CVNA - NYSE

Good afternoon, and welcome to the Carvana Second Quarter 2025 Earnings Conference Call. [Operator Instructions] Please note, this event is being recorded. I would now like to turn the conference over to Meg Kehan, Investor Relations. Please go ahead.

Thank you. Good afternoon, ladies and gentlemen, and thank you for joining us on Carvana's Second Quarter 2025 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 second quarter shareholder letter is also posted on the IR website. Additionally, we posted a set of supplemental financial tables for Q2, which can be found on the Events and Presentations page of our 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 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. 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-K and Forms 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 metrics. Unless otherwise specified, all references to GPU and SG&A will be to the non-GAAP metrics and all references to EBITDA will be to adjusted EBITDA. 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 IR website. And with that said, I'd like to turn the call over to Ernie Garcia. Ernie?

Thanks, Meg, and thanks, everyone, for joining the call. The second quarter was another exciting quarter for Carvana. There are many financial highlights that we hit in the letter, and we will hit throughout the call, but I want to spend this time focusing on a subset and to discuss their implications for the future. We were once again the fastest-growing and most profitable automotive retailer, again by significant margins. Based on our best available data, the market grew by less than 5% in units in the quarter compared to our growth rate of 41%. We believe this growth rate speaks clearly to the desirability of our offering and our team's ongoing execution. When looking at our adjusted EBITDA margin, we once again set a new record for automotive retail and improved by 200 basis points year-over-year. This makes our model twice as profitable as other publicly reporting automotive retailers on this basis. Excitingly, not only with the most profitable adjusted EBITDA margin, but for the first time, we are also the most profitable as measured by GAAP operating income and net income dollars, another significant milestone along our path to becoming the largest and most profitable automotive retailer. Hitting these milestones and rapidly moving through the various definitions of profitability carry significant meaning. It means that when we set a completely different course for automotive retail 12 years and $10 billion ago, our underlying belief was correct. Customers were ready for something new and something tailor-made to serve their modern preferences generates a completely different customer response and completely different financial performance. That alone is extremely exciting and positions us very well for the future, but there's more to the story. Part of what is enabling this rapid growth at a scale of $4.8 billion of quarterly revenue and $500 million of quarterly GAAP operating income is that the market we are changing is enormous. We are currently about 1.5% of the U.S. used car market and approximately 1% of the total U.S. car market. We are also excited about the long runway and incredible potential we have. In addition, we benefit from unique competitive dynamics. Despite being so early in our maturation, we are already the second largest retailer of used cars with our eyes fixed firmly on becoming the largest soon. Our industry has structurally different and more favorable competitive dynamics than other large verticals, and we believe this bodes well for our ability to play a very outsized role in our industry in the long run. And lastly, what we are doing is hard. Hard is the ultimate competitive moat. Our business requires a complex mix of highly varied capabilities to deliver a simple and efficient customer experience. The difficulty of our business was a liability when we started, but today, it's a valuable asset. Big swings have always had that property, and we have always taken big swings. To drive our success over the long term, there are 3 primary areas where we will be putting our focus. Number one, driving significant growth over a long period of time. Number two, constantly improving the machine through fundamental gains across all areas of the business. Over time, we plan to share the majority of these gains with our customers, the same way many great consumer brands have before us. Number three, building additional foundational capabilities that will make our platform even stronger and will help us drive remarkable outcomes for our customers, our partners and ourselves over the long term. These efforts will continue to drive us toward our next goal of selling 3 million cars per year, 13.5% adjusted EBITDA margin in the next 5 to 10 years. The path from here is straightforward. To call on the metaphor, we've made it from 0 to 1. Now we're focused on a very big end. To get there, we'll remain ambitious. We remain focused on giving customers the simplest, most efficient, most fun and most satisfying experience we can give them, and we'll continue to work hard with people that we're proud to work alongside. The march continues. Mark?

