology so those issues come up less often. So I think that's something that we're definitely focused on. On the coverage and our goals for market share, I think I would point to our long term goal of trying to sell 2,000,000 plus cars per year. That goal implies about a 5% market share around the country, and that is something that we believe that we can achieve. We think it's something that there is sufficient evidence today in the way all of our cohorts are performing to believe that that's very possible. And so that's our goal. We are only a 7 year old company. And in that short period of time, we've become the 3rd largest automotive retailer. That speaks pretty clearly to customers' response to our offering of buying cars online. And then we also have a business that gets better as it gets bigger. We put some information in the shareholder letter where we recently launched 2 inspection centers in Indianapolis and Cleveland. Prior to that, we didn't really have inspection centers in kind of the Midwest. And so for many markets in the Midwest, suddenly cars got a lot closer to them. That meant the offering got better. They had broader selection. They had faster delivery times. What we saw there is we saw a 20% reduction in average miles traveled for all cars that were sold in those 10 markets that were nearest those ICs. And then we also saw sales more than double in those markets and grow over twice as fast as we would have otherwise expected. And so I think that just speaks to positive feedback. And that's why I think our view is with everything that we're seeing, the major question for us is can we continue to execute at this increasing scale because we've got a business model customers love, we've got unit economics that are really, really positive and we've got positive feedback in the business model. And so we're focused on execution. We think that's the most important thing that we can work on. Great. And a quick follow-up. As you said that on there that this isn't a particular cause of harm, but it's something that could happen. Just to square the circle here or whatever, have you seen any change as you've gone to 31% of your cars sourced from consumers? So has there been any change in your return rates or something like that? No. Cars sourced from consumers have very similar return rates to cars sourced from any other source. NPS scores are strong on Carsourced from consumers. I think there's all the trends in the business are very positive there. Great. Thank you very much. Thank you. The next question is from Nick Jones with Citi. Please go ahead. Hi. Thanks for taking my questions. I guess, first, can you talk about what drove the view that you'd like to reach 95% of the population in the U. S. From I think the original where it was like maybe mid-80s if you were looking at MSAs over 200,000? And then I guess second on top of that is what is delivery into these smaller markets look like? Is it flatbeds, are there 9 car haulers? Any color there would be helpful. Sure. So I would say at a high level, we launched a lot of smaller markets in 2019, and we've seen really, really positive response in those small markets. I think we've always talked about the trends that we've seen in our cohorts, where generally speaking, the older markets are still growing very quickly and the newer markets grow faster than that. And then we've got smaller markets that tend to grow faster than large markets. That really has continued to be true. And I think that's pretty exciting for the model because we think that unlocks this additional population that we can now go serve very efficiently. We also tested in 2019 many markets that we call virtual markets, which are basically markets that we deliver to from nearby markets without having to have a physical presence there. And that also worked very well and was very cost effective. And so I think those two things combined to kind of increase the population that we believe that we will be able to effectively and efficiently serve, and that's what's led to us increasing our expectations there. Great. Thank you. The next question is from Zach Fadem with Wells Fargo. Please go ahead. Hi, this is Eric on for Zach. Thanks for taking the question. Earlier in the call you talked about stepping up advertising in Q4. Just wondering if you can talk about what you're doing more this year and particularly as you look at sort of your Cyber Monday promotions? Sure. So let me take those one at a time. So I think in terms of stepping up advertising in Q4, there is a lag time from when you advertise until you see the response to that advertising. And I think given the quality of response that we're seeing in buying cars from customers, we're feeling pretty good about that and we're trying to prepare the business for higher volumes of cars bought from customers than we were previously anticipating in 2020. And so that's going to take many forms. One of those forms will be increased advertising directed at that offering. So that's one. On Cyber Monday promotion, we plan to run that again. Our rationale there is that it's an opportunity for us to have a different way to speak to customers and to continue to build the brand. Customers are accustomed to buying items from different retailers that are generally line during Cyber Monday. And that's an opportunity to brand what we do selling a car online being a different offering. And so that's something that we plan to do again this year. The next question is from Rick Nelson with Stephens. Please go ahead. Thanks. Good afternoon. I'd like to follow-up on the retail. GPU pulled back a bit sequentially. If you could discuss again the drivers there and what happened to average days to sale in the quarter? Sure. So average days of sale was 63 in the quarter. That was flat year over year. I think average days of sale, right now, our kind of expectation is that we're very comfortable with the level of average days of sale that we're operating at today. We see that bouncing around in a stable and reasonable range. In the long term, we think we have lots of upside to our current sort of high 50s, low 60s range that we've been in over the last several quarters. But for now, I think we're very comfortable with that range. In terms of the drivers of the sequential change in GPU, I