Carvana Co (CVNA) 2020 Q4 法說會逐字稿

Good afternoon, and welcome to the Carvana Fourth Quarter 2020 Earnings Call.

(Operator Instructions) Please note this event is being recorded.

I would now like to turn the conference over to Mike Levin, Vice President of Investor Relations.

Please go ahead.

Thank you, Andrea.

Good noon, ladies and gentlemen, and thank you for joining us on Carvana's Fourth Quarter and Full Year 2020 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 fourth quarter shareholder letter is also posted on the IR website.

Joining me on the call today are Ernie Garcia, Chief Executive Officer; and Mark Jenkins, Chief Financial Officer.

Before we start, I would like to remind you that the following discussion contains forward-looking statements within the meaning of the federal securities laws, including, but not limited to, Carvana's market opportunities 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.

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.

Thanks, everyone, for joining our call.

2020 was a defining year for Carvana and one that we will likely look back on as a significant milestone in our journey.

When we launched the company 8 years ago, we set out to be the first online seller of cars by building modern technology, a new optimized supply chain and integrating vertically to serve modern customer preferences and deliver the best customer experiences available.

When we went public 4 years ago, we spoke of our most mature market, Atlanta, as a proof point that we could build a sizable business, given its 1% market penetration, and we set a mid-term goal for the company to reach $3,000 GPU.

And 2 years ago, at our Analyst Day, we outlined the audacious goal of becoming the largest and most profitable automotive retailer of buying as many cars from our customers we were selling and of selling more than 2 million cars per year.

Today, we are the leading online seller of cars by a wide margin.

And for 8 years, we've been the leading -- the leader in providing the best customer experience available in buying a car.

Today, we have over 100 markets above 1% market penetration.

For 7 consecutive years, we've grown GPU by $400 or more.

And in 2020, we exceeded our $3,000 GPU goal for a full year and hit $4,000 in our best quarter.

Today, we are the second largest seller of used cars in the country, marking the final milestone on our path to becoming the largest.

And we have levered EBITDA margin for 7 consecutive years, demonstrating significant and consistent progress on our path to becoming the most profitable.

In the second half of 2020, we bought more cars from our customers than we sold to them, and we are celebrating our third consecutive year of being the fastest-growing automotive retailer in the country, lighting our path to selling more than 2 million cars per year.

Overall life is a company, we have set goals, and we have hit them.

And we have many more goals in front of us.

But before diving into more detail about where we are now and how we plan to continue hitting our goals in the years ahead, I want to talk for a moment about what got us here.

Our engine of growth has been incredible people with high ambition, boundless energy and discipline, who have chosen to care a little more than most, who learn and get a little better every day, who feel like they're part of something and who have fun along the way.

This has always been a close group that we are proud to be a part of.

But 2020 drew us even closer and made us even prouder.

We came into the year positioned for another year of hypergrowth before rapidly transitioning to a defensive posture in the pandemic hit and rapidly reverting back to growth posture as demand shifted in our direction.

Our team made those adjustments, managed through 3 waves of COVID and delivered another year of being the fastest-growing automotive retailer in the country while simultaneously growing GPU and levering EBITDA margin.

When we move to a defensive posture, our team stuck together.

We managed through the pandemic without a single layoff and without furloughs.

And when we reduced hours for our operations teams, our team came together.

We created the We're All In This Together Fund with contributions from hundreds and hundreds of people throughout the company that generated sufficient funds to ensure that the Carvana team members that were impacted by reduced hours, were still able to maintain over 80% of their take-home pay over the entire effective period.

To every member of our team, thank you.

You have an unbelievable mount to be proud of.

So where are we now?

And where do we go from here?

In the short-term as a result of the accelerating adoption of our model and the strength of our customer experiences in our brand, we have far more demand than we are able to satisfy with our current supply chain.

As evidenced of this demand, in January, we grew sales by 80% year-over-year, and did so with just half the immediately available inventory that we had a year ago.

This demand paired with the operational stresses of 3 successive waves of COVID, have led to constraints across our operational chain.

But given that they are most pronounced and hardest resolved in our IRCs, we will focus our comments on our progress there.

We opened 4 IRCs in 2020 and plan to open 2 more by the end of 2021, and the team is working at max speed to ramp hiring and training to catch up to demand as quickly as possible.

Since December, we have increased production by 40%.

In the medium term, we are working even faster than before to prepare the business to handle the demand.

In 2022, we plan to open 8 additional IRCs, bringing our total IRCs between now and the end of 2022 to 10 and bring our total facility capacity to 1.25 million per year at that time.

We're also scaling our logistics network capacity and capabilities further and making additional investments in technology that make our customer experiences even better to make us more efficient and enable us to scale more quickly.

In the long term, the plan remains the same.

We built a platform that delivers the best customer experiences, the best unit economics and is the most scalable.

This is a powerful combination.

We got here with execution, innovation and ambition.

Execution allows us to cover ground along our path.

Innovation allows us to uncover additional opportunities.

And ambition keeps us charging forward, and we are charging forward.

The opportunities in front of us are broadening and are even bigger than they were in the past, fully achieving our potential will demand that we continue to improve, that we move even faster.

Our ambition is only growing, but we must also maintain focus to move fast.