Thank you, Ernie, and thank you all for joining us today. Our second quarter results once again showcased our team's ability to deliver fundamental improvements in operating efficiencies while also driving significant year-over-year growth. For the sixth sequential quarter, we earned positive net income, and we set new records for retail units sold revenue, adjusted EBITDA, adjusted EBITDA margin, GAAP operating income and GAAP operating margin. Unless otherwise noted, all comparisons will be on a year-over-year basis. Retail units sold totaled 143,280 in Q2, an increase of 41% and a new company record. Revenue was $4.84 billion, an increase of 42% and also a new company record. Consistent with past quarters, our growth in the second quarter was driven by our 3 key long-term drivers of growth. A continuously improving customer offering, increasing awareness, understanding and trust and increasing inventory selection and other benefits of scale. We believe as we continue on our path of profitable growth, each driver will improve, creating more positive feedback in the model. Our strong profitability results in Q2 were again driven by sustained and fundamental improvements in GPU and operations expenses as well as levering our overhead expenses. Non-GAAP retail GPU increased by $195. This change was primarily driven by reductions in reconditioning and inbound transport costs and an approximately $100 benefit from tariff-related impacts. Non-GAAP wholesale GPU decreased by $85. This change was primarily driven by faster growth in retail units sold than wholesale marketplace units, partially offset by lower wholesale vehicle depreciation rates. Non-GAAP other GPU increased by $126. This change was primarily driven by better cost of funds as well as a higher attachment rate on vehicle service contracts, partially offset by a positive impact of approximately $100 in Q2 2024 from selling additional loans. Q2 was another strong quarter for demonstrating the power of our model to lever SG&A expenses. Our 41% growth in retail units sold led to a $460 reduction in non-GAAP SG&A expense for retail units sold. The Carvana operations portion of SG&A expense decreased by $147 per retail unit sold, driven by our operational efficiency initiatives. The overhead portion of SG&A expense decreased by $328 per retail unit sold, driven by higher retail units sold. Advertising expense increased by $29 million or $44 per retail unit sold. On a sequential basis, advertising increased by $12 million. We believe we are still in the early days of automotive e-commerce adoption, and there is a significant opportunity to further invest in building awareness, understanding and trust of our customer offering. As such, we expect a larger sequential increase in advertising spend in Q3 versus Q2. We continue to see opportunities for significant improvement in per unit SG&A expenses over time and as we scale, driven by both continued efficiency and operational expenses as well as leverage in the fixed components of our cost structure. We continue to pair industry-leading growth with industry-leading profitability, not only by adjusted EBITDA, but also for the first time by GAAP operating income and net income. Net income was $308 million, an increase of $260 million. Net income margin increased 5 points year-over-year to an industry-leading 6.4%. GAAP operating income was $511 million, an increase of $252 million and a new company record. GAAP operating margin was 10.6%, a 3 percentage point increase and a new company record. Adjusted EBITDA was $601 million in Q2, an increase of $246 million and a new company record. Adjusted EBITDA margin was 12.4% in Q1, a 2 percentage point increase on a new company record. As previously discussed, our adjusted EBITDA is very high quality compared to many rapidly growing companies due to our relatively low noncash expenses, which will continue to lever with scale. We converted approximately 85% of adjusted EBITDA into GAAP operating income in Q2. This compares to adjusted EBITDA to GAAP operating income conversion of 73% in Q2 2024. As previously noted, we currently carry many expenses that support retail unit sales capacity of over 1 million units and expect our GAAP operating income to grow faster than adjusted EBITDA over time. Our results in Q1 and Q2 position us well for a strong Q3 and Q4. Looking forward, we expect the following as long as the environment remains stable. A sequential increase in retail units sold in Q3 compared to Q2, and adjusted EBITDA of $2.0 billion to $2.2 billion for the full year 2025, an increase from $1.38 billion last year. In conclusion, our results in Q2 were exceptional. Our team's focus is unwavering, and our opportunity remains clear. Thank you for your attention. We will now take questions.

[Operator Instructions] Our first question today is from Daniela Haigian with Morgan Stanley.

So my first question comes from incremental adjusted EBITDA margin. It came in at over 17% this quarter. Is there any reason why that wouldn't be indicative of future incremental margins?

Daniela, I think that's obviously a great number. I think it reflects the general leverage in the business and the improvement that we've had. As discussed, we improved EBITDA margin year-over-year by about 200 basis points, while growing at 41%. I think when that's true, your incremental margins, I think, have to be in a pretty good spot, and that was clearly the case. I think our goal is absolutely focused on getting to 3 million units and 13.5% EBITDA margin and I think we're excited by the potential of the business to keep getting fundamental gains to continue to print great margins and also to have significant value to share with our customers. So that's the plan, and we'll keep marching to it.

Great. And then, Ernie, second one, a bit of more of a longer-term question. I'm sure you saw Avis announced a partnership with Waymo this week as an autonomous mega fleet manager. While there's obvious synergies with Robotaxi and rentals and logistics, maintenance and infrastructure, ADESA gives you hub-and-spoke network across the country. You've built out real reconditioning chops with digital integration. You see opportunity for Carvana TAM to expand beyond used cars?

Part of the sickness that, I think, we probably have as we always see opportunity everywhere. But I think part of what we try to do is stay focused and figure out where is the best place to put our energy. So I think we're extremely excited by what we see in front of us. That just requires blocking and tackling and continually making the customer experience better and getting more efficient all the time. So that is absolutely our primary focus. We'll always be paying attention to everything else, but that's what we're focused on.

The next question is from Jeff Lick with Stephens.

Absolutely magnificent quarter. You guys are keeping track of that. So in our work, it appears this quarter, you did some experimenting or just some toggling with APR raising price, and I was wondering if you could talk about that and if that's indeed true and what you're seeing there?

Sure. I mean I would say I think the way that we try to focus on this is we built a vertically integrated machine that I think has a lot of levers. And then I think we try to put a lot of effort into making sure that we're building intelligence into those levers that make intelligent decisions in any given environment, try to make sure that we're always moving forward and then we try to always make those levers a little bit smarter and make the business a little bit better. I think you can see some movements sometimes in the various GPU line items, but I think the goal is to always make progress. And I think that's where we would kind of push your attention. I think quarter-on-quarter, you can see things jump around, but the overall results is what we're focused on.

And Mark, a question for you. Our work kind of shows you've got about a 300 to 400 basis point higher APR, but you're only -- maybe costing you 60 bps of 61 day plus delinquency. I'm just curious, I guess, how are you doing that? And what are the metrics that help you deliver that?

