Okay, let's get started.
Let's do it.
Ernie Garcia, thanks very much. President, CEO, and Chairman of Carvana. Ernie and I have had a number of these fireside chats. Sometimes you wear socks, sometimes you don't wear socks.
Good memory. I wore socks because you called me out last time.
Oh, come on!
I did.
You have an idea.
Your ankles are fire, though.
Oh, gosh.
We do have a full room as well. A lot to talk about. I mean, I just start off by saying a year ago, a year and a half ago, it wasn't looking good. I was talking to you. We would have our quarterly calls and sometimes more frequently, and I'd ask, like, "How you doing?" Either you were a very good liar, but you always said, like, you sound. That's the answer.
Well, you sound you were you had your head down, and a lot of people were counting you out. One of the classic buy signals on Wall Street is when a sell-side analyst pulls a rating and moves to non-rated. I remember on the call with you, and I said, "Ernie, look, I believe in you. I believe in the story. I think you're going to make it. But this is really, really hard. There's so many binary outcomes here. I have to pull the rating, and it's probably a buy signal." And you were like, "Dude, I get it." You kept it cool. You kept it calm. You didn't go bankrupt. You didn't wildly dilute.
The low bar.
You did wildly well, I mean, this is the auto. The auto's tough. You got to know your industry.
Yeah.
Know your audience. You did not wildly dilute shareholders. You put your head down and kept your team together. You kept a lot of the core team together, which is critical. How do you keep talent from, you know, evacuating? You, I would say you stopped the bleeding. You got bankruptcy. I think you did. You took steps to kind of move that off the table on the liquidity side, right-size the business, focused on unit economics. Now here we are talking about growth in the first quarter, maybe double-digit growth kind of thing, and being, you know, record levels of EBITDA profitability. How the hell did you do that? And, you know, kind of what are your messages just to begin the discussion, and then we'll go through it.
Oh, man. Okay. Well, that's a wide opening. The last couple of years have obviously been crazy, I think. I think something that happens whenever you go through any of these periods like this is people take some collection of the very complicated set of facts that exist in your past, and they select the ones, you know, that they want that kind of fit the narrative they believe at any point in time. And I think that across Carvana's life, there's been a very varied set of facts. And I think there were many times when you could have selected different data points that tell a lot of different narratives. I think our narrative would be that it was always a very hard business to build. I think we would like to believe that we always had a team that was very capable.
I think we would like to believe that we have always offered customers something that is unique and differentiated. And it was always a business model that if executed well and at the right scale would deliver better unit economics than the traditional model does. And I think the very hard part of that, I think, has been expressed in what we've seen from others who have, you know, endeavored to build something similar to what we've tried to build. And then we've definitely seen it in our own journey where, you know, there were moments of euphoria, and there were moments that looked really, really dark where, you know, we went down 99%, which is kind of, like, amazing that that's even possible.
But to me, what those facts have always been true, and it's just always been a question of where you're going to make it through. Were you going to be able to get to a spot where the sum of your scale was kind of sufficient to where the business model made sense? You were still delivering experiences customers love, and investors ultimately kind of were along for the journey. And we did not see kind of 2022 and 2023 being part of our story. But I think that the good news is it caused us to basically shift all of our energy toward part of the business we were always going to have to focus on. You know, we've always had our top three priorities have been one, growth; two, GPU; and then three, driving down expenses.
I think we had investor support that allowed us to have those priorities for a long time. And then when that investor support evaporated and we clearly had to be capital-independent, it forced us to basically reverse those priorities. But I think the fundamental strength of the customer offering and the underlying business model enabled us, as well as the team, which I would like to think our team is pretty great, enabled us to go kind of, you know, unlock those positives of the business model relatively quickly. And now I think we're sitting in a spot where, you know, we're in a great competitive position. I think the unit economics are very clear.
I think we still have a lot of room to grow into the kind of ultimate manifestation of what our business model can be because we still have way more fixed costs than would be ideal. But we're tiny in a massive market offering something customers love that is very, very hard to replicate. And I doubt that there's going to be a lot of you that are going to line up to fund the next version that's going to try to build something similar. And so I think now, you know, the ball's in our hands, and we got to execute well. And I think if we execute, we have an opportunity to build a huge company. So I think it's really exciting. But it's been exciting in all the kind of different meanings that word can have, journey from the beginning to here.