think there were a few to point out. One was that wholesale prices were relatively high in Q2 and Q3. That had an impact on cars sold in Q3, particularly since depreciation rates were relatively high in late Q3, which also had an impact on sold GPU for cars that were sold in the quarter. The other thing that I called out or that I would call out is, we saw a decline sequentially in shipping revenue per car And that was largely driven by the increase in inventory in our East Coast or Eastern Half IRCs, which came online in late 2018, early 2019, those being indeed Cleveland and Nashville, which tended to have lower shipping fees. And so we saw a sequential decline in shipping revenue. Got you. And would you expect some of the pressures caused by higher auction prices and the depreciation challenge that that will continue into the Q4 or can you adjust pricing for those vehicles? Well, so the Q4 is typically weaker from a depreciation perspective overall. We've certainly seen seasonal depreciation rates really starting in September. And so that definitely has historically and we would expect to be a factor in retail GPU in every Q4. That said, I think the flip side of high depreciation rates is it means wholesale prices are lower than they were a month or 2 before. So that can benefit margins other things equal. Great. Thanks and good luck. Thank you. The next question is from Armintas Sinkevicius with Morgan Stanley. Please go ahead. Great. Thank you for taking the question. When I look at the markets available that you haven't yet accessed, markets like Minneapolis, Portland, Seattle. Are there any reason is there any reason to think that you won't enter those markets next year? And if not, how do you think about rolling out the reconditioning centers and such to prepare for those launches? Presumably, that's something that we should be looking for in advance of you opening them. Sure. So there's a couple of points I would make there. I think one as it relates to IRC expansion. So we've definitely proven that we're capable of opening markets, which Ernie outlined a few of those key benefits earlier on this call. In terms of expanding out to new regions, I think we'll provide some update on our expansion plans as we go forward into 2020. We provided some guidance that we plan to resume a relatively rapid pace of market openings in 2020. And so we'll provide some more color on that looking forward. Okay. And then the other question I have is you have a nice chart here on the SG and A per retail unit by cohort. Looking out, it looks like it takes about, call it 4 years for the cohort to reach breakeven looking at the 2015 cohort and then it picks up call it 100 basis points a year. Just trying to think through how I could extrapolate this to the entire business model. Is there any reason that we can't get the profitability within a year or 2? Sure. So I think this chart definitely lends itself to thinking through how the entire business works. I think there's a lot of things to cover there. So you can look at SG and A per unit you've outlined and obviously as you move to the right to older cohorts that gets significantly better. I think the other line that's on that chart is the total GPU line, which is a fundamentally important line because when that crosses over your total SG and A, you're at a spot where that market is effectively EBITDA positive. And so we've got those 3 cohorts that are performing very well and contributing positive cash. And then you can basically see the investments we're making in growth, which is happening to the left in the newer cohorts, which those cohorts are ramping faster than the older cohorts were at the same point in their life, but they are earlier in their life. And so they continue to require some additional cash before their cash flow or actually before their EBITDA positive. So I think that's a really good way to think about it. You can think about as we add more markets, we're shifting over to the left, but then as time passes, we're shifting to the right. All those markets are kind of aging and moving further and further to the right. As we make additional investments and improve GPU, that GPU line is moving up. And as we make investments to make our SG and A more efficient through different technology and everything else, the entire line is kind of shifting down. And so I think you can kind of think about all those different dynamics at play. I also think it's important to think about the weighting of those different cohorts. So those last 3 cohorts are positive. If we look at the market level, we have 14 markets that were EBITDA positive in the Q3. Those 14 markets were representing 35% of our sales. And that's a natural kind of overweight lean towards those older markets because they do tend to have higher market shares. But I think you can think about that as well. So yes, those are the 3 major dynamics in that chart that I think are really useful. We're always shifting to the right just with the passage of time. We're really focused on pushing GPU up. And then as we get more efficient with all of our different technology, we're shifting that entire blue line down. And I think that tells you how the whole business works effectively. Okay. Is there any reason why the GPU would be higher in any given market by cohort or the GPU is fairly consistent? They can vary. So they're not going to be precisely the same. But in general, a benefit of our model is that all markets are sharing the same pool of inventory. And so consumers are picking the same inventory, and they tend to select financing and warranties at very similar rates. So first order, the right assumption is basically flat GPU across all of our markets. The next question is from Ron Josey with JMP Securities. Please go ahead. Great. Thanks for taking the question. I want to ask about just the progress on building the brand online specifically. And I ask because when I look at your web traffic, I think in the Q, you reached 6,300,000 uniques in the quarter, which is pretty substantial growth. And as you get to that 95% coverage or shall we say market expansion acceleration to 2020, Can you just talk about those sources of traffic? I think SEO was a big focus at last year's Analyst Day. So trying to understand sort of where these uniques are