Well, I think what you're asking is about the strength of our vertically integrated finance platform. And I do think the -- I think one of the things that we've been excited about for a long time is, by vertically integrating finance and other parts of the transaction, it affords a lot of advantages versus a non-vertically integrated model. We've talked about some of those advantages over time. We intimately know the car that we're selling to the customer. We've done 150-point inspection on that, ensured its quality also added a 100-day limited warranty to make sure it's a great car that every customer is getting. We intimately know the customer because we're interacting with them directly rather through an intermediary, like is often the case in indirect finance. And so intimately knowing the customer is a powerful component of our vertically integrated business model as it relates to financing. I think, obviously, we take great -- we've invested a great deal in making the best possible use of our data. So we're rapidly growing the number of transactions that we've executed and all of the data that our finance platform is growing, that leads to better models over time, and that's a place where we've really placed a lot of focus. And so I think there's some real meaningful advantages of being vertically integrated, being very large scale, and those can lead to very strong outcomes in something like the finance platform.

The next question is from Brian Nagel with Oppenheimer.

Great quarter, congratulations. As we're looking the volumes here continue to strengthen, if you will. I mean how should we think about particularly over the balance of this year, the new reconditioning capacity coming online, whether -- what you're doing at your legacy IRCs or with these converted ADESA centers. So how should we think about the pace of the cadence of that? And then any commentary as we look at these recent trends as the capacities come on, to what extent that's helping to satisfy demand and bolster sales?

Sure. Well, I think at a high level, I think the simple answer is we're on plan. We feel very good about it. We grew sales by 41% year- over-year. We grew inventory available for our customers by 50%. So we're obviously growing production a little faster than we're growing sales. We now have 12 sites integrated with ADESA. We've been doing that at a pretty quick clip. We've been averaging on the order of 3 per quarter, give or take. I think we have a number of sites left to continue to integrate. I think the team is executing exceptionally well. You continue to see improvements across many parts of what they do that is leading to the gains that we're seeing in retail GPU. And I think excitingly, I think there's a lot more for them to do. I think we have this planned cadence where we aim for Q2, and we set a new set of projects, new set of goals and we aim for the next Q2. And I think looking forward in our reconditioning group and in all the groups across the company, I think we continue to have ambitious but achievable and exciting goals that are clearly articulated with clear underlying projects behind them. So I think we feel like we're on path. We've got our eyes focus down the field. It's a $3 million goal, and we're just going to keep moving as fast as we can towards that milestone.

That's helpful. I appreciate it. And then one quick follow-up. And I don't know if you addressed this in your comments or maybe I read it in the letter, too. But was there -- did you notice any type of demand choppiness just as consumers may be reacting one way or another to the tariff environment?

I would say -- at a high level, I think what matters the most is probably pretty consistent. I think we put a chart in our shareholder letter that shows that -- I think for the other public automotive retailers, the growth year-over-year was about 1% on average. I think last quarter, that same number was about flat. So I think that suggests that kind of the -- for the full quarter, the seasonal trends were pretty consistent. I think there was a little bit of kind of pull forward and then maybe a little slowness immediately thereafter. But I think for the most part, it was relatively flat. We called out $100 retail GPU impact due to some pricing changes we made as we are riding through that period. I think our goal there was to make sure the machine was just operating in as balanced a way as we possibly could keep it. I think there will be other dynamics and catalysts and little moves quarter-to-quarter in the future as well. And I think the most important thing that we can do is just try to keep the machine moving forward effectively and build tools that enable us to absorb whatever bumps and make whatever changes we need to and if we see them. So I think overall, nothing super material, a couple of little things week-to-week, but nothing that matter at the level of the quarter.

Next question is from Sharon

I wanted to dig in a little bit on the marketing side. So it completely makes sense you'd be reinvesting in more marketing. But I'm curious where your brand awareness now is nationally. And if you have a figure for where that might be in Phoenix? And then -- I'm sorry, not Phoenix, Atlanta. Phoenix too, if you want to give it to me. And then as we think about marketing, is this more brand building? Or is there a specific message that you're trying to get across in the new campaign?

Sure. Well, so I think let's start with what's the goal of all this. I think the goal of all this is to make sure we're laying foundations for outsized growth for a long time. That's the general goal. And I think the fundamentals are we are now at a spot where the gap between our total GPU and our operating expenses is very, very large. And I think that suggests that there are lots of opportunities to lay those foundations. I think one of the things you can compare that gap to is just our advertising expense. And I think there's clearly a big gap between our advertising expense and the difference between those 2 numbers. So that suggests opportunity. We run various surveys every quarter that we update to try to get a sense of where all 3 of the considerations. So awareness, understanding and trust is kind of the general frame that we use, but we try to understand where all that is. And I think what we see is we see constant progress, but we still see a lot of opportunity in all of those. I think there's significant opportunity in awareness and gets through across the country. I think it's true in many even parts of what we do. I think understanding there's even more opportunity. And then I think trust is mostly about making sure that we give customers a great experience one at a time. But I think we're leaning into marketing a little bit to test some things and to see how that works. I think there are many different marketing channels that exist on a spectrum from more direct marketing, where the response is pretty immediate to more brand marketing, where the response is long term. I think we're testing both. I think that marketing in general is something that's very difficult to precisely attribute to sales. I think it's a little easier on direct marketing and even harder on brand marketing. But as we discussed in the letter, we're putting out another brand campaign. We're testing a number of different channels. So I think we see that as one of the many areas of opportunity where we can continue to lay foundations for a lot of growth, and we'll see where that takes us, but we're excited.