Yeah. So you took advantage of a crisis, and it seems like also survival is winning. And that was kind of what I think you got is just, like, if we can survive and get the other side of this, we're winning. What did you see? What was a moment maybe you can think of in the last 12 months where you're like, "Wow, this is really going to we got this"? Was it something on the execution side, on the costs, or, you know, if you wanted to point out one or two examples where you felt like, "All right, yeah, we're going to make it"?
Well, sure. My version of the story was, like, I thought the moment where it was like, "We got this," that to me was more, like, 2018 or even earlier. I think, like, at that point in time, we were starting to roll out kind of, you know, some underlying visibility into our cohort, cost structure. People could see our GPU moving in the right direction. People could see our market share in different markets, and you could extrapolate it out. And, like, at that point, I thought that kind of the path was clear, and it was at that point, it was just execution. And then I think that the way that the narrative collapsed as we headed into 2022, which I think there were all kinds of things that came together. You know, we bought ADESA. We funded it with debt.
I think it wasn't our best execution. I also don't think it was as bad as maybe the story, you know, got collapsed to be, but it probably wasn't our best execution. Rates went up. Car prices went up. Everything got really hard. And I think, like, that was a moment where then it just turned into, "Okay, this was, like, a zero-and-straight phenomenon. It never made any sense." And I think people kind of threw the past out, and it was like, "Reprove it." And so I think in the second proof, I think that we as a company, you know, the plan we put together last year in kind of, probably April, May generally, we have, like, an annual planning cycle, and so we're putting together all our specific goals around that time.
I think we looked at that plan, and we were like, "If we can hit this plan, this is going to be a really clear story to a lot of people, and people are going to see where we are." And I think probably by about the end of 2023, it began to look like we probably were going to hit that plan. And I think that was, like, for me at least, kind of, like, the second time that I felt like, "Okay, like, we've got it, and people are going to believe us again as long as we just keep marching.
Okay. It's a tech conference. You're definitely a tech company. What you do with data and how you take a pretty traditionally horrific experience and make it a really pleasurable experience buying a car on your platform. I don't want a show of hands here. Is anyone who has bought or sold a car on Carvana in this room? Okay. Okay. That's some representation.
Yeah. So we obviously have a lot of market share opportunities.
That's true right here, even at a Morgan Stanley.
That's the upside.
Even at a Morgan Stanley tech conference. By the way, my hand went up too. I sold a car. I didn't knock on it, but I sold one, and it worked pretty well. It worked pretty well.
Oh, God. Oh, God.
Here we go.
Okay. What are we going to do?
We'll take that one off mine. No, it was good. It was good. But contingent to this is the used car market is unique, maybe to people in this room. It's a gritty market. It's a physical market. How is the used car consumer doing just in terms of the setup for, you know, the year ahead?
Sure. Well, I think that the used car market generally, I think, like, the simplest way to think about it is that it in many ways is led by the new car market. And so I think that the story that we tell about, like, what happened in the used car market is post-pandemic, there was just so much less new car production. There was a massive supply shortage there that led to new car prices going way up. That then pulled used car prices up with it. That really impacted affordability.
It also sort of caused traditional buyers that would have traditionally been new car buyers to substitute into used cars because they could afford them and because, you know, by virtue of, like, their habit generation over, you know, many, many years, they were used to going to new car dealerships where they were going to kind of buy a new car, and then all of a sudden they showed up, and there wasn't a new car for them, but there was an off-lease used car they could buy. And so I think that for the actual traditional used car buyer, I think that the market shrunk even more than used car sales did because it was, like, a substitution for, you know, used car inventory into traditional new car buyers.
And I think there's that, you know, kind of, like, famous saying, "There's never been a shortage without a glut that follows it." We clearly had a shortage. I think that it's early to call, like, a glut in new cars, but I also think it's not totally irrational that there may be headed at least directionally there. And so I think we've seen new car prices start to come down. We've seen incentives for new cars start to go up. I think it's, you know, relatively early in all that. I think that the sum of that for used cars is on net good. I think it's good because affordability, I think it's, a little bit negative because it means customers are going to be a little less likely to want to sell their car.
It's a little bit negative because it impacts consumer credit performance a little bit negatively when their car value, you know, decreases. But on net, we think that that's good, and that helps more customers come back into the market. So I think that, you know, that feels pretty good.
Affordability, still a problem but way better or way up?
