Carvana Co (CVNA) 2021 Q4 法說會逐字稿

Good day, and welcome to the Carvana Fourth Quarter 2021 Financial Results Conference Call. (Operator Instructions) Please note this event is being recorded.

I would now like to turn the conference over to Mr. Mike Levin. Please go ahead.

Thank you, Chuck. Good afternoon, ladies and gentlemen, and thank you for joining us on Carvana's Fourth Quarter and Full Year 2021 Earnings Conference Call. Please note that this call will simultaneously be webcast on the Investor Relations section of the company's corporate website at investors.carvana.com. The fourth quarter shareholder letter is also posted on the IR website. Additionally, following the announcement of our acquisition of ADESA's U.S. physical auction business from KAR Global today, we posted the press release and slide deck on the Events & Presentations page of our IR website, where more details can be found.

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 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

(technical difficulty)

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. Unless otherwise noted on today's call, all comparisons are on a year-over-year basis.

Our commentary today will include non-GAAP financial measures. 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.

And now with that said, I'd like to turn the call over to Ernie Garcia. Ernie?

Thanks, Mike, and thanks, everyone, for joining the call. 2021 was an extraordinary year for Carvana, full of landmark accomplishments along our path to achieving our mission of changing the way people buy and sell cars. We started the year being named to the Fortune 500, tied to the third-fastest company ever to do so. We sold our 1 millionth car through organic growth faster than any other automotive retailer in history. We had our first positive earnings quarter. We had our first positive EBITDA year, excluding onetime items, and we became the fastest-growing e-commerce company in U.S. history.

While I'm extremely excited and proud of the team for that blockbuster list of accomplishments, I'm even more excited about the fact that we were also named to the top 10 of Forbes Best Large Employers to Work For list. We're ranked ahead of any other retail company and ahead of any other technology company. That set of accomplishments can only be achieved with a clear customer-focused mission and enormously scaled opportunity and the compounding that results from the efforts of great people who care are directed toward long-term value creation. Thank you to everyone inside Carvana for always choosing to care and for always giving that little extra effort that adds up to make a huge difference.

We've always framed our opportunity simply. Customers desire the experience we provide for them. Providing it is hard. There are approximately 40 million used cars sold every year in the U.S. The unit economics of the industry viewed over any reasonable time frame have been stable for a long time. That unit economic stability at the industry level is structurally driven by the fact that there are tens of thousands of dealers out there providing customers with similar customer experiences and who share similar cost structures. That simple frame is clarifying. To achieve our goals, we must continue to deliver great customer experiences, we must continue to differentiate our unit economics, and we must continue to scale. That is our path.

In alignment with that path, we are extremely excited to announce the acquisition of ADESA's U.S. operations. ADESA is the nation's second-largest auction company with 56 locations, completing about 1 million auction transactions a year. This acquisition has 4 primary justifications. Number one, it solidifies our path to becoming the largest and most profitable automotive retailer. Number two, it provides us with a nationwide inspection center network that we estimate will increase our production capacity by approximately 2 million units per year when fully built out.

Number three, it substantially improves our logistics network. Demonstrating the breadth of these locations, we will move from currently having inspection centers within 200 miles of 56% of the U.S. population to eventually being within 200 miles of 94% of the U.S. population. Demonstrating the quality of these locations, we will move from being within 50 miles of 16% of the U.S. population to being within 50 miles of 58% of the population. This will reduce shipping times to our customers nationwide and lays the foundation for eventually offering same-day delivery to many of our customers.

Number four, it significantly increases our auction capabilities and kick starts or deepens our relationships with many large and important players in the automotive industry. We look forward to working with our new customers to creatively finding new and interesting ways to work together and to valuing them in the same way we have always valued all of our customers.

We also expect ADESA to accelerate our path to our long-term financial model given the many powerful benefits that derive from being closer to our customers. For perspective on how powerful proximity can be, sales that are delivered to customers from an inspection center within 200 miles of our customers today have unit economics that are about $750 better than our average transaction as a result of lower inbound costs, lower delivery costs and higher conversion rates leading to lower customer acquisition expense.

Now I'd like to turn briefly to the current environment. Starting in the late fourth quarter, we, like everyone else, were hit pretty hard by the Omicron variant. At different points in time, we had up to 30% of our people in various operational teams simultaneously called out. It is obviously very difficult to deal with in any system. But in systems that rely on chained activity like our inspection centers and our logistics network, it is even more difficult. This led to the most severe logistics network constraints we have seen in our history. While we are largely out of the Omicron wave, it takes time to work out of our backlogs, and this year's severe winter storms have slowed our progress. Today, we remain severely constrained, but we're working hard to work through it as soon as we possibly can.

These constraints, paired with the recent rapid appreciation of vehicle prices as well as rapid increase in interest rates, have colluded to make this a challenging time. While this has undoubtedly been complex operationally, as our team has in the past, they are rising to the challenge. We've managed to grow our inventory available to our customers. We've grown our operational capacity to handle more volumes throughout our operating groups in anticipation of alleviating our logistics constraints and in anticipation of tax season. And we have made changes to the mix of cars we are purchasing and reconditioning to help our customers find more affordable cars despite the car pricing environment.

And our long-term indicators continue to look incredible. In our oldest cohort, we are now up to 3.5% market penetration. This may not sound like much at first, but if we apply that nationwide, it extrapolates to about 1.4 million annual sales. On top of that, our oldest cohort grew by 50% in the last year and therefore clearly has a long way to go. Lastly, 95% of our 311 markets are ramping faster than our 2013 cohort was at the same time in its life. 2 million-plus car sales per year is no longer the goal we're aiming higher. We have the team, the business model and the ambition to do it. The march continues. Mark?