But Ernie, do you have any metrics for what aided brand awareness would be for Carvana versus maybe some of your peers?

Of course we do, but we very purposely didn't provide them. And then you followed up and made it awkward.

The next question is from Brad Erickson with RBC.

I had 2. First, just for the capacity expansion going forward. Mark, this might be a few. Can you give us a sense of just kind of what that might look like in terms of facility integrations and line expansions and so forth and just how to think about investment necessary to do that? And then I have a follow-up.

Sure. Yes. There's lots of things to hit in that question. So I think the -- just to talk a little bit about our strategy. So core to are production expansion strategy right now is ADESA integrations. I think we've had great success with that over the last year or so, marching up integrated ADESA locations steadily over time. I think that is something that we intend to continue through the back half of the year, but also I would expect that we continue to march out integrations in 2026. Those are obviously CapEx light, integrating these ADESA locations. We're really utilizing existing structures at those ADESA facilities and just layering in Carvana, Carli technology to run the centers layering in Carvana retail reconditioning processes. So those are -- that's sort of the first phase. The next phase that we'll enter into as we proceed with our growth plan is to actually start building out some of those ADESA locations and really actually doing more fulsome build-outs of the locations to increase the capacity at ADESA. There, I think the way to think about that is back at the time of the acquisition, we basically gave a number, we thought it would cost roughly $1 billion to build out all the ADESA locations. That would obviously be something that would take place over a number of years. There's been some inflation since then, so it might be a little bit higher than that now, but I think that still gives a pretty reasonable sense. And then that would be something that would play out over time.

Got it. That's helpful. And then how should we think about kind of your ability to source vehicles from consumers and you have more of these IRCs up and running? Like does that ramp up kind of in a linear fashion? Or anything else we should be thinking about kind of in the equation for supply acquisition as you grow your capacity?

Sure. I'll take a swing on that one. I mean, I think it just makes things better and it makes the machine more efficient. I think we talked about in the letter, our inbound transport is down about 20% in terms of miles traveled. As we open more facilities that should continue to go down as we convert all these ADESA facilities, we have room for more than twice as many inventory pools as we have today. And when you think about kind of what that means for the reduction in distance of miles, the cars have to travel, that's meaningful, and that's kind of fundamental value that has to show up somewhere either in bottom line or in bids where we're sharing it with customers. So I think that that's one of the areas where I think that there's fundamental gains that are because we make the system smarter or do things differently and more efficiently. There's also fundamental gains that just come directly from scaling the system and getting the benefits of positive feedback that exists in the model, and I think that there are many of those in this area. I also think that we just spoke about advertising and brands generally a moment ago. I think a number that is pretty interesting is in the last 2 years, we've grown retail sales by approximately 80%, give or take. And if you look at, say, wholesale to retail ratio, it's been approximately flat at that entire time. And I think that speaks to the fact that these 2 businesses generally grow together, but they also have completely separable product pipelines, right? There's different things that the different teams are working on to make that those different experiences better. And I think those teams are both doing incredible things. But so far, we've remained very well in balance, and I think we're benefiting from continuing to scale and continue to open up additional sites. And as I said, I think that creates opportunity to share some of those gains with our customers over time, which we think can power growth even further.

The next question is from Andrew Boone with Citizens.

I wanted to ask about retail GPUs in the quarter, had a really strong quarter. Mark, I know you called out the onetime $100 benefit. But is there anything else you can help us understand there? And then, Ernie, in the letter, you guys mentioned word of mouth, understood you guys aren't going to disclose anything on awareness. But bigger picture, as you guys are just a bigger part of the car economy. Can you talk about the benefits of word of mouth and just having more people talk about the actual experience of Carvana is showing up in a driveway?

Sure. Yes. Let me start with the first question. So retail GPU was up about $200 year-over-year on a non-GAAP basis. That really breaks down into 2 primary categories. One was just improvements in reconditioning and inbound transport costs. I think that's an area where we've had a focus for a long period of time and saw a nice year-over-year gains there. I think some of those year-over-year gains were driven by continuing to integrate ADESA locations, which leads to lower inbound transport costs, which is one of the -- those retail cost of sales. And so I think just fundamental gains and reconditioning costs and inbound transport cost was about half of that year-over-year gain. The other half was an impact where we saw in April where our April retail GPU was higher than the GPU we saw in the months and the remaining months of the quarter. We think, overall, that impacted the quarter by about $100 for the full quarter as a whole. That higher April retail GPU really linked to the announcements of auto tariffs in late March that drove stronger demand and higher margins.