Yeah. I think high-level reductions are, relative to other goods. Like, on a CPI-adjusted basis, vehicle prices are, like, around 10% higher than they were pre-pandemic, and vehicle payments, including rate, are about 20% higher than pre-pandemic relative to other goods. So I at least, like, the simple mental model that we use is that should roughly equilibrate with, like, room for fluctuation depending on, what's going on with new car supply because that's, like, a long product cycle that can be out of whack for a while.
On rates, you know, how's that? What's the direction of travel there? Have we peaked, or are we kind of?
Oh, you know better than me. Who knows? I hope we've peaked, but I have no idea.
Okay. You restructured the business. You right-sized, and that gave you the opportunity to focus on unit economics. And now a lot of people in this room and I think the message from your fourth quarter was, "We're kind of growing again," right? Because I remember last, at our Laguna conference last September, you were thinking, "I don't want to think about growth right now. I just want to, like, just focus on unit economics and focus on cost, cost, cost." And it seems like you're at that point now where you may be able to get, you know, decent growth and show unit economic improvement at the same time. Is that true? Like, are you at a point where you're not making a trade-off anymore between growth and unit economics, and you can actually get both simultaneously?
I think, you know, we're trying to complete the set of projects that we've set out to complete a year ago. I think we still got a little bit of work to do on that, but that would take us through, kind of, the middle of this year. I think that's where our primary focus is. I think that.
Fixed-cost projects or?
cost mostly variable cost reduction projects, but also, gross profit increasing projects either through, like, more intelligent, you know, credit pricing and vehicle purchasing or through, COGS cost reduction, like inbound transport and reconditioning. But we've got a number of projects we're still trying to complete there. I think that for the last 18 months or so, we've been shrinking, and it is a business that benefits or suffers from feedback. So I think when we're growing, you know, inventory size goes up, that drives up conversion. That automatically kind of causes other things to get better. You then invest in those other things, and it allows you to carry more inventory. The reverse is also true. So I think over the last 18 months, we've fought the headwind of basically negative feedback as we've shrunk inventory and shrunk marketing.
I think we're now at a place where basically, you know, inventory has been flat-ish for, you know, several months, and I think marketing has been first-order flat-ish for a bit. So I think we're now, like, moving away from the negative feedback that we've been fighting, but we're not yet leaning into growth. I think we're trying to kind of, use this terminology of a transition period which we'll head into next. And I think the goal of that transition period is going to be to shift from what has been almost a well, 100% unit economic focus in the direction of growth.
As we're making that kind of shift, feel out how well we're executing and how much we're still driving fundamental gains and how much growth we're able to drive, and then kind of determine where we kind of, you know, leave the settings of the business, based on the progress that we're making. So we're trying to not too precisely call our shot, but our goal is absolutely millions of cars, and our goal is absolutely most profitable automotive retailer. And to do those two things, we have to do both. So, you know, we're going to shift in that direction over time.
What would be the biggest opportunity on the variable cost side that you might want to highlight?
I think the biggest.
I know it's a long tail of things. Is that?
Yeah. Exactly. There's lots of things that are, like, $25 or $50 at a time. The biggest single bucket is logistics. There's still logistics is a variable function, but it's got a fixed component. You've got, like, you know, logistics hubs and management structures, and technology that is largely fixed, but flows through logistics. So I think that there's that can leverage reasonably well. Also, as we kind of build out reconditioning at ADESA locations, I think there's room to decrease both inbound transport and outbound transport. Reducing the inbound transport will be, you know, retail GPU benefit. Reducing outbound will be a SG&A reduction, through the logistics line item.
But that's just because, you know, in an ideal setup, you want to buy a car you want to ship a car from a customer that you bought it from as short a distance as possible. You have to ship it to your nearest reconditioning facility. So as you have more, there's less average distance. And then as you have inventory in more locations, through both customer preference expressed in time and also through shipping fees, customers are more likely to buy cars from a nearer location, and so you can drive down average delivery distance as well. So those are probably the biggest areas, but there's in every part of our business, there's room for improvement.
Just from the outside, it seemed like the logistics fees that you added used to not charge a logistics fee, and now you do. I don't know if you want to share what percentage or did you share what percentage of your transactions have a logistics fee, either inbound, outbound, or both. Did you?
Yeah. It's a $200 change. The majority of sales have logistics fees, but customers have a free option.