Thank you, Ernie, and thank you all for joining us today. We are pleased to report another year of strong growth in both retail unit sales and revenue. Revenue totaled $12.8 billion, an increase of 129%. And retail units sold totaled 425,237, an increase of 74%, making us the fastest used automotive retailer to sell over 400,000 vehicles in 1 year. For the fourth quarter, retail units sold totaled 113,016, an increase of 57%. Total revenue was $3.8 billion, an increase of 105%.

Our exceptional growth in 2021 was driven by rapid growth within our market cohorts. In Q4, our 2013 through 2020 cohorts grew retail units sold by 52%, and our oldest cohort of Atlanta grew by 51% to 3.53% market penetration. As of Q4 2021, 95% of our 311 markets are ramping faster than Atlanta was at the same age, and record numbers of markets are crossing key market penetration thresholds.

In 2021, we completed our eighth consecutive year of $400 or more GPU improvement, our eighth consecutive year of EBITDA margin leverage and our first full year of positive EBITDA margin, excluding onetime items. Total gross profit per unit for the year was $4,537. Our growth in GPU in 2021 was broad-based, including increases of $166 in retail, $308 in wholesale and $811 in other. Total GPU in Q4 was $4,566, an increase of $1,187 year-over-year. Year-over-year changes in GPU were driven by gains across all components in Q4 as well.

Retail GPU increased by $230 in Q4, primarily driven by an increase in the percentage of retail vehicles sourced from customers, partially offset by higher reconditioning costs and higher wholesale acquisition prices. Wholesale GPU increased by $441 driven by gains in buying vehicles from customers and wholesale market appreciation. Other GPU increased by $516 driven by strong finance execution and higher industry-wide vehicle prices on average loan size as an additional impact.

In Q4, EBITDA margin was negative 2.5%, including a 0.6% impact from onetime items, an improvement of 1.4%, reflecting gains in both GPU and SG&A leverage. We ended the year with $2.3 billion in total liquidity resources, giving us significant flexibility to execute our plan.

Today, we also announced that we have signed a definitive agreement to acquire ADESA's U.S. physical auction business. The acquisition is expected to close in the second quarter of 2022 and will be financed with $3.275 billion in committed debt financing to fund the $2.2 billion purchase price and an additional $1 billion in improvements across the 56 sites. Upon development of these 56 sites, we are expected to unlock approximately 2 million units of incremental annual production capacity at full utilization.

This year, we made significant progress scaling our vehicle production capacity. In 2021, we added 3 inspection and reconditioning centers, increasing our total IRC count to 14. Following year-end, we also opened our 15th IRC near Cincinnati, Ohio, bringing our total capacity at full utilization to approximately 880,000 units as of February 24. We remain on track to open 6 additional IRCs by the end of 2022. We plan to open 5 of these on schedule, leading to more than 1.2 million units of annual capacity at full utilization at year-end. And with our acquisition of ADESA U.S., we are currently evaluating our preferred time line on opening the sixth.

In 2021, we also opened 45 new markets, bringing our year-end total to 311. With these new market openings, we now serve 81% of the total U.S. population, up from 74% at the end of 2020. We will continue to expand in 2022 and continue to expect to serve 95% of the U.S. population over time.

As we look toward 2022, we expect another strong year on retail units sold, revenue, GPU and EBITDA margin. We expect to grow retail units sold to over 550,000 for the full year. Following the first quarter, in Q2 through Q4 taken in aggregate, we expect total GPU over $4,000 and approximately breakeven EBITDA margin. We expect the first quarter to be impacted by supply chain challenges brought on by the Omicron variant and severe winter storms and the recent rapid increase in short-term interest rates. We expect these effects to have a significant impact on Q1 total GPU and SG&A per retail unit sold, leading to an expected EBITDA margin loss in the mid-single-digit range.

Since becoming a public company nearly 5 years ago, we have made tremendous progress toward our long-term goals and achieved many major milestones such as achieving our first profitable quarter this year. We're excited about what comes next.

Thank you for your attention. We'll now take questions.

(Operator Instructions) And the first question will come from Zach Fadem with Wells Fargo.

So over the second half of 2021, your business had been averaging roughly 8,600 unit sales per week. And this number is obviously constrained by a host of factors that appear to have worsened in January. But now that you've opened up 2 new facilities and a few more on the way in the upcoming months, can you talk about how this weekly run rate should change as the new capacity comes on and how this dynamic fits into the '22 unit outlook?

Sure. So what I would say is I think the primary driver is going to be how quickly we can alleviate the constraints in logistics network. Yes, as I said in my prepared remarks, Omicron really set us back. The logistics network is a system where many drivers connect in order to deliver a car to a customer. And at kind of the peak of the Omicron wave, we did have very high call-out rates, and that made it hard for those connections to occur and then set us back and got us a bit behind.

And then I think we've been working hard. The team has done a great job trying to catch up. We've got a really strong network when it's fully functioning that should enable us to catch up relatively quickly. But we've also been hit by a number of winter storms. As recently as this week, we are hit by another one. And those storms definitely set us back again.

So I think when you look at the remainder of this year, I think the primary driver right now is our logistics network, which is much more constrained than any other part of the business. And then I think from there, it's about catching up, and then it's about figuring out where we are for the remainder of the year. Right now, the logistics network is absolutely constraining sales. We have also seen car price appreciation that was pretty meaningful in the back half of Q4. And then we've seen interest rates go up quite a bit as well.

And so we don't feel like the impact of that has been that severe on us just yet because we've been kind of impacted by the logistics network constraints, and that's really what's holding us back. But we have seen industry-wide sales drop by kind of 15% to 20% relative to 2019 or 2021, depending on where your reference point is. And so we're taking that all into account and trying to put together our outlook of 550,000 units, and that's what underlies that.