And then trying to hit on word of mouth. I mean I think that's ultimately, in many ways, I think, the name of the entire game. And we think building a better business as the foundation of word of mouth. I think there's a lot of stuff that we are doing and a lot of goals that we have, but I think one of the frames that we've used in the past, we want to give customers the best selection, the best experience and a fast fun fair experience and a low price. And I think that we've got the ability to continually build that machine that just provides all of those things. I think as we continue to grow, it's very reasonable that many customers across the country will have access to more cars on Carvana than they'll have to some of all dealers in their city or state. And if we do a good job making easy for them to find that car, they can find it more easily and get a simpler experience and a better price. We talked in the letter about a number of ways we're trying to make the experience simpler. Faster delivery times, fewer customers calling in when they call in, having quicker calls. That's all driven by the customer being in control and us building tools that make it to the customer doesn't need to call in or if it's the type of customer that doesn't want to call, they don't need to, but if they want to, we're there for them. And I think we gave an anecdote of a transaction we bought a car from a customer, where in 38 minutes, they went from getting a value on our site to getting money in their account. And I think you can imagine a version of that conversation that hopefully happens between millions of customers in the future that is they had every car they could possibly want. The prices were incredibly fair. And I got the car faster than they could have driven down to the dealership and come through the transaction myself. And I think if that's the conversation that's happening, I think we don't understand why we're not at least in the conversation for every single customer thinking about buying a car anywhere in the country. And I think -- that's our goal. But the foundation of that is build a better system that's simple, that allows customers to tell each other simple stories that are very compelling by making it fundamentally better for them. And so I think there's a ton of work to do. I think we've done a ton of that work, and we're very proud of the experience that we deliver and the progress that we've made in all these different dimensions. But there's still a ton more work to do. And it's still very early in that game. We're 1.5% of the used car market. So there's a lot of room to run and a lot to do, but I do think that, that story consumers tell each other is probably the single thing that matters the most, and we got to make sure that story is very compelling.

Next question is from Rajat Gupta with JPMorgan.

I think in the letter, you had a comment that you want to look at the business more holistically going forward, focused just in units and EBITDA, could you elaborate a little bit more on that? What that means, what those levers are? I think you talked about advertising as one. And maybe if you could just elaborate on what different kind of levers that you can manage within those different line items? And I have a quick follow-up.

Sure. Well, I mean, that's a big opening to run through because I think that basically every number that we tried to provide in the letter is a lever that we're looking to improve. So again, I'll go to customer experience, a fast, fun fair. Make the experience extremely fast, make it fun, make it fair, make sure that we're getting more intelligent all the time, make sure that we're utilizing all the data that our system kicks off to make better decisions about which cars we're buying and what price we're paying for those cars and how we're pricing those cars, how we're merchandising those cars. Build tools and make it easier for customers to find those cars. I think all of those are things that we're continually working on. And I think every one of those items are levers that we have. We try to build systems that are intelligent and as autonomous as possible. But obviously, that have our oversight. So we decide on key levels like overall pricing level of any given lever, but we try to be intelligent and use analytics to determine where those prices are placed on various cars. And so I think continually improving those tools is an opportunity we've got. I mean, I think there are so many things to talk about there. We want to give customers simple experiences with an efficient machine that's getting smarter all the time. And I think every single thing inside that system is a lever for us.

Understood. Yes, I figured towards the and that it could have framed it differently. The other question was just on the cohort data. In the past, you've given us an anecdote around the Atlanta, some of the early cohorts. Anything incrementally can give us how the Atlanta or like some of the 2013 or '14 cohorts performed relative to the overall company this quarter?

I think we'll kind of stick with what we've said in the past, but all the trends that we discussed in the past remain there. I think it's been pretty consistent, broad-based progress across the country and across cohorts in every way.

The next question is from Michael McGovern with Bank of America.

On the really strong leverage in operations cost per unit, can you provide a little bit more detail on what exactly goes into that bucket of costs? Is that mostly labor and logistics? And what else is driving the bulk of that operational efficiency?

Sure. Yes. So let me provide a quick breakdown. So I think some of the key categories of expenses that are in that operations expense line item are fulfillment expenses. So that would include our multicar hauler network that connects our IRCs to the markets where we serve our customers as well as our last-mile delivery network that delivers the cars to the customer store. It also includes customer care and title and registration expenses as well as limited warranty expenses and some miscellaneous other expenses as well. So those are some of the major categories. I think that over the last couple of years, we've certainly made gains across the board. I think the -- we've talked about some of the efficiency gains in fulfillment, for example, that come from technology, process as well as adding things like adding ADESA inventory pools, which puts cars closer to customers. We've talked a bit about some of our early efforts in using our large data sets to fuel AI models. I think we're still in the relatively early days of that. But some of the early applications have been in customer care, and document processing, where AI can really make us more efficient in the way that we communicate with the customer, improving customer experience and also lead to some cost efficiencies as well. And so those would be a few of the examples. I think operations expense per unit, it's an area where we've seen really strong gains over the past couple of years, but an area where we do see opportunities looking forward as well.

Got it. And you also called out you grew selection of 50% in the quarter versus last year. Can you provide a little bit more detail on what that means? Is that just like a number of vehicles in your inventory? Or does that mean kind of breadth and depth and different makes and models or maybe more inventory of specific models that are really in demand? And where does that stand today relative to where you want it to be?

Sure. So I think the -- what we mean by that is really just the count of units that are immediately available for sale on the website as a measure of selection. Typically breadth increases with inventory count as well. So that metric specifically points to just inventory count of immediately available units, but breadth typically increases as well with that metric. In terms of where we want it to be, I think we still have a very clear opportunity to provide more selection to our customers. We view selection as a powerful long-term growth driver that interacts with some of our other growth drivers to create positive feedback cycle is one example of that. The more selection you have on the website, the more customers you convert, the more efficient your advertising becomes. And so the more you can leverage advertising to build understanding awareness and trust. So that's just one example of the type of positive feedback that selection can't create. I think in terms of our -- where we stand versus the long term in our selection, I think we're small compared to what we ultimately want to be. I think the used car market has a very unique property where there are a very large number of unique SKUs when you combine year, make, model, mileage level, trim, set of packages, exterior color, interior color, material and so on and so forth. You really get down to a very large set of unique combinations. And so compared to that very large set of unique combinations, our inventory today is small. And so we really do view growing our effective SKU count, i.e., growing our inventory and the breadth of our inventory as a powerful long-term growth driver.