But they're, I mean, elasticity. It seems like again, you were shrinking at the time, so maybe it's hard to identify, but like a lot of other tech platforms that, you know, of companies that are presenting here, when you add that extra few bucks a month and you test the elasticity, and you realize that, you know, there might be some churn, but for the most part, did it seem to?
So I think we try to equilibrate elasticities across all of our settings. So whether it's shipping fees or vehicle price or bids or rates or whatever it is, we try to equilibrate marketing spend or inventory size, we try to equilibrate those things. So roughly speaking, we're trying to make it so those things are all balanced, and so that's our goal. But I think the other major gain in logistics that has been pretty big, like a big cost reduction benefit, has just been balancing the system. So before, we didn't have a lot of these tools that enable us to intelligently set shipping fees.
So today, for example, if you've got a route that is a, you know, 250-mile route in one direction, and then, you know, it's busy in one direction, so you're constrained one way, but you're unconstrained the other way, we now have the capacity to basically charge, you know, double distance.
Like a surge.
In the direction that you're constrained and have it be free in the direction that you're not constrained so you can balance the system again. So our whole system is just a better balanced system as we built these tools that enable us to kind of control the flow across the system.
Again, at the time you did ADESA, everyone thought, you know, it was easy to be critical of the timing and the magnitude of that shift.
Yeah. Everyone was happy at first and then mad very quickly.
Now our goal is to make them happy again.
Yeah.
But, like, I remember at the time, it was like, "Do you regret it?" You're like, I mean, when it was when you were going through the it was looking bad part, and you're like, "No. I mean, like, it's painful. You know, you can't pick the timing." But you're like, "I really, really think that we will look back and be very, very glad you did this." Presumably, you feel the same way. But I didn't know if you wanted to add any more examples of how having that that logistics and fulfillment infrastructure beefed up, even though you're running at an extremely low capacity of what you have there, why that's helped you make more optimal density decisions and fulfillment.
Yeah. So I mean, I'm obviously just slamming the Kool-Aid as fast as I possibly can. So, like, everyone here is supposed to be a lot more skeptical than I am. But my view is just it's a very you know, we now have 6,500 acres and 500,000 parking spots, connected by a logistics network that is customer-facing with a scheduler that allows you to select, you know, to the hour when you're going to have a car delivered to you across thousands of cars. It's you know, earlier, you characterized us as a technology business. Like, I don't know exactly what kind of business we are. Whatever has higher multiples, we're definitely that. But we're whatever we are as, like, a sum of those things. But part of what we are, without question, is a big physical infrastructure business.
Like, that's part of what we have to be. And I don't I don't know it's very, very hard, I think, to put together the infrastructure that we've put together. And as long as you believe that customers care about our offering and want it, you know, to the degree that we think that they're ultimately going to want it, you don't have many opportunities in time to acquire assets in that kind of chunkiness. I think, you know, we had, we had a long-term relationship. The two biggest auction houses in the country are Manheim and ADESA. We've had a long-term relationship with both over, you know, for many, many years prior to the acquisition of ADESA.
You know, there were always smaller-level conversations about, you know, "Could we acquire this site or that site or whatever else?" And I can just tell you, like, when the auction business is going well, they print money, you know, per acre, and there's no desire to ever get rid of those sites ever. You know, when you head into a time where the auction business isn't going that well, availability is there. And if you believe in the long term, then, like, it's a good time to go for it. We'll see if we end up being right on that, but we think we will.
Remember, I seem to remember, the NADA and a bunch of other dealer groups were kind of boycotting you or boycotting ADESA. Is that calmed down? I don't know if there was, like, a—was that ever material, or was that just talk? And where's that now, that animosity?
Yeah.
I always view that as a positive. Like, "Wow. If you're that worried about these guys, like, they must be doing something fucking right." But, like, I mean, yeah.
You can't say that. Can you say that? Do you want to?
Oh, sorry. Oh, sorry. Sorry. It's right. I mean, excuse me. Don't, don't.
All right. I know. We're good.
It says [Prick Hank].
Yeah. No, I heard you.
Thank you.
Yeah.
What did he say?