We'll continue to prioritize growth, and we must also stay disciplined to continue demonstrating operating leverage.

We will do all of that while always keeping our customers at the center of every decision we make and while delivering to them the best customer experiences around.

And we'll get a little bit better every day and have fun along the way.

It will be hard.

All important things are.

Our team is up for the challenge, 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 units and revenue.

Revenue totaled $5.6 billion, an increase of 42%, and retail units sold totaled 244,111, an increase of 37%, making us the second largest used automotive retailer in the U.S. For the fourth quarter, retail units sold totaled 72,172, an increase of 43%.

The total revenue was $1.8 billion, an increase of 65%.

Our exceptional growth in 2020 was driven by rapid growth within our market cohorts.

This strong growth overpowered COVID-19-related headwinds and our significant supply chain constraints.

In Q4, our 2013 to 2019 cohorts grew retail units sold by 40% and our oldest cohort of Atlanta grew by 12% despite an industry-wide decline in used vehicle sales and our election not to run our Cyber Monday promotion this year.

Our growth accelerated as we exited the year.

With our 2013 to 2019 cohorts growing by 75% in January and our oldest cohort of Atlanta growing by 45%.

In 2020, we completed our seventh consecutive year of $400 or more in GPU improvement and our seventh consecutive year of EBITDA margin leverage.

Total gross profit per unit for the year was $32.52.

Our growth in GPU was broad-based, including increases of $132 in retail, $43 in wholesale, $134 in finance and $90 in ancillary products.

Our GPU performance was bolstered by buying cars from customers.

In 2020, we acquired 204,000 vehicles from customers, an increase of 95%.

Total GPU in Q4 was 33 79, an increase of $549 year-over-year.

Sequential changes in GPU were primarily driven by retail GPU, which saw approximately $300 of normal seasonal depreciation, $100 of higher than normal seasonal depreciation and $200 of transitory costs, primarily driven by rapidly ramping our reentering capacity in the midst of the third wave of COVID-19.

We expect these transitory costs to flow into Q1, but to moderate beginning in Q2 as we ramp our new IRCs and move further away from the third wave.

In 2020, we took another meaningful step forward in our finance platform.

In Q4, we successfully completed our first SEC registered securitization, selling $405 million in principal balance and contributing to a $421 increase in finance GPU year-over-year.

EBITDA margin was negative 3.9% in Q4, an improvement of 4% versus the prior year, reflecting gains in both GPU and SG&A leverage.

We ended the year with $1.7 billion in total liquidity resources, giving us significant flexibility to execute our plan.

In 2020, we opened 120 new markets, bringing our year-end total to 266.

With these new market openings, we now serve 74% of the U.S. population, up from 67% at the end of 2019.

We will continue to expand in 2021 and expect to serve 78% to 80% of the U.S. population in more than 300 markets by year-end.

In 2020, we also made significant progress scaling our vehicle production capacity, and this continues to be an area of focus for the business.

We added 4 inspection and reconditioning centers in 2020, bringing our annual production capacity at full utilization to over 600,000 units at year-end.

We expect to open 2 IRCs in 2021 and 8 in 2022, ending 2022 with more than 1.25 million units of annual production capacity at full utilization.

Our eyes are squarely focused on achieving our goal of selling more than 2 million units per year.

And we will continue to maintain a healthy pipeline of future IRCs to support our growth.

As we look forward to 2021, we expect another strong year on retail units sold, revenue, GPU and EBITDA margin.

We expect an acceleration of growth in retail units sold in '21 and expect the level of that growth to be governed primarily by the speed at which we scale our production capacity.

We expect revenue growth in 2021 to be in line with retail units sold growth.

Finally, we expect total GPU in the mid 3,000 and a small EBITDA margin loss in 2021, improving both metrics year-over-year while continuing to invest for the long term.

Since becoming a public company nearly 4 years ago, we have made tremendous progress across all aspects of the business and achieved many major milestones, such as achieving our first EBITDA positive quarter this year.

Since 2017, we have consistently rapidly grown retail units sold, increased GPU and demonstrated operating leverage.

We've made significant progress and are excited about what comes next.

Thank you for your attention.

We'll now take questions.

(Operator Instructions) And the first question will come from Sharon Zackfia of William Blair.

Thanks for all of the commentary on the logistics and the reconditioning and kind of what you're facing there.

I guess I'm curious as well, there's been a lot of talk about stimulus checks and what that may or may not have meant to January this year.

I know you talked about January.

Have you seen trends continue strongly end of February?

And then as you think about delivery times, and I think this was commented on in the shareholder letter, what is the kind of optimal delivery timing?

I mean, where is that line in the sand where the customer starts to bulk and looks elsewhere?

And could you give us an idea of what delivery times are now versus maybe pre pandemic?

Sure.

Okay.

So I believe so January 1, with respect to stimulus.

I think it's hard to disentangle probably all the different things that have been going on.

Stimulus did not result in a large discontinuous super noticeable bump in demand.

So I don't think that we would have saw too much of what we saw in January to stimulus.

I think instead, what we would have signed the majority of the January strength too is just ramping up our production capacity.

As I said in my remarks, we ramped up our production capacity -- or excuse me, production level by approximately 40% from December to January, and that unlocks additional inventory, which then very rapidly sold.