The next question is from Chris Bottiglieri with BNP Paribas.

The first one is, can you talk more about the large build of inventory in the quarter? Does this primarily relate to the expansion selection or the change in the agreement with the commercial party just have an accounting impact that drove that balance higher? I just want to think if it's growth or just accounting that drove that?

Sure. Yes. So there are 3 drivers that our inventory growth in the quarter. The first was just sales and selection growth, as you just alluded to. The second was a change in essentially the contract structure with a large retail marketplace partner that has us holding the inventory on our balance sheet rather than the partner holding the inventory on their balance sheet. And then the third that I would layer in is we did see our average selling prices/average cost of our vehicles increased in the quarter, which also was a driver of that quarter-over-quarter increase in inventory.

That actually leads me to my next question. Can you kind of talk about that mix shift into more expensive vehicles and kind of what's going on there? Are you -- I assume you're after more prime customer, are you doing anything differently in terms of the rates that you charge prime and shipping fees like -- or is this just an inventory initiative right now?

Sure. Yes. So this is one of those areas that relates to Ernie's answer earlier, where I think we're not particularly focused on specific ASP. We're really focused on units and driving strong company level outcomes. But we have some mix shift into more expensive vehicles. That's largely just driven by our algorithms that are tracking real-time demand trends, tracking real-time supply trends. And then selecting inventory on the basis of what they're seeing in supply and demand. That has led to an increase in ASP. We called out that we do expect a further increase in Q3 beyond where we were in Q2. But it wouldn't surprise me at all if after Q3, we saw it declined. I think it really just depends on what are our models are telling us at any point in time. And I think most importantly, we're really focused on some of the bottom line metrics like growth in units and then also, obviously, our profitability metrics.

Your next question is from Michael Baker with D.A. Davidson.

Congratulations guys on a good quarter. Can I ask about the guidance for the back half of the year. It implies about, at the midpoint, 30% EBITDA growth. So obviously, a little bit of a slowdown from the first half. I presume it's just the law of large numbers. But does it signal increased investments? You already talked about increasing the advertising investment, but does the signal increased investments around price or bids or anything else along that or as we just simply we're coming up against big numbers last year?

Sure. Well, I think just to reiterate -- I think our guidance was for sequential increase in units in Q3 versus Q2 and then for $2 billion to $2.2 billion in EBITDA for the full year. I think we're not going to give a ton more color than that, but obviously, those are big numbers and exciting numbers. I think we grew by 41% in Q2. Last year was a record EBITDA year for us. It was incredibly exciting at just shy of $1.4 billion. So I think those are big numbers that suggest we're on a good path. And so we're just going to keep marching. I think that's another milestone on the way to $3 million, which is a milestone on to wherever we end up after that. So we're just going to keep trying to march through these various gates and make improvement along the way.

Okay. Fair enough. Maybe one other one, and it will probably be the same type of answer. But to get to $3 million in 5 years, I understand you have a range of 5% to 10%. But to get there in 5 years, off of 2024, that's like a 48% annual growth rate. You're having a great year this year, but I think your unit growth rate year-over-year is up like 43%. So again, do you have to like add some investments? Or how do you get that to accelerate to get to that $3 million in 5 years?

I think at a really high level, the 5-year time line was approximately a 40% compounded growth rate and the 10-year time line is approximately a 20% compounded growth rate. We just printed 41%. We're excited about that. I think very importantly, we're 1.5% of the used car market and 1% of the overall car market. So I think there's a lot of headroom. I think our machine is getting simpler. We're adding additional locations to hold inventory and to recondition inventory. We're making it so there's less work per transaction. And I think all of that aids growth, and we're trying to push back value into the customer offering, which makes those stories customers tell even better. So I think we feel like we're on a very good path. There's no question that growing at 40% for 5 years is an ambitious target, but we're an ambitious group, and we're going to try to get there somewhere between that 5- and 10-year target and we're just going to go as fast as we can along the way.

The next question is from Chris Pierce with Needham.

Just a quick question on other GPU, 2 of them actually. Just first, in the quarter, you talked about a higher VSC attach rate. I just -- I'd love to hear some comments about where you are on attach and warranty attach in general, how difficult it is to attach warranty online versus in person? Or maybe that's not true, just sort of runway and open space you have on warranty attached broadly?

Sure. Yes. So I think the attach rate of ancillary products and I would include BSE and also other ancillary products in that is definitely an area where we've made progress over time. And through constant testing and iteration, I have been able to identify wins the quality of our communication of those products, the way those products are structured, things like that. I think despite the gains that we have made, I certainly think there's opportunities for future fundamental gains really by running the same playbook on just constantly iterating and seeking to get better. One of the areas where we're always getting better is each year, we have more and more data. So we have more and more observations to help us evaluate what exactly do our customers like in these products? What are they looking for? And that helps us improve attach through communication and product structure, things like that, that I mentioned previously. So to summarize that, it's definitely been an area where we've made gains. Customers definitely like and get value out of these products, but it's also an area where we see opportunities for future fundamental gains just like in other areas of the business.