So I think that was a story to some degree. I think it is always at any time it's, like, media or anything like this, it's like there's always a loud minority. So I think there was more noise than there was ever business impact. But we're at bigger market shares today than we were at the time of the acquisition. It did go down a little bit post-acquisition. It's coming back now. So I think overall, the, I think our pitch like, when we talk to dealers, we just try to be straightforward with them. We're like, "Okay. Yes. We are competitive. That is true. We're trying to sell cars. You're trying to sell cars. Like, that's fine.
What's also true is we buy cars and sell cars all day long, and we're trying to build a business that serves our purposes using our tools, you know, and then make it available to you. And you decide, you know, what you want to do. And I think most ultimately decide they want to transact.
Let's talk about capital structure, Ernie. Your market cap's around $15 billion. You have around, you know, $6 billion or $7 billion of debt, maybe closer to $7 billion of debt if you include the PIK, you know, T+1 or T+2 years. A lot of people in this room are wondering.
You can't include future interest. Come on.
Well, sorry. Well, tell me.
Well, well, well, I mean, if you're but you're adding it, if it's payment in kind, doesn't it accumulate in debt?
Well, yeah. Fair enough. But it's also like a PIK cash, right? It's like you, if you're not paying it, you're at least preserving the cash. So it's if you're going to count our future interest, you got to count everyone else's future interest too, I guess is all I'm saying. But no, we clearly have more debt than is optimal. I think, you know, I go back in my head sometimes to the conversations that Mark, our CFO, and I had when we were buying ADESA and kind of making the choice to finance it with debt. And, you know, like, we tried to have all the smart discussions around, you know, what's the worst thing that can happen? What happens if the stock price drops? What happens if things get tougher?
You know, like, can we cover this from a cash flow perspective? And I think, I'm sure, like, in those conversations, Mark was closer to right than I was, although in my memory, I was exactly right. But my guess is in real life, Mark is usually the one who's right on those things. But neither one of us had on our board, "Okay. Like, people are just going to decide this model was, like, a zero-and-straight phenomenon. It doesn't work. And, you know, like, stock's going to go to basically zero, and it's game over." So I think that the lesson you have to take from that is, like, the future the distribution of possible futures is wider than you kind of thought it was, and that means that debt's a little more dangerous than you thought it was.
I think all else constant, like our optimal capital structure, we'll have less debt in the future. But we got to be smart about how we get from here to there. There are people in this room that are wondering whether you should do something, to whether you should increase equity to make that balance happen now, or are you confident that you don't have to do that? You've been able to delay it and take steps to avoid having to do that. Do you have confidence in the abilities and the ability of the business to generate enough cash to achieve that lower level of debt over a couple-year period?
Yeah.
How do we think about that? I'm not. I don't want to bait you into a conversation about cash flows or not.
Yeah.
I mean, you know, for the Q&A too.
No, you kind of do. You're saying you don't, but you kind of don't.
Well, I mean, it's a big topic. I happen to think since you asked, I mean, I think if you were to I mean, we'll never know. Maybe we will. But if you were to do some, you know, a 5%-10% type equity event, that I think your stock might actually go up on that kind of thing. Maybe not on the day, but I think that there it would be one of those, you know, based on my discussions, some investors might view you as a better risk-adjusted investment. But, you know.
Yeah. Yeah. So I would say I think, like, just to start at the end there, I do think over time, we would like to have less debt than we have today. So I think we'll manage to that over time. We'll see the pacing and the tools and everything else. What I would also say is I think, you know, I can never decide how much of your motivation should ever come from trying to prove people wrong, but I'm pretty sure the right answer isn't zero.
And I do think that, like, you know, if any of you want to take a trip down memory lane and go back and listen to, you know, any of our conference calls over the last two years where, you know, we always got the question, you know, "How much volume do you need to cover your interest expense?" And Mark would always say, "We think we can do it at our current volume." And you could basically hear the investors and analysts guffaw on the other side of the phone. And I think today, like, hopefully, if we do what we're supposed to do, I think that it'll maybe turn into, like, a maybe, and then it'll turn into, like, a head nod.
But we do think that we could cover interest and ultimately pay back debt at volumes where we're at today, and we absolutely plan to grow from here. And that's because we think the business model is a really powerful model, and we think that it lends itself to larger GPUs than many other companies have been able to achieve because we're vertically integrated, and we think the value of that vertical integration is greater than, kind of the customer offering benefits that we pass through to customers in the form of lower price. And we think that at scale, the cost can be much lower. So we think we can, you know, generate meaningful cash even at today's volumes, and we plan to grow from here.