So Ernie, just to be clear, should we assume that the average weekly units in Q1 is below Q4 and then improves sequentially throughout the year?

So I think we're going to stick with what we've got in the outlook in terms of specificity. I do think Q1 is going to be a tougher quarter. That's just the way it's going to unfold. We've been heavily constrained, and that's going to show up in the volume. It's also -- in the inspection centers that were hit really heavily in Q4 with Omicron, they were still able to produce a lot of cars because we built a lot of inspection centers and built out a very strong capability there. But the efficiency certainly went down, and that will show up and impacts the COGS.

So I think those impacts will show up in Q1, and that's why we're trying to guide to Q1 to provide kind of clarity there. And then we expect to emerge from that and look forward to a really strong Q2 through 4.

The next question will come from Sharon Zackfia with William Blair.

I got excited when I heard Zach. I don't know if you got that. On the ADESA acquisition, can you talk about the buy versus build kind of analysis you went through there, what you assumed for the attrition potentially of customers of ADESA as they might not want to kind of line your pockets, some of them? And then the $1 billion in improvements, on unlocking the 2 million in incremental capacity, what's kind of a reasonable time line to get there?

Sure. Well, I think, first of all, as you can imagine, we're extremely excited about this acquisition. We're extremely excited to welcome the team. We think it's a pretty incredible fit. I think whenever businesses come together, you always talk about all the synergies and all the things that you hope to occur. But I think what's so great about this one is it's very clear to see how we fit together. They've got an unbelievable footprint of 56 locations around the country, in great locations that are very close to our customers. You add up those locations, it's over 4,000 acres. For perspective, if you add up all of our inspection centers today, it's probably close to around 1,000 acres, so a really meaningful increase in just kind of the physical infrastructure that we will have access to.

That obviously means incredible things for what we can do from a reconditioning perspective, both in terms of volume where we expect to get 2 million units plus incrementally out of there versus our existing capacity once we fully ramp it up and then also from a location perspective because we can deliver cars to customers where we'll be much closer to them. So I think that's really exciting.

And then we also think the business is a great fit. We've often used this mental model of -- the entire used market is basically a big machine with lots of players that helps customers trade cars with one another. And there's all kinds of players inside that system. There's dealers. There's finance companies. Auctions are a really big player inside that system.

And when we talk about vertical integration, what we generally mean is we want to be as big a part of that system as we can be so that we can cut out as much expense and difficulty as possible for customers that are trading cars. And we think working together with ADESA is really interesting strategically for that reason as well.

When we think about build versus buy, it's actually a fairly straightforward analysis that you can kind of walk through in that same kind of framework. And I think the simplest way to think about it is just with respect to our inspection centers. I think in order for us to get similar production capacity out of the inspection centers that we would build, we would expect to need to build on the order of 30 inspection centers.

And we would expect that cost to be pretty high. We would especially expect it to be pretty high when you look at the locations that we're picking up through this ADESA acquisition. They've got many locations on the West Coast. They got many locations in the Northeast, in areas of the country that are very expensive in dollars but also very expensive in time in terms of how long it takes to find those locations and get them zoned and get them and get them operational.

So we think that from a build versus buy perspective, with just that simple analysis, you're probably pretty close -- you're probably very close actually. And then we get the instantaneous logistics benefits and access to this great business. So we're really excited about it and just think it lays a very clear path to the future.

I talked a bit in my opening about our oldest cohort of Atlanta now being up to 3.5% market share. And we think that's really important in this context because it does just very clearly light the path to 1.4 million units that you can just kind of see as long as we execute. And then beyond that, you know that you have the 50% growth that is continuing to happen inside of Atlanta. We don't know exactly where that growth is going to compound out over the next many years. But I think if you look at other businesses that are growing even at rates materially slower than 50% and look out 5 or 10 years, you'll generally see businesses that grow by a pretty significant multiple to where they sit at that point in time.

And so we think that lights the path to volume. We think if you look at the unit economics, we just had our best GPU year ever, which is extremely exciting. But I think every bit is exciting. We've been marched up year after year after year very consistently. We've been supporting all this growth and all the investments we've been making, and we've been marching up our EBITDA margin to the point we had our first positive earnings quarter and our first positive EBITDA year, excluding onetime adjustments.

So when you put all that together and then you lay that on top of this footprint that we're going to get access to that we think is acquired at a cost that is very reasonable relative to what it would cost to build out and then comes so much faster and with so many other great assets and opportunities, we basically just think that this is even more clearly an execution game from here. And so now the ball is in our hands, and we got to run with it. And we're extremely excited to be in that spot because this is a really big opportunity and the path has never been clearer.

The next question will come from Michael Montani with Evercore ISI.

I wanted to ask about -- if you can help to tease out a little bit the impact of the higher used car prices and rates vis-a-vis some of the impact you saw to being able to process the vehicles in the quarter and kind of how you see that playing out over the course of the year. And then I just had a follow-up on costs.

Sure. So I think if we are digging into the first quarter, I think there are a number of different impacts. And I think I would start with Omicron in particular, which is going to have a meaningful impact on our retail cost of goods sold in Q1. And Ernie touched on some of the reasons for that. It's really about the efficiency of the inspection and reconditioning centers. And then in this case, with Omicron, it's really been about the logistics network as well. And so we're expecting to see higher reconditioning costs and higher inbound transport costs that we expect to have a meaningful impact on retail GPU in Q1.

I think as it relates to the rapid rise in short-term interest rates, I think what we've seen over the last few months is a relatively historic rise in terms of the magnitude and the speed. I think you have to go back basically 10 to 15 years to the global financial crisis to see anything like this. And so because we typically will offer a customer rate and then there's some lag time between when we offer that rate and we monetize our loans, that rapid rise will impact other GPU in Q1. Now we have raised rates following the benchmark moves, and we're also -- we've scaled up our hedges to lessen the impact of future rate changes, but there will be a transitory impact there on Q1.