Okay. And then just a real big picture question on gain on loan sale. If we think about -- before Carvana came on the scene, there was a pool of ABS investors, yield-seeking investors who want to own auto loans. You guys have uncovered new investors in this space. Is sort of the obvious conclusion, the correct one that as you find more of these investors, yourself and other auto dealers will have more pricing power on the loan side of the world, and that is a tailwind to other GPU? Or is that too simplistic because of your market share?

That's an interesting question. I think we often hear a related but different question as well of are we going to outgrow that market. And so let me let me just try to give you the way that we at least think about it. I think we are a small part right now, 1.5% of a large mature market where that large mature market kicks off a lot of very high-quality consumer loans in the form of auto loans. And those auto loans are being purchased by someone today through various different channels. And I think as we take market share and displace some of those originators, there are still hands that are hungry to hold those loans. And so I think we've been, in many ways, expanding that market as we've been kind of growing into it. I think we've also been expanding the buyer base more generally than just the ABS market. Sometimes we meet buyers through the ABS market that we then move on to pool sales, and it can go the other way as well. So I think our focus is, we believe that the receivables that we're generating are very high quality. We believe it's highly desirable to a lot of investors, many investors that have traditionally been buyers of auto loan assets and probably a number of investors that maybe traditionally haven't had access to them. So if anything, we think that there's room to expand the total buyer base for auto loans, and we think that, that could bode well. And I think we found that as we've gotten bigger and we've established more of a brand and people have seen more of our performance history, it's gotten easier to attract additional buyers. And as we've gotten bigger, it's gotten easier to be a meaningful partner to more buyers because we can originate enough volume to be meaningful for them just by the work that they do to start the relationship and evaluate a pool of loans. So I think as a general matter, that's another area where we're very optimistic, and we think our unique model and being vertically integrated, where there isn't slippage between the incentives of the retailer and the incentives of the finance company is a valuable thing, and it's allowing us to originate lots of very valuable receivables.

Okay. Just you actually said it. You might outgrow that market. How would you outgrow that market though? Because you're just selling a car that someone else would have sold and another loan buyer would have bought that loan. So how -- I just want to make sure I understand where you're coming from with that comment, and then I'll pass it on.

Yes. No, sure. That was not my intention. I think I was trying to point to a question that we sometimes get. I think our general belief is that we will not outgrow the market. We think that there are already buyers of auto loans through many different channels. And we think that the manner in which we originate auto loans and the way that we make those liquid and available to investors if anything, should expand the total buyer base for auto loans. And we think we have access to the traditional buyer bases as well. So we have found empirically as we've moved through orders of magnitude of scale that it has gotten easier to sell loans, not harder as we've got larger. And I think that our best expectation is that will continue to be the case.

The next question is from Marvin Fong with BTIG.

Congratulations on a great quarter. Just maybe 2 around other GPU also looking into that topic. But you referenced a lower sell- through rate in the quarter. Was that related to, I think, another player in the space in commented that there was more cash buyers that kind of came up because of the tariffs in the quarter. Is that related what you saw? Is that related to that dynamic? And has that sort of cycled out of what you're seeing? And then secondly, you cited improved cost of funds as benefiting other GPU. We just kind of love to just understand, obviously, we can always drive that lower. But if we sort of level set maybe in a 1- to 2-year horizon, how much more benefit do you think you can extract from lower cost of funds? And how could you kind of give us a relationship to how that translates to actual GPU? That would be great.

Sure. Yes. Let me take that one. So I think on the lower sell-through rate, I think what we're calling out there is actually really an impact in Q2 2024 more so than an impact in Q2 2025. So in Q2 2024, we had stored some more loans in Q1 that we ended up selling in Q2. So we actually sold more loans than we originated in Q2. And I think that had a positive impact on the order of $100 per unit in Q2 2024. This quarter was relatively normal, give or take, a small amount in terms of the ratio of loan sales to originations. And so nothing really to call out there in this quarter. That was really about approximately a $100 impact back in Q2 2024. That was a positive impact in that quarter. The question about cost of funds. So I think what are some drivers of cost of funds. One I think is the number of loan investors and buyers that we're selling loans to in the finance platform. That's been something that has steadily increased over time as more and more investors become aware of the quality of the assets that we're originating, do the relevant research and start to invest in or buy the loans. So I think expanding the pool of buyers is a driver of cost of funds. Other drivers of cost of funds, I think, are just continued strong performance, and we talked about this earlier in the call, but our origination platform has generated assets that have performed very well, offering very strong returns to investors. And so I think continued strong performance from a vertically integrated platform is a driver of cost of funds gains as well.

Your next question is from Alex Potter with Piper Sandler.

So you mentioned that 40% CAGR, if you're able to sustain that. Obviously, you've been doing very well recently, that sort of implicit in your comments earlier about that being a difficult thing to achieve in 5 years is that, obviously, 40% growth kind of year in, year out over time, something could break, right? I mean it's difficult operationally to sustain that sort of growth. So I'm wondering, obviously, you're growing at that pace now. What is it that you think in your system could break? Is there anything getting close to breaking with you sort of redlining at the top end of that growth range right now.