Okay. I mean, you don't target positive free cash flow this year, but you said you can generate meaningful cash at today's volumes. This year?
I'm not going to. I don't think we want to provide more guidance than we've already provided, except we're in the abstract. We think we could generate positive cash flow at these volumes.
Okay. And is there anything that you'd highlight as to why you wouldn't be able to generate free cash flow at today's volumes without?
I think, I think the biggest input to that question will always be execution. And then I think, secondarily, it would be environment.
Okay. I'm just going to come up for air for a second. Any questions for Ernie, please? Happy to wait here. We have a microphone in a crowded room. Bob. Thank you, Bob.
So timely. I sold a Range Rover Sport this weekend to yourselves.
Oh.
It was this New York City experience. It was seamless. I got the ad through Meta, just so you know, on Reels on Friday night. I went to a dealership on Saturday morning, half a mile from where I live on the Upper West Side. Wasn't satisfied with the price that they offered. Came back. As a function of the Meta ad, I went on, and I was shocked at how seamless the whole process was. So this is not a rah-rah for yourselves because there's a pointed question.
We'll take it there.
There's a pointed question coming in a second. But it was it was seamless. Your guy picked the car up yesterday, and the cash was in my bank this morning. Staggering, but the price offered was 50% higher than the dealership down the road.
I hope it was a $100 car.
It was a Range Rover Sport, so it wasn't a $100 car. It wasn't crazy money, but it was enough. The 50% is different. So my question is, of course, you know, as you try and bring vehicles into the funnel, is there some way you can explain that dynamic other than he's got a, you know, large, dealership that he needs to sustain, and you guys don't, and you're going to take it up to the Bronx? Can you just explain that a little bit more as it stands right now?
Sure.
But the experience was incredible. I would never sell another car to a dealership ever again.
Yeah. Well, good, mostly on that. So I think here's what I think is, like, the most interesting dynamic about that question. I do think if you had visibility into each of our, let's say, wholesale transactions because that gets monetized in a way that's clean and simple, right? We buy a car from you. If we're not going to sell your car retail, we just go take it to an auction, and we sell it. And then you have, like, a known profitability on that transaction. You have we know what we paid. We know what we got, so it's very clear. The distribution of that profitability is very wide.
It's wide enough to where if you saw no one else's distribution, you would think that we were heavily exposed because it's like, at some point, someone's going to be better at this than you are, and you're going to end up with the bad side of that distribution. It's going to be brutal. But I think what is the truth is everyone has a pretty wide distribution there. You know, like, there was an experiment early on when we were launching Carvana. I'll try to tell this story somewhat quickly, but this was, like, 2012. This was before we even launched. One of the key questions was, you know, will you be able to sell cars online?
One of the key questions was, can you buy cars from people sight unseen, or are you going to get absolutely destroyed because you're at an informational disadvantage? And so we took a bunch of—at the time, I was working at DriveTime—DriveTime buyers who had previously worked at CarMax, and we randomly selected a pool of 100 cars that were trading at the Phoenix ADESA auction that next day. And we basically ran our algorithms, and we valued them. And then we had the buyers go and value all those cars. And then everyone stepped back, and no one bid on them, and we ran them through auction. And we just said, what are the values of these cars, and who had bigger average errors? And our average error on these were relatively inexpensive cars at the time.
They were on the order of, like, $10,000. Our average error was, like, I can't remember what the exact number was, but I think it was $1,200 or $1,300. And the average error of these buyers was, like, $1,100 or $1,200. It was, like, $100 smaller than our average error. But, like, the distribution was extremely wide for both of us. And I think the reason is because it's very, very hard to value used car very well. And the ultimate valuation machine, which is the auction, also is, like, it's a noisy machine where there's, like, even the same car being run through 50 times would have 50 different values. It could be pretty wide. The good news on that is that distribution as long as your distribution has a similar width to everyone else's distribution, that's just opportunity.
And that's just an opportunity to get better. So I think the internet is full of examples of where we massively overpaid for cars, and then people say, you can't make any money. But then, like, that all comes out in the wash, and you see what that is. It's also full of examples where we, you know, underpaid for cars, and people were, you know, kind of frustrated by the offer that we made. And I think our goal over time is just to get smarter with all the various data sources that we have and all the data that we're collecting from all the different transactions to narrow that distribution. And I think the better we get at that, I think we are likely already at a place where our distributions are much narrower than most.