So I would call out those effects. I think -- clearly had severe weather, as Ernie mentioned, as well in Q1, which should be having an impact. So those are a number of things that are impacting Q1. I think when we take a broader view, obviously, we're very excited about the progress in the business, made tremendous progress in 2021. I expect to be over 4,000 GPU for the Q2 through Q4 in total. And so I think we're feeling really good about our overall progress.

Okay. And then if I could, just on the behavior of your consumer as it relates to the credit side, we're seeing some signs of normalization in some of the filings as well as from some competitors. So just wanted to get some added color from you all about the credit behaviors that you are seeing in terms of delinquencies and charge-offs and kind of how to think about that evolution as we get further beyond stimulus payments.

Sure. So I guess I would start with -- again, like the way that our business model operates is we generally sell that credit risk off at the time that we complete our securitizations through the sale of the residual asset kind of as well as the underlying bond. And so that doesn't flow directly through our financial statements. We kind of neither benefited from the extremely strong performance over the last couple of years nor would we necessarily get directly impacted if performance were to get worse in the future, outside of kind of the expectations for future losses that would be kind of priced into any bonds that we sell. So that doesn't kind of directly flow through our financial statements, just for clarity.

I think what we're generally seeing is the same thing that the industry is seeing, which is I think there is some approaching of normalization of consumer credit behavior. It still, on average, is better than historical, if we define historical as kind of normal times pre-2019. But it's not better by as much as it was better in kind of mid-2020 through maybe mid-2021, where performance was unbelievable. So I think there may be some normalization of those trends, but it's been, I would say, very orderly so far. And I don't think there's any reason to expect something different just yet.

The next question will come from Chris Bottiglieri with Exane BNP Paribas.

So first one -- so ADESA already got title processing and fund clearing for dealers. How does the title processing compare for dealer-to-dealer transactions versus dealer-to-consumer? I'm trying to get at -- are there synergies somehow that where this makes you more efficient on processing titles out of your core business? Like are you able to utilize their people? Or is the process completely different? And then I have a follow-up.

Sure. So as a start, I would say it's a simpler process, generally speaking, doing dealer-to-dealer transfers. And so I think you can think about those processes as existing pretty independently. Certainly, our teams will share learnings that we found over the years, and there may be some benefits there. But generally speaking, I think the simplest way to think about that is separate.

Now because you brought it up, I do think the registration team and title team inside Carvana definitely deserves a shout-out. They've done an unbelievable job over the last probably 6 months or so, making a lot of progress. We're now kind of approaching similar levels of success in title and registration to what we were experiencing pre-pandemic. So I think that that's really exciting, and that's hard to do in this environment. So there's a lot of great things going on there. They're working very well with our partners in the various DMVs to make sure that we make the customer experience as simple as possible. So I think we're excited about the progress we're making there.

And then I think on the ADESA front, we plan to continue operating that business as we have -- or as ADESA has over a very long period of time. So for all the people inside ADESA, it will be business as usual, and they'll continue to handle their title processing in the wholesale business the way that they have.

Got you. Okay. Follow-up question is I think you -- I'm not sure I heard this correctly, but I think you said the logistics benefit would be pretty instantaneous. I'm not sure if I heard that correctly. And then two, you're already using auction companies for reconditioning prior to the acquisition. So my question is, on day 1, can these facilities handle reconditioning for you and your partners? Or do you need to make that -- do you need to deploy some of that $1 billion to recondition any cars?

Sure. Well, I appreciate you clarifying because I want to make sure that we don't generate too high of expectations instantaneously. I think probably that language you're referring to, we instantaneously spin to an incredible network that will enable us to build a very meaningful reconditioning capacity and significantly improved logistic capabilities. That will take time. And I think we're not planning to provide precise time lines on any of that just yet, but we certainly have plans, and our teams will be hustling to get as much benefit from that as we possibly can as soon as possible.

Got you. And the Manheim -- or the...

Sure. So yes, so we do partner with both Manheim and ADESA on third-party reconditioning. I do think that offers some capabilities. But again, I think we'll stay away from giving too precise of guidance there.

The next question will come from Brian Nagel with Oppenheimer.

So my initial question is on the ADESA acquisition. First off, congratulations. And maybe a bit of a follow-up to some of the prior questions. But just want to understand the kind of mechanics here. So the ADESA facilities you're acquiring, when all is said and done, will they act just like an IRC that you built organically? Are they going to function the same? Or is there some type of relationship where the ADESA facilities will operate differently than the greenfield IRCs?

Sure. So I think a good starting point is they will operate largely as kind of 2 businesses on the same land is maybe a good and simple way to think about it. We've got our inspection center capabilities, and we've got our logistics capabilities that we've built out over time. And so I think we can go and we can kind of place those on these 4,000 acres of land around the country and operate them very similarly to the way that we historically have. And then I think the auction can continue to run the way that it historically has as well, so it will be business as usual sort of for both sides.

The place where we will be working together very quickly is in the retail, recon work that ADESA does. We'll be working to kind of merge those teams together. So that's kind of the single place of integration or at least the most important place of the integration between us.

And then part of why this is such a great fit is there's kind of a secular and a cyclical trend going on with auctions where, secularly, many more components of auction transactions are happening digitally even though there's a very large desire by many customers, both buyers and sellers, to go actually transact in person. And that often means that cars can spend less time on the ground at auctions, meaning that their space can be utilized in other ways. And then cyclically, we're at a time where auction volume is very low, historically low, in fact.