So I think that would probably be the driver.

Heading into February, I think from a demand perspective, we've largely seen the same.

February can be a bit of a tricky month because there's really nothing about February that you would expect to be different from January with the exception of the arrival of tax season, which this year has reportedly been pushed back a couple of weeks, although I think we're starting to see some evidence that it's showing up at least moderately over the last couple of days.

So prior to kind of tax season hitting, I would say, February looked a lot like January and the same trends continued.

Those trends are just that we've got a lot more demand than we're able to satisfy given our limited inventory.

And so our focus is definitely on trying to ramp that up and produce inventories, by far and away, the biggest impact, but you also brought up delivery times.

The rest of the supply chain is also impacted.

That obviously includes logistics.

Roughly speaking, year-over-year, more recently over the last couple of weeks, logistics has probably been on the order of kind of -- delivery has probably been on the order of 40%, give or take longer than they were a year ago.

I don't think that there's a magical moment where delivery times get too long, and that starts to materially crimp sales.

I think it's much more of a continuous effect.

And I think that part of that increase in delivery times is just due to the underlying strain of ramping the business as quickly as we are.

Although I think a very meaningful portion of that strain has also just been the very large storms that we saw in the middle of the country over the last week or 2, which we would expect to be transitory, and we'd expect to work out over the next week or so.

I guess a follow-up.

I mean, knowing what you know now.

I mean, at what point do you think you've got reconditioning where it needs to be?

You've got logistics where they need to be, and we're not talking about the pinch points related to demand.

Sure.

So I mean, I think -- I do think it's a high quality problem.

I think we've dealt with supply constraints many times throughout the company's life.

Generally speaking, the form that takes is we'll see some supply side constraints, and then we will resolve those over maybe a 3-month period or so.

I think 2020 was pretty unique in the sense that we came into the year positioned for our largest growth year ever.

And then we were hit by the pandemic.

We rapidly pulled in.

And then kind of customer preference has shifted in our direction.

We saw more demand than we had ever seen before, and we tried to rapidly catch up.

And then that was kind of overlaid with 3 successive waves of COVID, which obviously have kind of impacts across our entire supply chain.

So I think this has been a difficult year, and I think we can almost think about it as being kind of 3 separate events that led to some supply side constraints.

I think the team has done an incredible job working out of all 3 of those different events.

And I think that's evidenced by the growth that we saw in the year despite the pandemic.

And then obviously, the growth that we saw in January is tremendous.

And I think at different points in the year, we kind of worked out of some of the supply side constraints started to build inventory again, and then we got hit with the next wave, and we had to kind of go back and start again.

So I think as long as we head into a more normalized world, I think we'll work out of this pretty quickly.

We put some materials together in the shareholder letter that are hopefully helpful about the IRC capacity that we are building.

That's facility capacity, which is different than active production capacity is producing cars, but we're clearly building for a very large future, and we're trying to make sure that we get in front of ourselves because when we look at our supply chain, the longest lead time components of that supply chain is building the physical facilities to do the reconditioning.

The second longest lead time is training and hiring all the people to turn that capacity into active capacity.

But we're working very hard to get in front of it.

And I think we feel incredible at the plan, and then I would just reiterate the statistic again, increasing total production by 40% from December to today is a pretty incredible move.

And I think, as I said, the team has done an unbelievable job in light of the circumstances to rapidly catch us up.

The next question comes from Zach Fadem of Wells Fargo.

So the quarter-to-date commentary for January would suggest that you're tracking at about 6,400 units per week.

And if we look back since you've been a public company, you've been able to step-up that weekly run rate every single quarter.

So when I think about the commentary around capacity and production and inventory constraints, is there anything out there besides those things that would preclude you from continuing to ramp up that weekly unit run rate?

That's the first question.

Could you talk me through that?

Sure.

So we put some commentary in our outlook about the expected differences in shape this year.

And I think that kind of cuts directly to the point that you're making.

The demand that we're seeing, I want to reiterate is very, very strong.

We're clearly seeing materially more demand than we're going to be able to satisfy this quarter and probably for the remainder of the entire year, and we're going to work very hard to try to catch back up.

And I don't think we want to dive into weekly sales rates, but kind of the numbers that you're pointing to aren't unreasonable.

I think it's important to look at those numbers in light of the graph that we provided around immediately available inventory on the website in our shareholder letter, where as of the end of January, we had 9,500 cars that were immediately available for customers.

That means cars that a customer can purchase and are ready to be loaded on a truck right when the customer purchase them.

So selling that number of cars against an inventory of 9,500 cars is an extremely rapid turn.

And there is just something to kind of a natural limit of how quickly customers can get on the site, find the car they want and start to go through the purchase process.

So I think our expectation is that probably relative to previous years, sales volumes will be pushed a little bit into the future.

Partially, that's going to be because of tax season being pushed back a couple of weeks.

But I think the bigger impact is probably just us catching up to demand and catching up our production capacity.

And so I think you've seen these big discontinuous moves in the past.

Those really big discontinuous moves have been supported by a much larger ready inventory than we have today.

And so in order to take as big of a discontinuous step, we would have to be turning inventory extremely, extremely quickly.