I think there's clearly plenty of room to narrow them further. So that's the kind of opportunity when we talk about, like, fundamental gains in retail GPU. That's one of those areas. There are always more data sources coming available. And there's always more kind of depth, and intelligence we can build into our systems to collapse that distribution. And that's something that we're always working on, and that's something that we've, I think, seen gains in over the last year or two. And I think it's an area where we have three or four projects where we're trying to narrow down to one or two projects that we're going to work on over the next six months. So I think that's a bummer that there's that much width in a way, but I think it's opportunity. And I don't think we're disadvantaged there.
I think we're actually probably advantaged.
Any other questions right now? Ernie, what's your latest experience with electric vehicles on the platform? Anything you'd point out in terms of, I don't know, how quickly they move or unique issues in terms of repair or working with them on the IRCs or the values or popularity? I don't know if you had any because it seems to be quite popular, quite fashionable to beat up on electric vehicle demand lately. And yeah, I suspect that might get a little worse before.
I think at a high level, we view it as an opportunity. I think that we, our customer base skews, electric. So, like, we've always had, like, a bigger electric market share than ICE market share. I think it does require, some, like, process differences, that, you know, many of which aren't, like, particularly, like, intellectually exciting. But you do have to have just different processes to make sure that you're at the point of intake and the point of sale. You're doing different things, to give customers a good experience. And then it requires some infrastructure investment, in the form of chargers and sometimes in the form of just bringing more power into your facilities, to even support the chargers. And I think that's an area that we're, like, believers in EVs over time.
I think, EVs are at an earlier part of their technology maturation curve, obviously, than ICE cars. But I think even different places in the world point to very clear improvements in kind of customer-facing attributes of those electric cars. And we think it's a matter of time. And so we think it's a smart place to be prepared.
I didn't know what portion of your volume is EVs right now? I mean, I didn't know if it skewed higher than market just given the kind of audience that you're targeting.
It is. We haven't given that number, but, you know, we have on the order of 1% of market share, and we've definitely got a higher market share in EVs than we have in ICE.
Okay. Any other opportunities beyond used car? I mean, we talked about this. We wrote about this at Morgan Stanley quite a bit that, you know, is the used car to Carvana, like, the book to Amazon, where you start that way, it's a product that, you know, requires some unique assets in the field and fulfillment, but then over time could have multiple ways to address global mobility market, whether it's in the new or parts and service or, you know, fill in the blank, lots of anything that moves and needs to be maintained and fulfilled and kind of this very specific logistics problem that you're solving.
I didn't know if this was still something that you felt was relevant for, you know, next one to two years or whether it was something like, you know, let's just keep we just got enough to do on the used car side, and let's see where it takes us. But,
Yeah. Well, I mean, I would say it's the used car market's a 40 million car a year market, and it's a, you know, roughly $1 trillion market. That's a very, very big market. And I think it's a market where we are very confident we can offer differentiated experiences with differentiated unit economics. And so I think we're supposed to be smart in our investment and focus there. We would like to think that we have built a transaction engine that is generalizable and interesting. And I think we're undoubtedly in a world where, like, there's a lot of dynamism around what's going on, whether it's electrics versus ICE or, you know, new OEM entrants or whatever else. And I think it'll be interesting to see how that all plays out over time.
We like where we are on the board is kind of the analogy that we've used there. But the focal point is growing used.
Biggest risk facing Carvana in 2024?
I think it's always execution. I again, like, as I said, like, I'm a huge believer in it, and I'm a natural optimist. So I'm not objective or even close to it in these statements. But I just think that this is our game to win or lose from here. And I just think that we have to execute really well. I think the market's huge. I think the experience we deliver to customers are great. I think customers are going to shift in our direction over time. I think they're getting more and more accustomed to buying everything they buy online. I think there's normalization of even buying cars online, mostly through new OEMs and EVs. I think there's clearly room in our unit economics. I think we found another gear in execution over the last 18 months.
I think it's our job to kind of try to maintain that speed. But I just think that it's like, we couldn't ask to be in a better spot. And I think we also, though, we have a very complicated machine. So it's like, we have to do a good job taking that machine and scaling it up to what it can be. And that's not going to be easy. So I think execution is our biggest risk by a long way.