And so that means there's a lot of excess land at these locations. The average location is less than 50% utilized today, meaning there's a lot of opportunity for us to work together and for these functions to coexist. So again, we think the path that this lays for us is very, very clear. And now we need to just go execute. And I think that's our job for the next many years to take full advantage of this opportunity.

The next question will come from Rajat Gupta with JPMorgan.

Great. Congrats on the deal. Maybe a question on metering. You talked about that on the last earnings call. Where are we with that exercise? Are you still leaning to meter sales? Maybe if you could comment on how you are with respect to labor or logistics bottlenecks today. And I have a follow-up.

Sure. So the specific way that we "metered" demand last quarter was in suppressing the amount of inventory that was visible to customers. That tool is still being utilized today to a very similar degree, in fact, given the logistics network constraints that we see. So we are continuing to utilize that tool. That means that the inventory growth that we've seen that, I think, is evidence of the great work our team has done to expand our production capabilities is not really flowing through to customers yet in the form of as much growth in kind of available inventory. So we are very focused on alleviating the constraints in our logistics network. And once we do that, we have the ability to make more of that inventory visible to more customers, which is something we look forward to being able to do.

Understood. And then the comments in the shareholder letter around the second quarter to fourth quarter normalization on some of the metrics, is getting to the 550,000 units for the year assuming that some of these bottlenecks go away? And is it primarily capacity-driven as well? And maybe if you could comment on just the demand aspect. You mentioned that the low-end consumer is seeing a bit of an impact due to the pricing environment. Do you expect a normalization in terms of demand from that consumer segment as well in order to successfully hit the 550,000 number?

Sure. So again, I think we want to be careful about giving too much kind of guidance on guidance. But what I'll say is what basically underlies our expectations for the year is that we do resolve the constraints that we see in the logistics network. And then the demand environment, for the industry in total, remains somewhat similar to what it looks like today. And so I think that's basically what's underneath our expectations. As I said, I think today, we have -- we're part of the industry. So when there's affordability impacts to the industry, that's going to impact us as well. But given how constrained we are, it's probably not impacting us as much as it might be impacting the rest of the industry.

But we have seen sales slow down quite a bit industry-wide as car prices went up and interest rates went up and that impacted affordability. And so depending on what your comparison period is, sales look like they're down in the industry 15% to 20%, give or take, on the used side. And then we do see some evidence of that in our data as well. While our sales are largely driven today by the logistics network and not by the demand environment, the demand environment still flows through to the sales that you realize.

And so if you look through at our customers that are making less than $50,000, for example, they're growing about 30% slower than the company is year-to-date. If you look at customers making $100,000 or more, they're growing about 30% faster than the company is year-to-date. And normally, when you kind of do any reasonable kind of demographic cuts, you tend to see similar growth across all these different consumer segments. And so I think that pretty clearly shows the impact that affordability has had on consumers, on lower-FICO consumers and lower-income consumers and younger consumers. It's kind of showing up in the data.

So we're assuming that, that environment will remain somewhat similar. We would actually hope for a decreasing price environment. There's a lot of potential impacts that would result from that. But most importantly, it would improve affordability for customers and most likely drive more customers to a place where they're ready to buy again. And so we think that, that is probably what we're hoping for, but our outlook assumes kind of a similar macro environment to what we see today.

The next question will come from Colin Sebastian with Baird.

All right. So a couple of follow-ups. I guess first, curious if you could add some color in terms of unit sales and profit trends here later in February versus what you were seeing earlier in the quarter with Omicron, if that's part of what's giving you the visibility into the material improvement expected beginning in Q2. And then, Ernie, just on what you said right there around pricing and lower price environment you think would sort of be a favorable trend, in terms of embedding the current macro environment and the outlook, how would a faster declining pricing environment impact the guidance that you gave?

Sure. So I think a lot of these impacts we can start to see through, right? So the first quarter is going to be tough for a number of reasons. I think we'll start with Omicron and the inspection centers. That just flows through in COGS. In kind of smaller COVID waves earlier on, we saw impacts on the order of $150 to $200 that flowed through COGS. This was a bigger and more severe wave, right? So that then means that the inspection centers are less efficient. And so those costs get spread onto cars. And then when those cars sell, it shows up in lower margins. But as you kind of move through Omicron, the inspection centers are starting to recover, and so you can already get visibility into the direction that COGS is heading there. We still have some work to do, but you can get visibility.

I think interest rates, as Mark said earlier, moved really quickly. And when they move that quickly, we tend to do our securitizations toward the end of the quarter. And so the originations that we had earlier in the quarter, where they were originated in a different environment, are expected to have a different spread and then you see that impact as well. And so that can kind of flow through there.

And so I think when we look at margins on the cars themselves, we've also recently seen a little bit, for the first time in a while, of car price reduction. So we've seen kind of wholesale prices drop by maybe something like 5% for, say, 2019 model year cars year-to-date. We've seen retail prices drop by maybe 2%. And so we can start to see as well, while car prices started to reduce, that at least little cohort of cars, the margins there look potentially a little bit better despite the fact that car prices are dropping. So I think that that's helpful in giving us visibility beyond Q1 as well.

And then from a units perspective, it's really all about, or at least primarily about, alleviating constraints in our logistics network. And then from there, it's going to be about what does affordability look like. And I think it's hard to know exactly what's going to happen over the next 12 months with car prices or with interest rates. I think your guess is as good as ours there. But car prices going down is, all else constant, helpful for sales; interest rates going down, all else constant, helpful. The reverse is the opposite.

And then from a margin active, I think it's actually unclear how that cuts because, generally speaking, margins are driven by kind of the gap between wholesale and retail prices. And so it's all about how those 2 markets move in relation to one another. So I think we have pretty solid visibility into the next several quarters from where we are because a lot of these impacts, you can already see starting to abate a bit.