So it's likely that in the immediate term, the step-up will not be as severe.

But then because we have excess demand as we move through the year and catch up production, we may see kind of a more beneficial shape later in the year.

Got it.

Okay.

That one makes sense.

And then, Mark, as I look at your business today, you sold about 244,000 cars last year.

And it took you an operating expense base of about $1.1 billion to support that.

So as I look to 2021, can you talk about incremental SG&A per unit, and walk me through how to think about the SG&A step up?

And what is expected to be a fixed cost, which I guess we can call it investment?

And then how much would be a variable cost?

Sure.

So as you sort of alluded to, as we've grown the business, we've been able to show leverage in SG&A.

I think that will continue to be the case going forward as we march through our long-term model.

As we take a narrower view into 2021, I think one way that we think about this question of leverage is through our cohorts, where we've definitely seen leverage over time as cohorts age.

I think that, that happens in several of the line items.

One, I think that we've historically seen advertising leverage in our cohorts as they age, as word-of-mouth grows, as awareness grows and the cohorts, we provided some data on that in the shareholder letter.

As it relates to some of the other expense line items, we've also seen leverage in the cohorts there.

I think that takes the form of benefits from scale.

I think it takes the form of benefits of utilization in the logistics network.

And then there's also benefits as we scale of spreading some of our fixed cost base out over more and more units and more and more markets.

And so I think we've seen lots of great trends and leverage in the business.

We provided some guidance on EBITDA margin, where we expect another year of improvement.

We think that will be coupled with improvement in SG&A as we march toward our long-term model.

And I think we feel pretty good about the path that we're on.

Your next question comes from Michael Montani of Evercore ISI.

The first one I had was just to clarify a little bit in the guidance, you had mentioned a small EBITDA margin loss.

Does that basically mean like a down low single-digit EBITDA margin?

Or were you talking about the actual dollar amount itself?

So I think what we were guiding to is mid-3,000 GPU and a small EBITDA margin loss.

I think that again will be driven by some of the sources that I was just describing in response to Zach's question.

I mean I think that will be our eighth consecutive year of EBITDA margin improvement.

Okay.

And then the other question I had was on total revenue growing roughly in line with retail units sold.

I was thinking it might be up faster given the success that you've seen in wholesale, ancillary and then just the ASPs being up year-over-year.

So I'm just wondering if there's something there that I'm missing that we need to consider as well on some of these other lines of business?

Sure.

So I think you're hitting on a couple of good points there.

I think the major drivers of revenue are vehicle ASP and then wholesale and other revenues.

And we've obviously made progress in all of those, really the latter 2. I think one point that I would make is we've seen ASPs drift up here a little bit in the back half of the year.

Part of that is just supply and demand and our purchasing algorithms responding to what we're seeing in supply and demand.

Part of that is also an intentional to a degree.

As we're constrained in the reconditioning centers, we've tended to shade a little bit more toward some cars that maybe require less reconditioning on occasion.

So that's been a driver.

I think the gains that we saw in ASP in the back half of the year, we expect to moderate a little bit in 2021, so the balance of all of those things is what led to our guidance.

The next question comes from Naved Khan of Truist Securities.

This is Robert Zeller on for Naved.

Congrats on the good -- on the great performance of a quarter.

A couple of questions.

So can you just reiterate -- just discuss about what made Atlanta, which is one of your first markets, what made that market very attractive to enter?

And then stepping back, as we speak broadly about the rule of 40 with -- in the Internet space.

We've seen that become the rule of 50, the rule of 60, the rule of 70.

And Carvana has shown really great top line growth and improving EBITDA margin.

So curious how you're thinking about that general phenomenon moving forward?

Sure.

Well, first of all, thank you.

I can tell you were internally debate and good quarter and great quarter.

We appreciate you going with a great quarter there.

So thank you for that.

What makes Atlanta an attractive market to enter?

That's going way back in our history.

The short story is basically, we had some kind of operational benefits to scaling out of a pre-existing reconditioning center in that area.

And so that's what kind of led to us choosing that market.

There was nothing unique about the market that we thought would uniquely fit our model.

It was more about just operational ease that caused us to choose to go there.

And then obviously, that market has responded really well ever since.

And I think we put some commentary in the letter that 90% of the markets that we've launched are continuing to ramp even faster than Atlanta did at the same point in its life.

So I think there's nothing special about the choice of that market.

It's been a great market for us.

It's continued to grow.

It had a great January, growing at 45%.

And I think the other markets are performing very similarly.

Mark, do you want to hit the rule of 40, 50, 60?

I'm not sure I'm familiar enough for that rule.

Sure.

Yes, yes.

I mean I think we're very happy with the comp to fast-growing Internet companies.

Thank you for that.

I think historically, we fared extremely well on any rule that you could put together.

I think our growth has been historically strong compared to other e-commerce companies.

I think we've continued that strong growth while also showing significantly narrowing EBITDA loss margins.

So I think we continue to fair well on those metrics.

We're going to look to obviously continue to make progress on both top line growth and simultaneously on EBITDA margin leverage, which is something we've been able to do for 7 consecutive years now.

So we're going to keep trying to keep that number as high as we possibly can.

The other thing that I would just say is, as I try to infer what the underlying drivers of that rule are.