coming from and maybe as a secondary mark, in your comments you talked about gains in retail GPU specifically some benefits from incremental shipping revenue. I'm wondering if that's delivery from cars outside of Carvana's markets and maybe you put that all together that's a lot of traffic and delivering outside of your current markets that talks about acceleration? Thanks. Sure. So let me hit traffic first, then Mark can roll over to hit the second question. I think traffic comes from many, many sources. I think the first place to start is we do have significant traffic now and I think that's very exciting, but we're also very, very small compared to the traffic in the industry. And there's just a lot of room for us to continue to build brands. So I think most importantly, we've got a real focus on brand building as being something that's really important to business to drive more traffic and just more customers who are aware of Carvana and understand what our offering is that we can continue to grow sales. I think that takes many forms. We're obviously advertising on TV. We're advertising across the vast majority of other channels that you would think of. And we're testing constantly which channels work better in which markets and which attributes of which markets tend to lend themselves to which channels. And we've got a really, really impressive team of highly quantitative analysts that focus a lot on that. And so I think that's something that we're very proud of. You specifically call that SEO. SEO is something that we've definitely been focused on for probably the last year or so, give or take. And it's an area where we are making progress undoubtedly. SEO is kind of an interesting area because there's kind of many steps you try to make. The first, you have to make sure that your site is crawlable and easily understandable by all the different search engines. The next thing you're trying to do is make sure you get all your pages indexed. Those search engines are kind of aware of what pages you have. And then over time, you build credibility for each of those pages. And really, the payoff doesn't come until you start to make it into the 1st page of search results. So we're seeing a lot of gains in the background, and I think we're starting to see some of the nice gains that actually do drive traffic, but it feels like it's still very, very early days there. And so I would say across the board, it's early. We're excited about our traffic. It is growing very fast. It's growing in lots of with our sales, but we also think there's just a ton of opportunity there across all the different marketing channels, across brand building and across SEO. Sure. And then on the point on, ink rental shipping revenue, I'd say over the last year or so, we've been testing in network shipping fees for cars that are typically far away from customers. It's relatively small, but it's been some incremental contributor to GPU. I think in the quarter, looking sequentially, we brought down shipping revenue in large part because of the inventories ramping in some of our Eastern IRCs, particularly in the Cleveland and Nashville. Got it. Thank you, guys. Thank you. The next question is from Lee Crowell with B. Riley FBR. Please go ahead. Great. Thanks for taking my questions. Just first on IRC expansion, you guys highlighted that you guys have five locations penciled out. Do you kind of expect to roll out those locations at the same cadence it took you to get to 8? And then my second question, just talking about cohorts in terms from a little bit different angle, but cohorts in terms of customers to which you source cars from. Curious how across the age of cohorts whether or not you see a higher uptake of cars from customers, just based on either branding or familiarity with the product? Sure. So first of all, in IRC cadence, what I would say on that is, it's really important for us to make sure that we stay ahead of our IRC pipeline because that's definitely the longest lead time part of the business and that's a big operational facility that we have to build to be able to support our growth. So we're very excited that we've invested in a lot of effort in preparing for the next several years of growth in inspection centers. And that team has many, many sites that they're looking at and then the ones that you called out were already making positive progress. I think in general, the way we think about IRCs, we want to make sure that we roll them out in anticipation of coming growth. So I think we need to be careful about telling you the cadence at which we plan to roll them out because I think that starts to look a lot like forward guidance. But what we're going to do is we're going to roll them out to support the growth that we anticipate coming. And then on across market or across cohort dynamics in buying cars from customers, I think generally speaking, we tend to see higher penetration rates or higher overall volume in our older cohorts, not unlike retail sales, although the patterns can be or the sort of overall levels can be slightly different. I think that's basically driven by brand awareness. We have more traffic coming to dotcom in older cohorts, more familiarity with our brand and that drives increased volume. Got it. Thanks. The next question is from Brad Erickson with Needham and Company. Please go ahead. Hi, thanks. Just a follow-up on the cash flows. Can you just walk through cash flow from ops where I think the burn kicked up a little bit quarter over quarter? Maybe just talk about some of the meaningful inflows and outflows there, particularly around the financing business and just kind of what we're seeing? Sure. Yes. So of the major drivers in cash flows for operations, I'd start with EBITDA, which we've always mentioned is one of the key drivers of our cash use. I think that another is you mentioned finance receivables. I think when you account for the securities that we retained as part of the securitization is roughly flat on total net change in finance receivables. Some of that will some of that flatness will come from change in financing activities where we finance those retained securities. Some of it will show up in the cash from operations. And then inventory has a small change, obviously, generally finance that with floorplan liquidity. And