Next question will come from Adam Jonas with Morgan Stanley.

So look, in our discussions with clients, including right after the results, people are really concerned about 2 main things, right, growth and liquidity. Let me take this just one at a time, one question for each, Ernie. On growth, if I do this math, if I take your fourth quarter, or frankly, even your third quarter retail units, and I annualize them, I would get to around 450,000 units. So your guide on retail for this year is about 22% above kind of the annualization of either the third or fourth quarter, give or take. You with me, Ernie?

And I think we agree that the second half of last year was capacity constrained, and you're still experiencing those constraints now, but you anticipate those get better. So you're only guiding to 22% growth on what should be, over the balance of 2022, less capacity constraints than we've seen, let's say, in the fourth quarter. So adjusted for capacity constraint, your growth is well below 22% in your guide, maybe 10% or 15%. And that's a big change from what people were thinking of, maybe 30% growth not too long ago.

So is -- you're citing logistics and price shock. I just don't know -- it's just people are left with not much visibility on when the logistics get better. You see it in real time. What kind of metrics can you watch out for in terms of people or the size of that logistics pipe so we can determine whether this is an industry growth scare or a Carvana-specific thing?

Sure. What I would say is, I mean, from a demand perspective, we're very confident. I think you saw all the kind of market penetration growth, the number of markets that crossed over all these different lines. You saw our average cohort grew by 50% last year. Atlanta grew by 50%. It's our oldest cohort, and there's a lot of signals there that are very clear. If you look at website traffic, which website traffic is not a perfect top-funnel metric because website traffic quality can vary, but it's at least a reasonable top-funnel metric. We continue to see extremely fast growth there.

So I think what's driving our forecast is exactly what we said. The first thing we have to do that's Carvana specific is we have to alleviate the constraints in our logistics network. And then that will sort of get us to a baseline where we see where we are.

We are subject to the conditions of the industry. And so I think when we look at the industry today, and we see that many retailers are down between 15% and 20%, which we attribute to these affordability issues, we're taking that into account when we think about where things are likely to head for the remainder of the year. And so that's what's built in.

Got it. Ernie, just a follow-up on liquidity. Listen, and we've seen this with other companies like Tesla in the past, too, where kind of you get -- the market can get ahead of itself and your capacity for growth is kind of higher than what you're bringing in. I'm getting questions daily. Is this Peloton? Is Carvana going to run out of money?

Now you got $2.3 billion of liquidity. You got over $3 billion of used cars. You obviously got some great financing in place presumably with your financing partners for the acquisition. But just level set here, tell us why you're not running out of money or if you need money to kind of do what you're planning over this year, a, do you need it? And b, if you do, how much and how are you going to get it?

Sure. Yes, I can take that question. So at the end of the year, we had $2.3 billion in total liquidity resources. I think if you look at the way we used money in 2021, it was primarily investment in inventory and investment in finance platform assets. I think if you exclude inventory and finance platform assets, our cash flow from operating activities was only minus $82 million. That was a $200 million -- approximately $200 million, improvement year-over-year. And so I think the core business got to a point in 2021 where, again, like other than investments in inventory and investments in finance platform assets, it's really not using a lot of cash from operating activities. And so I think that's one important point to make.

I think we view the investment in inventory as a very positive thing for the business, obviously. Investment in inventory increased the selection for our customers. I think right now, we're actually metering visibility meaningfully. So we're not getting the full benefit of the investments that we've made in inventory so far. I think we're looking forward to the logistics network strengthening over time, coming off of these weather events and the impacts of Omicron so that we can un-meter some of the limited visibility of the inventory and get even more benefits. But I think we feel really good about that.

I think that one of the things that's frequently missed about our liquidity and cash needs is that we are an automotive retailer that invests a lot in hard assets. We invest in inventory. We invest in finance platform assets, and we invest in real estate. Those are highly financeable assets using traditional sources of financing that match those assets, which we used extensively in the past and expect to use in the future. And so I think we feel really good about all of those elements of our business, and those are some of the key things to be thinking about.

And then if I could jump in really quick, too. I do think we were lucky enough to have a very strong last couple of years. And I think that sometimes we can get roped in as kind of a pandemic growth story because those 2 years, we are a more visible company than we were prior. But that said, I think if you go back to 2017 when we went public, we were 1/10 the size that we are today. We grew on the order of 4x over the next 2-year period, pre-pandemic. And that growth was driven by the exact same forces that we expect our growth in the future to be driven by, which are these cohort curves and visibility into the future as long as we execute.

So I think all these things will work themselves out. It's all going to become visible over time. But we feel like the visibility into demand is very clear. And then hopefully, that was a helpful answer from Mark on the cash front.

The next question will come from Seth Basham with Wedbush Securities.

I guess my question is around gross profit for you on the retail side in the first quarter and how we should think about that. What's embedded in your expectations on used car prices? If we see an even sharper drop in wholesale prices or really retail prices, do you think that will have a more negative impact on your GPU outlook?

Sure. So I mean I can tell you some of the key factors that are going into it. I think we've called them out previously on this call. I think the key factors are really going to be the COGS impact of Omicron and, to a certain extent, weather on reconditioning costs and inbound logistics costs. I think that's going to be the biggest impact on retail GPU in Q1. And then we do think -- moving to wholesale, we are seeing higher depreciation rates in the wholesale market than we have seen in the past few quarters of 2021. And so we do expect that to have an impact on wholesale GPU in Q1. We expect it to be lower as a result of the higher depreciation rates than we've seen in previous quarters.