I also think our business has the unique benefit of having visibility into the underlying cohorts as well.

We shared data in the third quarter that our oldest cohorts were already approaching our long-term EBITDA target range.

And that several of our cohorts were positive from EBITDA perspective.

So I think that really does help to clarify that kind of the -- while we're making all this progress and constantly narrowing our EBITDA margin losses, we have shown negative EBITDA over time, but that kind of investment is clearly going into the growth of the business.

Your next question comes from Rajat Gupta of JPMorgan.

I just had a couple on like the GPU.

The $200 transitionary impact on retail GPU in the fourth quarter, what is assumed for the 2021 mid 3,000 guide with respect to that?

Does that all come back?

Do you still have a of that flowing through into the guidance?

And just relatedly, could you give us a sense of like how that mid-3,000 breaks up narrow across the 3 different businesses?

That would be helpful.

And then I have a quick follow-up.

Sure.

Yes.

So as it relates to the transitory costs, we do expect them to flow into Q1, but then to moderate beginning in Q2.

And the moderation comes from -- as we ramp our new IRCs and as we move further away from COVID, the third wave of COVID, we expect those impacts to moderate.

So that was -- on your first question.

On your second question, consistent with our sort of past outlooks that we've given for the year, we don't plan to break down the mid-3,000 GPU into various components.

But I think we've given a pretty good sense there of the progress that we expect to make next year.

And I think we're feeling really good about that total GPU progress.

Not unlike EBITDA margin.

Next year, we'll make our eighth consecutive year of improvements in total GPU, and we're feeling really good about that.

Got it.

That's helpful.

And just on the February comment, you talked about trends continuing like January.

I'm not sure if I missed it, but was there any impact from the adverse weather that hurt February?

Or despite that, you continue to see the same level of growth that you saw in January?

Just wanted to clarify that point.

Sure.

Yes.

I think there are several concepts in there.

So in terms of underlying demand, we've seen underlying demand be very consistent throughout the early part of this year so far.

In terms of sales, we've seen kind of the demand for sale to be very consistent, but there have been some delays related to the storms in the middle of the country.

We expect to resolve those over the next week or so.

As it relates to year-over-year growth, that will be impacted by the storms that we saw, which impacts the timing of sales, but then also by the timing of tax season, which is the major catalyst for growth that happens sometime between kind of mid-February and mid-March.

And we expect to be later this year than last.

But as it relates to kind of modeling underlying demand and getting that signal for what that implies for the strength of the business and what we expect the year going forward, it's been very consistent so far this year.

The next question comes from Edward Yruma of KeyBanc Capital Markets.

I guess 2 for me.

First, on the consumer sourced vehicles.

It looks like you guys continue to push penetration higher.

Just wondering if you're starting to get to the point where the consumer is potentially pricing against some of your competitors, are you having to pay more for vehicles at your consumer sourcing?

And then maybe just as a second question, now that we're in a rising rate environment, how should we expect that to flow through the P&L, particularly out of the fact that you're doing more of your securitization yourself and trying to net those 2 out?

Sure.

So first of all, on buying cars from customers.

That business obviously continues to do incredibly well and grew at 96% for 2020.

I do think the most impressive statistic there is that in the fourth quarter of the cars that we sold to customers, 65% had been bought from customers.

To put that in context, when we had our Analyst Day 2 years ago.

We set out a long-term goal of 38% to 52%.

So for roughly 2/3 of our cars that we sold to customers we bought from other customers, that's a pretty remarkable number and something that we're extremely excited about.

As it relates to the implications for the future and the remaining opportunity, please keep in mind that on the retail side of the business, we're still about 0.7% of the entire market.

We remain very, very small compared to the business.

And on the purchasing side of the business, we're probably a little bit bigger than that, but not materially bigger than that.

And those 2 businesses are largely independent.

They kind of intersect and move together as it relates to trade-ins when a customer simultaneously buys and sells a car.

But the majority of each business runs independently.

So I think looking forward, we expect those businesses to grow independently to some degree.

We expect them to basically race against each other in moments of strain when we are constrained in our logistics network, we will tend to favor retail transactions to vehicle purchasing transactions in our delivery schedule.

So that can have some impacts in the near-term and kind of makes the businesses interact together.

But over the medium and long term, we don't expect that to be a big driver, and we're just going to work to grow both those businesses as fast as we possibly can as we think that they are mutually reinforcing and enormous opportunities independently.

As far as selling loans goes, there's a lot that goes into that.

I would say, generally speaking, over any kind of medium-range period of time kind of consumer spreads or kind of interest rates that we would fund at and consumer-facing rates have spreads that are fairly consistent.

And so I think, generally speaking, the expected finance margin is very similar across time.

I do think in a lowering rate environment, sometimes those margins get a little bit bigger for a quarter or 2, and in a raising rate environment, they can get compressed a little bit for a quarter or 2. But those impacts are generally relatively small compared to the entire finance margin.

So I think kind of the directional impact you're pointing to is a reasonable factor, although there are many factors, including just the underlying improvement we've seen in that business over the last year or so.

So I think you could model that in, but I don't think it's a big contributor.

The next question comes from Chris Bottiglieri of Exane Paribas.

So I want to ask you about inventory selection on our website.