so I think for the most part, when you take into consideration the with financeability, that offsets a lot of the working capital cash flows. I think as we think about liquidity in the big picture, it's always really important to think about working capital line availability. I mentioned on the earlier portion of this call, taking fully into consideration available on our short term working capital facilities, as well as real estate on balance sheet, which gives us a lot of flexibility. The next question is from Colin Sebastian with Baird. Please go ahead. Hi, thanks. This is Dalton on for Colin. Just one question. When you laid out in the letter talking about the impact of IRCs in Indianapolis and Cleveland driving lower logistics expense and faster sales growth. How do you think about that in terms of the strategy of building out IRCs moving forward? And maybe how the boost to the local markets might be offset by some of the lower shipping revenues you called out from shipping from further away markets and kind of if anything you've learned in the recent build out influences any of your long term strategy in the IRC build out moving forward? Sure. What I would say is that I think it empirically reinforces it. So I think many of these questions about how many inspection centers should we build, how big should our inventory be, how much should we spend on marketing, there's a lot of positive feedback in all of these different areas in the business because we do share a single inventory. And then that means as we grow inventory, customers have access to larger selection. And as we increase inspection centers, they have shorter delivery time. And yes, as we invest in our brand, it gets less expensive to enter a new market. And so I think all these things feedback really positively. And I think what's great is just that we're seeing that empirical feedback show up in real world data points like looking at the Midwest where there weren't inspection centers nearby and then we went and placed them down and we saw the effects that we would expect to see. So I don't think it changes our strategy. I think it reinforces what we believed in the 1st place. And then I think in general, our goal is definitely to move as quickly as we can, as quickly as we responsibly can to continue to grow this model because we feel like we've got an offering that is exceptional, the customers love and the business gets better at scale. And so that's what we're seeking to do and we're basically just trying to go as fast as we can. We're making sure we keep the wheels on. Operator, we have another one. Yes. The next question is from Rajat Gupta with JPMorgan. Please go ahead. Hi, thanks for getting me back on the queue here. Just had a question on just one follow-up on the finance GPU. Clearly, the ABS securitization is increasing in your mix, but you still have the MTA and the MPSA agreements. Do you still think you would need to have those agreements longer term? Or are there any other avenues of monetization that are being considered apart from that? Thanks. Sure. Well, sorry, I think part of it we've had for a long time that we sold receivables too that we generated in our platform is Ally, and they've been a great partner for us for a long time, and they're a partner we really value. And so I think we view that as a mutually beneficial relationship for sure to continue to have that line open. It's nice to have access to the securitization market. That's probably the most efficient market there is for monetizing finance receivables, especially once you kind of build liquidity in that market and familiarity there. But there's something really nice about having another very high quality stable outlet through Ally. So at this time, we don't plan to shift everything into either channel. Got it. That's helpful. And just one last one. Your full year revenue guide implies flattish sales going from 3Q to 4Q. Are we to think that the business is entering more of a normal seasonal pattern now going forward based on what you're seeing in your different markets? Thanks. That would be all. Sure. So we're obviously very excited about our revenue growth in Q3. We're very excited to raise guidance for the full year significantly. I think we feel great obviously about unit growth and revenue growth. There are some small seasonal factors that play out in Q4. One example is the Cyber Monday promotion, which is $1,000 sticker price discount on cars that are sold during the promotion. So there can be some small impacts there, but obviously we feel really great about where we're headed. Got it. Thank you. Thanks. We actually have time for one more question from Sharon Zackfia with William Blair. Please go ahead. Hi. This is Tanya Anderson for Sharon Zackfia. I just have a quick question on the customer exhibition costs. They were up for the first time year over year this year. What were the reasons for that? I think there's probably 2 that I would point to. So one would be to make sure that you're thinking about customer acquisition costs, not just in terms of retail units, but in terms of total transactions. We did start to invest in advertising that we buy cars from customers and that doesn't show up in units. So when you kind of divide by units, you're missing that impact. And so I think that's an important thing to look at. And then I think that probably would have solved your question. But I think even beyond that, we did generate pinch points in the quarter as we saw all of this growth in buying cars from customers and generally just saw great demand on the retail side. And when we see those pinch points, delivery times go out, conversion rates tend to drop and so that kind of mechanically reduces your marketing effectiveness and increases your CAC. And so I think that would probably be the other impact. Okay. Thank you. This concludes our question and answer session. I would like to turn the conference back over to Ernie Garcia for any closing remarks. Thanks everyone for joining the call and thanks everyone on Team Carvana. We had another incredible quarter and it's only happening because of everything you do and all the passion and energy you bring to it. We really appreciate it. Please keep it up. Thanks a lot, guys. The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.