And then my follow-up is your decision on build versus buy with ADESA. And you talked about the equivalent capacity of 36 IRCs with this acquisition, which is about $88 million per location, including the additional improvement costs. What does that compare to versus building an IRC? Obviously, you take into consideration the additional profit stream for their auction business. But can you help us frame that a little bit more?

Sure. I think at a very high level, it's probably not very different. And the reason is that's more expensive than our historical average inspection center has been by a pretty decent amount. Our average inspections, in the past, have probably been more like $40 million to $65 million or $70 million, something like that. But these locations are oftentimes in more expensive cities and also in more expensive parts of town. This footprint has been built over a 30-year period.

And so a lot of it is infill location. It's a little bit closer to customers and a little bit closer to employment than many of the locations that we would build out today if we were going out and trying to find farmland, for example, and to build a greenfield. So again, I wouldn't want to be too precise there, but similar, when you're just evaluating it through the inspection center lens, it's probably not way off.

The next question will come from Mike Baker with D.A. Davidson.

Yes. I just wanted to follow up on Seth's question, maybe along the same lines. I mean I think based on the slide presentation here, the acquired company did $100 million in EBITDA. You're paying $2.2 billion if you exclude the additional CapEx. That's about 22x. There's got to be some synergies in there. So I suppose there's sort of conspicuous bias options in that. You talked about synergies, but you don't quantify it.

Any way you can quantify it or give us some way to get to, if this is accretive, how accretive this is to earnings or when it would be accretive? If you take the $3.3 billion in debt and put a reasonable interest rate on that, you're probably going to get incremental interest at or above the $100 million in EBITDA. So if you could give us some help with how to build this into our model in terms of accretion, that would be great.

Sure. So I think -- to be honest, I probably go about the analysis a pretty different way. I think you can -- the EBITDA is present, but I think that that's -- and it's important, and it represents the incredible business that's been built by ADESA over a long period of time. And I think that's an incredible asset that we're extremely excited to be getting. But I think we also can think about it in terms of the value that will be created through our other kind of pillars of rationale for this.

And so there is a large portion of the real estate that is owned by ADESA. That's a very valuable asset that can be added to this. And on that land, we can build these inspection centers. To the extent -- the reason that we thought it was important to put our cohort curves into this presentation was to give visibility into our ability to fill that production capacity.

And if you think about what 2 million units means to incremental EBITDA to the business, if you're able to produce an incremental 2 million units because you do have the demand at our long-term financial model, it's on the order of $5 billion annually. And then if you just look at our logistics benefits, which today, if we look at sales that happen within a 200-mile radius of our customers and compare that to the average sale that we have across the entire country, we're $750 better off in unit economics through lower inbound costs, lower outbound transport costs and lower customer acquisition costs because of higher conversion due to faster delivery. So that's a big number.

And so when you start to apply those sorts of numbers against our sales and when you start to imagine those incremental sales, you can get to really big numbers really fast. And I think the math just kind of isn't even important, right? The math isn't even -- well, what's important is our ability to execute to unlock those extremely gutty possibilities. So that's where I say our job now is to execute because the path is clear. We can see the demand. We know the unit economics. We now have the facilities to build on top of, and now we have to go do that. And so that's where our work lies.

Okay. Well, so then just -- I appreciate that. And to follow up on that, the cost to build, I think you just said to Seth, given the great locations, would be similar. So is it more that it just gets you there quicker? I know it's going to take some time to get these things ready for your IRCs. But if the cost to build would be similar and it's not really about the $100 million EBITDA, is it about the time or other sort of synergies? Or if you could just help us with that a little.

Yes. So I think all of the above, I think, time, I think, location, and then I think end point is also potentially different. So I think that you can think about all those things as being large benefits of this transaction. When I say locations, I mean quality of locations.

Next question will come from Nat Schindler with Bank of America.

Two quick questions. One, on the financing and other revenue and the GPU you've created from that has been obviously very good this year, how much of that is due to the fact that, at this point, default rates kind of don't matter anymore because you can sell the used car after someone has bought it for more than they paid for it? I mean so there -- how much was gain on sale from this unusual pricing environment, not just the fact that the car price was higher to begin with, but the fact that the risk in the loan had disappeared because of rising prices on cars?

Sure. So I mean I think that -- I think the simple answer to that question is very little benefit at all. And that benefit would have gone to the purchasers of our residuals who then observed better performance on the loans that they purchased than was expected. The economics that we get are a function of the expected losses over the life of the loans at the time that they're purchased. And as you can imagine, despite the fact that losses were great from kind of mid-2020 to -- well, I mean, even now, they still are great generally across the industry.

Investors aren't assuming that, that will be the case forever when they buy a multiyear asset class. And so the underlying assumptions that have driven the prices that we have received really haven't moved in a material way as a result of the last couple of years. But those who have bought the loans have significantly outperformed probably what they expected as a result of that great performance.

But then with a significant normalization in prices, significant, not 2% here and there, but a significant normalization in pricing on used cars, take it back so that there is a significant risk to those credit hedge funds that are buying your loans that they simply can't get back on the loans that they're used to because the car is going to drop in price?

So I'm going to try to answer this concisely because I think we could talk about it for a long time, but I do think that's just all about kind of the interest rates that exist on loans that are originated, which are set by the market. And the market is generally seeking interest rates that would allow a loan buyer in any form, whether whole loan or a partial loan or residual or whatever it is, to receive the yield that they would expect given the losses that they would expect. So I think that's just -- that's really a question of how market interest rates evolve, and we will evolve with the market.

I think as long as our evolution is happening with the market, we wouldn't expect the impact to conversion to be all that high. We wouldn't expect the impact to finance profits to be all that large. Where there is an exception to that rule is when you have a very rapid move like we recently saw, which we expect to flow through in Q1 and impact us negatively. But in general, we would expect the market rate to absorb those sorts of expected changes.