Clearly, there's very prolific levels of demand that's outpacing your ability to scale production.

From my observations, you've been testing third-party part inventory on your website for at least a year or 2 now.

Can you talk about any initial results that you've seen?

What are the benefits and trade-offs with Carvana and the third-party dealer from this arrangement?

And then from an investor standpoint, how do we evaluate the economic implications of allowing more third parties on your website?

Sure.

So there's a lot in there.

Let me start with this.

In any given quarter, we probably have 100 to 200 tests that we're running across the business.

There's a much smaller set of potentially very large tests.

But we have many, many tests that we're running.

We have -- depending on how you break it down, probably 60 to 75 different independent product groups that are made up of engineers and product designers and visual designers and analysts and infrastructure developers and database developers that are all running kind of in their own area and building their own technology.

So there's always a lot going on.

And in general, our policy is going to be to just discuss things that we believe will have early to have the potential to have a material impact on our P&L in the near term.

And prior to that, we will probably just choose to not discuss these underlying tests too much.

That said, the particular test that you are talking about, we are going to be moving into a second phase of that test to test it at slightly higher scales.

We still expect it to be small.

It does not have a material impact on the business to the extent that changes, and we will start to discuss it with you more.

But we will be doing a little bit more of that.

So you may see some more partners and more inventory showing up there.

You then ask a question about how we collectively can evaluate all the opportunities that Carvana faces.

And generally speaking, there are many lens, in many ways that we think about these different opportunities, but the lens that we use the most often is we're always trying to pick the opportunities that we think offer the best customer experience, the most compelling unit economics and the most scalability, and we're trying to make good choices in the trade-offs of those 3 dimensions.

And that's how we will determine where we put our energies.

And ofttimes, there will be many tests going on in the background that are helping us to understand what some of those trade-offs may be with some of the things that we could invest more energy into.

Yes.

That's really helpful.

And then just a follow-up is I want to follow-up on the Atlanta performance is really eye-popping the 24% to 45% growth in January.

What would you attribute the acceleration?

When I look at the 2 IRCs you opened in West Memphis and Orlando, they seem to flank Atlanta from a geographical perspective.

Is it reasonable to think that Atlanta, some of your earlier markets disproportionately benefited from these IRCs?

And are there any other markets where you feel like when you open an IRC this year it coul sustain a similar benefit?

Sure.

So I think there's a lot going on there.

Let me start with what I think is most important, and then we can talk about some of the other impacts as well.

The most important impact, I think, by a long way, we talked a lot about the constraints that we face as we kind of dealt with COVID and then try to catch up with these 3 successive waves.

I think the right way to look at kind of the back 3 quarters of 2020 is that we really aren't seeing demand signals when we look at the markets.

What we're really looking at is we're looking at the way that we're allocating the very constrained supply that we had.

So if you see one market growing faster than other market, all that's really telling us is just that we were constrained in total sales and the market that grew faster had even more aggressive underlying demand, then the markets that grew a little bit slower.

And so therefore, more of the cars were shifted in that direction.

And so I think when you're trying to assess demand growth over the last 9 months, I just it's difficult looking at sales data because we have been so severely constrained.

So that, I think, is the biggest thing.

And then I think that, that -- while we still are constrained, I think what we saw in January is we saw the benefit of this 40% increase in production that happened from December to now that has been very helpful.

So I think that, that's probably the biggest underlying driver.

I do think that in general, as we've opened inspection centers near markets, those markets have benefited from that because those cars are closer by.

And therefore, oftentimes, those customers will buy them because the delivery times are faster.

So they may have more demand for those cars than customers in other markets.

So I think that's an impact.

And then we also launched a vending machine in Atlanta recently.

So I think that, that is probably an impact as well.

So I think there's a number of things going on, but I do think the most important is just kind of the -- while we remain very, very supply constrained, I think slightly loosening those constraints was helpful in January.

The next question comes from Nick Jones of Citi.

As the demand for consumers to sell cars directly to you exceeds your original goals.

I guess, what have you learned?

And can you revise those goals potentially?

Or where do you think, I guess both ratios can go over time, customer source ratio and then also the ratio as a percent of retail sold?

So I think -- let me start with how we arrived at our original goals.

Our original goals were arrived at by looking at market leaders over time and what they had been able to achieve.

And kind of using the heuristic that, that was a good target for us to aim for.

And I think the results that we've seen over the last year are really, really exciting, and we think just speak to the extreme simplicity of our offering from a customer experience perspective, but also to the underlying strength of the model.

When we've got a model where we can sell cars across geographies and across makes and models and across the price spectrum.

And we've got very efficient ways to sell even wholesale cars.

It puts us in a position where we can be a very high-quality buyer.

And so we've seen a lot of success in that business.

So I think that we're excited by where we are.

I think going forward, it's -- there's no -- until we get extremely large, there isn't necessarily a fundamental ratio that we would shoot for with respect to cars bought relative to cars sold.

I think we will just seek to grow both businesses as quickly as we possibly can because we've set up the business to do so.

If we're buying significantly more cars than we're selling, then we have the capacity to sell the excess wholesale.

If we're not, then we have the capacity to buy cars from other channels.

And so we're just going to look to grow both businesses as quickly as we can independently.