Sure. I think it's a good question. First, I'm not sure today it feels like we're redlining. And I don't mean that to imply that we're going to Immediately grow a ton faster. I mean that to imply that it feels like the teams are executing very well and I think, confidently and comfortably. So I think we've talked in the past -- as a general matter, I think to use that term break, the things that are most likely to break are the things where you have the most work to do, the most people in a system to coordinate or the most stuff to move. And so I think reconditioning is probably the place that is operationally the most intense in the business. I think what's great is we are laying foundations for the future today. And in many ways, we're making investments in future growth today. What we easily could have grown into the existing inspection centers that we have in the Carvana network and that could have supported the growth that we're seeing today. But over the last year, plus a little, we've integrated 12 ADESA sites, that's been an investment in certain ways. The payoff has been that we've added additional inventory pools until we have less miles traveled. The investment has been that we've had to open those sites higher at those sites, find managers for those sites, run those sites at lower utilization, they tend to be a little bit more expensive. So that's flowing through our results today, but it also means that there's higher utilization to come in the future, and that will be a tailwind to results and we have more locations to hire more people and produce more cars in the future. So I think what we're doing in reconditioning, which is probably the operation most difficult thing today, is we're not only supporting the growth that you see. We are also laying foundations for easier growth in the future. And I think that, that's important and exciting. I think in logistics, we've made a ton of gains over the last couple of years getting more efficient and causing cars to travel fewer miles. I think logistics is probably the second most complex operational undertaking that we've got. And given that we've recently made a ton of gains in reducing total miles traveled, I think getting to a spot we're supporting high levels of growth over a long period of time is work, right? The total amount of miles driven has not grown as fast as sales have grown over the last couple of years. So we got to make sure that we're in front of that. And we've got plans, and we're working hard, and that's a very capable team that has a very clear plan for how they're going to continue to do that. I think market ops is probably the next most complex operational area. I think we've got great plans there. I think that team is also executing incredibly well and making a lot of progress and has recently been putting a little effort into outgrowing our growth to enable a little bit faster delivery times in one of the areas where we're back value to customers. And I think that, that's going extremely well. And then I think in customer care, we shared a number of data points where we're making very rapid progress there and getting more efficient. And I think that, that means over a long period of time, the amount of growth in people and things moving around can be less than the growth that we're showing as a company as we continue to get more efficient. So I think we've got a good plan there. Real life is always hard. And so there are going to be bumps in the road. There's going to be stuff that comes up, and we're going to perform great sometimes and worse than we wish other times. But I think we've got very capable teams and clear plans, and I think we're going to go, like I said, as fast as we responsibly can is our plan.

Next question is from Ron Josey with Citi.

Two please. Just can you remind us, Ernie and Mark, just when mega sites or the ADESA sites come online, just how quickly are these sites up -- get up to speed to become efficient? And when I mean efficient, meaning on par with maybe the IRCs out there. And then just, Ernie, you talked a little bit about improving processes internally. But in the letter, you also talked about improving e- commerce experience, making it easy to use maybe more fun. Would love to hear your thoughts or just maybe highlight there's one or two changes that have had an impact on improving those conversion rates.

Sure. Well, so I think first, it's going to take a little bit of time for the ADESA sites to get up to the efficiency level of the existing Carvana sites, I mean, as a very high level, kind of estimate. The Carvana sites are a little bit less than half utilized today versus their facility capacity. And the ADESA sites are -- that are up today are less than half of that. So generally speaking, I think utilization is a good first order measure of how efficient they're going to be in cost. And so today, the ADESA sites are more expensive and it is a bit of an investment to lean into growth in those sites versus just leaning into it in the Carvana sites. But for the reasons discussed earlier, we think that's smart and we expect to continue to get better there as we do scale those sites out over time. I think the same is true of logistics. By having those sites, we have fewer miles to travel. But we also -- because there's less utilization at those sites, we have more expensive cost per mile traveled. That will again normalize as we get up to utilization rates that are more similar to the Carvana inspection centers. So I think catching up is something -- it's kind of like an asymptote that we catch up to that will take some time. That is not something that happens immediately. But I think so far, we are on plan with our expectations or a little bit better as a general matter. And as I said, the teams are doing very well. So I think we're excited about what's going on there. And I think the fact that we're showing the gains that we're showing while making those investments and making future growth easier, I think means that the fundamental gains that the business is actually achieving are a little greater than are showing up in the financials. So I think that's great. I think as it relates to making things fun, I think the last couple of years are -- we've had a major focus on making the business more efficient. That makes customer experiences faster and simpler, and it also makes the business better from a financial perspective and more efficient overall. I think we still have a number of areas where we can continue to improve there. But I think that you calling out fun is absolutely correct. And I think also kind of minimizing anti-fund in the form of when we make mistakes. And how do we make sure that if we make a mistake, we do right by the customer, we treat them incredibly well. I think there's a lot of interesting ideas that we're working on to continue to make the experiences that our customers have faster and more fun, but I think we're going to keep those cards closer to our vest and maybe talk about it more in retrospect than we do ahead of time.

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

Perfect. Well, thanks, everyone for joining the call. Carvana team, awesome job. Truly incredible. I feel like I just keep saying the same thing every time on these calls, but the results that you've been able to put up over the last couple of years are something that nobody could have foreseen. I think this quarter is another example of it. Thank you so much. I hope that you're proud. I hope you find one another tonight and give each other over aggressive high fives with some serious eye contact because you've absolutely earned it. And I think we've got a lot more to do. We've got more high fives to earn. So let's put our heads down tomorrow and go do it again. Thanks, everyone.