The next question will come from Naved Khan with Truist Securities.

I have a question on the ADESA acquisition. So if I disregard the real estate play that you're focusing on and just look at the core auction business that you're also buying, how should I think about the dealer participation once it's operated by Carvana? And is there a possibility that maybe the business could burn a little over time? And how should I think about that? And the $100 million in EBITDA that it generated last year, is there a risk to that?

So I would say, first of all, our plan there is business as usual, as discussed, for all the people inside of ADESA and also for all of the customers of ADESA that we're extremely excited to work with. So that's the plan. There's always risk in any business. So I mean I don't think that we want to say that there's no risk. We also think that there's a lot of opportunity. I think the transaction that we announced last quarter with Hertz maybe provides a bit of a window into the types of opportunities that could exist in the future. And we think there may be other opportunities like that as well.

So I think the plan is business as usual. We'll see how that plays out in the immediate term. We're excited to welcome the people of ADESA to the team and the customers of ADESA, and we're excited to find increasingly creative ways to continue to work together to take advantage of the incredible assets that we've both built independently that now get to come together to give our joint customers even better experiences and options. So we'll see how all of that plays out.

Okay. And then a follow-up question, maybe just on the consumer sourcing. Can you just sort of give us an update on where you are? Approximately what percentage of the vehicles are currently sold direct from consumers versus from dealers?

Sure. So we haven't provided precise numbers there in a while, but we have pointed to the fact that we've had extreme growth there across total transactions. This year, we had another year of over 100% growth. That's happened every year in our company's history, except for 2020 in the pandemic year. And we also are excited to -- we bought our millionth car from a customer just yesterday. So that was a great milestone that was recently achieved.

So we continue to have a lot of success there. We've got an offering that is very high quality that customers love, that is a very natural extension of our business because it just takes our logistics network in reverse. And it's simple and easy, straightforward and fair for our customers, which is the mark that we're always trying to hit. And that's all showing up in the results that we see and the growth that we've seen in that business, which continues to be very strong.

The next question will come from David Bellinger with Wolfe Research.

Ernie, back on the 550,000 unit guidance for 2022, does that include some type of benefit from ADESA? And maybe in terms of a lift in conversion rate, just given you could get first look at some of these more attractive, higher-demand vehicles at auction, is any of that embedded in your unit guidance?

I don't think we want to break out our guidance, but we're, as a general rule, not assuming super material benefits to the core business in this first year as a result of the acquisition. We think that more of that will come in the future than will come immediately. We'll obviously be working hard to find those benefits as soon as we possibly can, but that's our operating assumption going in.

Got it. Okay. And then just can we get a quick update, too, on the marketplace initiative? Any comment on the early progress there and anything in regard to economics behind those sales?

Yes. I think we don't have a ton more to share there. As we announced last quarter, we've got this really exciting partnership with Hertz, which we're excited about and where we're seeing great results early on. It remains early, I think, to do anything great with any partner. Both sides have to really lock arms and put the work in to build a great customer offering. And I think so far, both sides have done that, and we're excited by that. But there's certainly work left to do to unlock all the potential there. So potentially more updates there in the future.

Your last question will come from John Blackledge with Cowen.

Two questions. Will every -- of the 56 locations, will they all have IRC type of capabilities? Or is there, for some of them, overlap in near existing IRC facilities? And how do you envision the $1 billion to be spent across the 56 locations? And then second question, you referenced the $750 cost savings, does that change the kind of the long-term EBITDA margin range?

Sure. So on the first question, I think we're not going to get too specific on our plans there yet. There likely will be some locations that we do not add reconditioning capabilities to. That is taken into account in our approximately 2 million incremental unit estimate, and kind of the $1 billion estimate of CapEx is also taken into account in all that. And then as it relates to $750, I would say that that's more of an acceleration to our long-term model than it is an addition to our long-term model.

For those of you that remember, at our Analyst Day in 2018, we had simulations that assumed that we would have 40 reconditioning centers around the country, and getting those 40 reconditioning centers would drive down our transport distances and therefore drive down a lot of the expenses that we expect to be driven down by this acquisition. And that was built into our long-term financial model. So this is more locations that they are even better than we assumed the 40 locations would be back then. But that difference is very small relative to just the clear pathway to that benefit showing up in our numbers in a meaningful way faster than it otherwise would have.

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

Thanks. Well, first, I want to address everyone on team Carvana. Another incredible year. Thank you for everything you guys do. The last couple of years have been absolutely insane, and you've continually risen to the occasion. As I've told you before, I could not be prouder of what we achieved with that Forbes list. I think that, that is such a cool achievement. And it's very hard to do when we're all working as hard as we're working. So I cannot thank you enough for finding meaning in all the hard things that we do together and for having fun doing it and making it more fun for everyone around you. That's incredible, and it's the reason for our success. And as I said, I cannot thank you enough.

To the ADESA team, we are extremely excited to welcome you guys as well. We'll be talking more over the next several days. I think we have an incredible opportunity together. I can imagine that this is always a bit destabilizing. Change can always be scary. But we want you to know that we're extremely excited. We have a ton of respect for what you guys have built. We really look forward to working together. And we think it's going to be fun, and we think everyone's going to benefit.

And to ADESA's customers, I would say the same. We're extremely excited to work with you. We think there's a lot that we can do together. I can imagine the news is shaking you a bit, too. I can't wait to meet you in person and talk about the things we can do together. And again, I want to make sure you hear our commitment that our goal is business as usual and not to shake things up too much too quickly. And so we're excited to welcome you guys as customers. And as I said in my opening remarks, we will value the same way that we've always valued all of our customers, which is centrally.

So thank you to everyone, and we'll talk to you next quarter.

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