Great.

Great.

And 1 follow-up on, I guess, on -- as you expand, they cover more population.

How should we think about that?

Is that -- are there going to be some maybe large new markets?

Are these mostly adjacent, or maybe there's already kind of high Carvana brand awareness?

Sure.

So I mean, I think like always, it will likely be a mix of both.

I think the way we've historically expanded our network after launching in Atlanta in 2013 is sort of a continuous expansion out from Atlanta and now crossing most of the U.S., there's still some room for us to expand.

There's still some fill in markets available to us within the existing footprint of our network.

And so we're just going to keep marching forward here toward our goal of near nationwide coverage.

The next question comes from Ron Josey of JMP Securities.

Maybe 1 bigger picture one and then more specific -- some of that...

Ron, I apologize, can you hear me?

We lost you at more specific, if you could start there.

All right.

Well, high-level question and 1 more specific question.

So -- and it gets to do -- it has to do with just more demand than supply, but then alleviating that in January.

So just bigger picture, why is 10 the right number in the next 2 years for IRCs?

If the goal is 2 million, why not just go after that, build more in '21, more in '22, et cetera.

Because you're clearly seeing the benefits here in January when more supply came on?

And then the second question is, it looks like advertising rose 40% in the quarter.

And so Mark, as basically demand outstrip supply, can you just talk a little bit more about your advertising plans and -- just making sure that as awareness grows that you've got the supply to sort of back it up?

Sure.

So at a high level, I think we -- I think the right way to think about how we're growing IRC capacity is basically as fast as we can.

And so that is a relatively long lead time undertaking.

You have to find different sites around the country in the areas where you want to be, you have to negotiate transactions, you have to work with the city to get your zoning approvals, you have to then build the physical sites and then you have to hire and train all the people in those sites.

So I think the simple way to think about that is we're going as fast as we possibly can.

And I think that we recently have added more capacity than ever in our history, and our plan over the next 24 months is to add significantly more capacity than we've ever had in our history, and we're working hard behind the scenes to fill in a robust pipeline behind that.

So in general, I think we don't intend to miss on the not ambitious side.

We're going to be working really hard to try to build as much capacity as we possibly can.

Yes.

And then on the advertising point, it's an interesting question that we always face whenever we get constrained in operations.

That question is, hey, do you start to pull back in certain areas of the business while you are constrained.

And I think when we're balancing those questions, we always, first and foremost, take a long-term view.

I think we want to continue to build our brand.

We want to continue to provide a great experience to the customer, great value to the customer.

And we're going to take -- you're going to look at those things through a long-term lens.

In addition, as Ernie has alluded to, we're also -- we have many teams that are working as hard as they can, doing an amazing job, working to scale our production capacity, scale us up in the logistics network and the other parts of our supply chain to help us meet all the demand that we're seeing.

And so that's where we're really focused.

We know we have a tremendous amount of demand, and we're working as hard as we can to make sure that we can fulfill it all.

And maybe just a little additional color there, Ron, in case is helpful.

I do think that we try to think a little bit differently about brand advertising and direct advertising.

And I think when we are severely constrained, we will pull back a bit on some of the kind of direct response advertising, but we still got a really long way to go in brand advertising.

Based on our internal surveys that we've done, we still would only have approximately 10% unaided awareness nationally.

So there's still a lot of room to go there.

And just as part of the long-term view, we do try to balance those things, even though we recognize it can look like a bit of a conflict in the near-term when we're clearly constrained on the supply side and already have an excess of demand.

(Operator Instructions) And our next question comes from John Colantuoni of Jefferies.

On capacity constraints, should we think of the 80% growth in January as a decent proxy for what unit sales could have grown in Q4 had capacity not been a constraining factor?

And I have a follow-up.

I want to be a little careful about diving into too much hypothetical, but let's just talk about January because I think we've at least provided the underlying data for that.

We grew at 80% in January, and we had half of the available inventory that we had a year before.

We also have longer delivery times than we had the year before.

So I think the underlying demand in January, we believe, was sufficient for sales to have grown materially more than they did in January.

And I think kind of on either side of that, that is likely true.

So I think that's the right way to think about it.

As we look forward, we feel like this is really a question of how quickly can we ramp up production capacity, and that's going to be what drives our sales growth.

There seems to be significantly more demand than we're able to handle today.

Great.

And can you help disaggregate the mid 3,000 GPU expected in 2021 between retail, wholesale and finance and other?

Yes.

So consistent with our past outlooks that we've given over the last several years, we don't break the total GPU components out in our guidance, but we hope that the mid 3,000s total GPU guidance for the year will be helpful.

The next question comes from Brian Nagel of Oppenheimer.

Nice quarter, congratulations.

So I apologize if it's a bit repetitive.

But just on -- with regard to supply constraints, you talked about in your letter, in your comments here that those were alleviated somewhat in January, and you saw sales improve significantly.

When you refer to supply constraints, is there a delineation between what's in your control and what you're in the process of building out versus external factors that may be persistent for a while longer?

Sure.

I would say, in our control would be the characterization of our supply side constraints.

And so just to be very clear, what we mean by that, our supply chain means buying cars.