How would you describe any the changes of some of your online competitors? There's been some that are either smaller or, you know, have liquidated. You're in the process of helping one or more competitors, I think, doing that. I didn't know if that's in any material way affected the competition in the market, or are you seeing it being replaced just as fast with some of the other more well resourced public dealer companies filling in?
I mean, I think the best production of that is I think that we are the combination of the gap of our offering to the closest competitive offering and the realistic scales that can be achieved by those competitive offerings. I think that we're more differentiated than we've ever been. I think that it's harder for anyone to enter now. I think, like, if the success story looks like you go down 99% in the middle and you spend billions of dollars along the way, I think it's like, that's a hard thing to get behind. I think we're in a really good spot there. That's where to me, it's just it's exciting. Like, I think we just have to execute.
I think we've got, it's very hard, I think, to find yourself in a market of this size where you effectively have, you know, versus at least, like, pure play competition, a 10+ year lead. Like, I don't know that many examples of that. And so I think we have to.
It's still less than 1% or.
Yeah. And we're less than 1% of the market. So, I mean, yeah. I think that's a good combination of facts.
More questions for Ernie over on the side? Just speak up.
New record. You know, it seems like the ABS market is more constructive for your financing deals. Like, what's preventing you from pushing harder into growth, and going from this phase two, phase three combo to phase three?
I think over the last year, there are two major lessons for us to take from the last year. One is working on fewer things is extremely valuable. And going from one thing to the next when it's done is a discipline that smart, ambitious people in general don't have. And it requires, like, significant discomfort constantly to maintain that level of focus. And I think that, by virtue of not growing over the last, you know, two years, it was easier to focus because there was just, like, a smaller bucket of things to even pick projects from. And I think that led to more focus than we would have otherwise had.
I think number two is, I think, like, the parent concept is pressure, but I think, like, the immediately, controllable concept is accountability. I think that the better a job that you can do maximizing accountability on a rapidly recurring basis, like, every week you have an expectation, and you compare reality to that expectation, I think the more you can just internally generate pressure that makes you execute at a higher level because there's no room for nonsense. And I think accountability is higher when there are fewer moving pieces, because there's just not as many room or as many ways to explain away bad results. And so I think that by focusing on just unit economics, accountability was higher over the last year.
I think in order for us to achieve our goals, we have to grow, and we have to keep getting better economically. So we don't get the luxury of that maximum level of focus and maximum level of accountability. Focus, I think, is relatively easier out of the two to try to maintain because we can just continually every quarter make sure that to the point of discomfort, we're taking things off the board, and we're frustrated because the thing that we want to do, we're not going to get to do yet. And I think that it's literally, I think, managing until you have that feeling in your stomach where it, like, doesn't feel good, I think, is where you have to go if you have a bunch of ambitious people because they will explain to you why they should do it.
And you'll explain to yourself why you should do it. I think on the accountability side, I think we've learned a lot about how we set targets that maximize accountability. And I think that we'll continue to try to set targets in the same way. And I think that there will be more moving pieces. And I think it will be harder because we're going to, you know, have a joint set of goals. But I think our job is to try to just internally generate as much pressure as we possibly can so we can keep executing at the speed that we did over the last two years. And I think if we do those two things and we're right about all the inputs to the equation, then I think we're going to build something awesome.
But I think those are the two things that are, like, most important that we have to do over the next couple of years. So, I mean, the simplest thing is basically just, like, hiring and training into that capacity and then, you know, growing inventory and starting to benefit from the positive feedback and then adjusting some of the settings in the business where, you know, at the intersection of transaction volume and monetization per transaction, we are probably not at, like, the peak of what we think is possible from, like, an EBITDA on cash flow perspective today because we've been holding our operational capacity flat. And so as we made fundamental improvements, it's forced us into more monetization per transaction. That's, like, more the place where we are.
So I think that, you know, we can kind of, like, reset those settings. But we think that, like, first order, that given our current assumptions, would actually be beneficial to cash flow. And then I think it's from, like, a project, prioritization perspective, we're trying to pick things that drive conversion more so than drive down costs. And I think we're out of time, and the thing's coming out.
Thank you for your question, Bob. Thank you for your question. Ernie, I want to thank you for joining us. It's, you know, there's still so much more to do. And, you're proving the skeptics wrong. So keep doing it.
Will do. Thank you guys. Appreciate it.
Thank you.
Thanks so much.
Thank you.