It means transferring cars to reconditioning centers, putting roughly $1,000 of parts and labor into those cars, photographing them, preparing them for sale, listing them on the website.

A customer calling into an advocate and asks some questions about the car and then going through the purchase process.

And then that car being delivered to the customer.

That is the sum of the chains.

And so I think the places where we historically have seen constraints are in customer care, the advocates in logistics and in the reconditioning center.

Today, they are clearly most pronounced in the reconditioning centers.

And so that's where the constraints are.

And we collectively refer to that supply chain as supply side constraints.

The constraints are not in just acquiring vehicles.

In general, we have ready access to vehicles.

Our final question will come from Adam Jonas of Morgan Stanley.

So Ernie, once upon a time, people used to describe Amazon as an online book retailer.

Today, Carvana is described as an online used car dealer.

Now, look, I think one of the themes from this call is you guys are really constrained to meet demand.

And so this question, obviously, is quite long term.

And I'm not asking you to break major new information on this call, right.

But could the Carvana vertically integrated software and logistics platform that you've built and are growing, could that have a use in addressing auto or fleet management-related markets beyond just used cars, specifically new or mega fleet management, stroke, rideshare, et cetera?

So that's a big question.

So let me start with this.

I think we are clearly constrained today, and we clearly have an unbelievable opportunity in front of us, just kind of blocking and tackling and doing exactly what we know how to do.

We have 0.7% market share.

It's a $40 million year per year unit.

There's a ton of room to buy more cars from customers.

I think there's theoretical room to accelerate the rate at which customers turn over cars between each other, making the market larger.

But then I do think that there's also no doubt that we have a lot of really compelling assets.

And I think there's probably lots of ways to think about our assets.

But I think one way that's maybe useful is to kind of think about kind of any transaction requires a customer.

It requires the transaction itself, and then it requires a vehicle.

And I think when you think about the Carvana business through that lens, kind of the customer is our brand, which I think we've built a pretty exceptional customer brand that's known for delivering great customer experiences.

And while there's still a ton of room to grow that, we're investing in it.

It's one of the better known brands now in automotive.

And so I think that's exciting.

I think when you look at the transaction, I think we've built a lot of technology that takes a very complicated vehicle purchase and makes it simple for a customer by kind of taking all the things that go into purchasing a car and the surrounding transactions and just making that easy.

And so that's a tool set that we've built.

And then when you think about the supply side, we've got a supply chain that enables us to deliver a car to a customer's door with a branded hauler and a uniformed Carvana team member to what we hope to be 80% of the U.S. population by the end of this year.

So I do think that those are very, very interesting assets.

And I think that as we think about the way that automotive -- the automotive industry at large evolves over the next decade or so.

I think as you think about many of those opportunities, there are many places where those assets or some combination of those assets may be an extremely important and highly differentiated part of various customer value chains.

And so I do think that it doesn't take a ton of creativity to see that.

Again, I want to reiterate, the opportunity in front of us is really, really large.

And I think that one of the great problems that we face.

And I think this is a very high-quality problem as well is trying to find the balance between being extremely focused on the things that we know really well, but making sure that we're ambitious enough to take full advantage of the position in which we find ourselves.

And I think that our ability to manage that equation and to continue to move fast, we continue to put space between us and others that are looking at the same problem set, I think is going to end being the story of Carvana over the next decade.

So we're aware of all that, and we're trying to make smart decisions as we move forward.

Okay.

Ernie, if I could just have a follow-up and then I'll bust out here.

As you're no doubt aware, there are a lot of new EV manufacturers entering the U.S. market.

And they all have one thing in common, they don't want to use franchise dealers, right?

Not at all.

They all want to go direct-to-consumer.

I mean, I count 8 to 10 significant new OEMs in the next 3 to 5 years.

And I'm counting 15 to 20 by the end of the decade, and that's a very open-ended number.

So I'm just kind of -- and these guys as they're coming to market, they're teaming up with partners.

I think sister is using Cox or Manheim.

Lucid is using Glovis.

Lordstown, I think, Camping World.

I mean, give me a break.

So again, could you confirm -- you're super capacity constrained, but is your team even at a high level entertaining discussions with this flood of new players that will need logistics and support and customer care support when they have products, but no way to kind of serve the customer?

Are there even advanced discussions -- not advanced, preliminary, highly preliminary discussions at this time?

Or would you care to avoid that question?

I think we're probably smarter to stick with the answer that we gave you to the previous question, and then I would just reiterate that our supply constraints are generally in the reconditioning center.

That's generally where they are.

So anyway, I appreciate the question.

Recognize that there are many opportunities out there.

And as I said, I think something we got to do a really good job of over the next 10 years is finding the balance between focus and ambition.

This concludes our question-and-answer session.

I would like to turn the conference back over to management for any closing remarks.

Great.

Well, thank you, everyone, for joining the call.

We really appreciate it.

And to the entire Carvana team, unbelievable year.

Thank you so much for everything you did.

This year threw more at us than we could have possibly imagined.

And I don't think there will ever be a year that I'll walk out of prouder, and I hope you feel the same way because you did an incredible job.

We stuck together as a team.

We're in a great spot as a result.

So thank you so much.

Talk to you next quarter.

The conference has now concluded.

Thank you for attending today's presentation, and you may now disconnect.