Recurve Capital's Aaron Chan on the scalability of Carvana's $CVNA business model
- Aaron Chan’s differentiated bet is that Carvana ($CVNA) has entered a third rerating phase in which scalability—not survival or proof of profitability—is the central question. Phase one, “it’s not going bankrupt,” carried the stock roughly from $4 to $40; phase two, proving the model could outperform the industry, took it toward $125. With market share near 1%, sell-side unit-growth expectations in the low-20% range looked too low against early production data tracking closer to 50%, while Chan believes incremental margins can remain positive rather than deteriorate with growth.
- Carvana’s 99% collapse reflected a rare collision of operating leverage, financial leverage, and dependence on freezing capital markets—not a structurally collapsing used-car market. Industry volume only fell from roughly 40 million to 36 million units, but Carvana simultaneously overbought inventory, acquired ADESA with expensive capital, and carried costly third-party logistics and reconditioning. Since 2022 it has internalized more work, rebuilt unit economics, and reduced its liquidity risk; against roughly $1.5 billion of floor-plan capacity, utilization was under $100 million last quarter.
- The moat is a physical and financial system that would be extraordinarily expensive to recreate at Carvana’s current service level. Its DriveTime heritage supplied early facilities, loan-servicing expertise, and subprime underwriting knowledge; ADESA then helped regionalize a production footprint that had inefficiently shipped cars from the middle of the country toward the coasts. A startup buying at auction, outsourcing reconditioning, and delivering locally would pay variable costs everywhere yet struggle to match Carvana’s “almost Prime-like” three-to-four-day delivery.
- Carvana’s reported profitability looks optically incomparable with peers, but Chan argues genuine vertical-integration advantages remain after correcting the accounting definitions. Carvana excludes logistics from gross profit and includes wholesale profit in GPU, while CarMax classifies items differently; apples-to-apples rebucketing is therefore essential. Even then, sourcing roughly 80% of inventory directly from consumers, operating high-volume inspection and reconditioning lines, owning logistics, and originating financing across the credit spectrum may deliver $1,000 to nearly $2,000 of efficiency per unit.
- The short thesis clusters around the Garcia family’s control, Ernie Garcia Sr.’s legal history, related-party transactions, and Carvana’s exposure to subprime borrowers, but Chan finds the alleged distortions too small or too externally testable to carry the bear case. His estimate was that doubling the cost of all related-party arrangements would reduce EBITDA by only about 2.5%. He also rejects the idea that a roughly $400 billion ABS market repeatedly buys obviously defective paper without noticing: Carvana’s prime performance was better than CarMax’s, while Chan said some weaker early-2023 cohorts led Carvana to tighten underwriting.
- The cleanest way to falsify the long thesis would be mature-market share flattening near 5% or 6%, because that would turn a long-duration growth asset into a business approaching a visible wall. Carvana could still be profitable around 3 million units, or roughly 7.5% market share, but its multiple could compress sharply if investors could not underwrite growth beyond that point. Chan worries less about autonomous vehicles, believes EV reconditioning can be cheaper when battery replacements are avoided, and says an extra $200 rural delivery cost is manageable against roughly $7,300 of GPU.
- At approximately a $55 billion enterprise value, the valuation depends on how quickly Carvana fills infrastructure built for far more than its current roughly 400,000 units—and whether its advantages strengthen enough to extend the runway. Chan’s illustrative math uses about $5,000 of incremental EBITDA per unit: 3 million units would imply $15 billion of EBITDA, although the timing and terminal share remain uncertain. His strongest formulation is that “growth is more of a choice for Carvana than it is for any other business I’ve looked at,” with third-party fleet inventory, peer-to-peer transactions, and future marketplace distribution offering additional paths.
1. Carvana turns a hated purchase into a sub-hour transaction
Chan’s opening frame: Carvana is a “fully vertically integrated retailer,” combining financing, logistics, delivery, inspection, and reconditioning inside one system. Founded in 2012 by Ernie Garcia Jr. and spun out of DriveTime, it now holds roughly 1% of a large, stable market containing more than 40,000 independent used-car dealers.
The investment pattern Chan seeks is “disruptive companies in non-disruptive industries”—a differentiated operator attacking a sleepy category where the incumbent experience remains miserable. Walker’s family anecdote captured the stasis: a new-car purchase took five and a half hours, and the finance employee explained, “Nothing has changed. You just do this so infrequently, you forget how bad it is.”
Carvana’s Netflix-like disruption is experiential: searching, purchasing, and receiving a vehicle can require less than one hour of customer labor—roughly 10 to 20 minutes to find and check out, then another 10 to 20 minutes at delivery. The limitation is frequency: “You use Netflix every day, and you buy a car every six years,” so consumers repeatedly forget the pain Carvana removes.
Chan had studied the company for years but waited until the 2021-22 drawdown because he had not yet seen enough evidence that growth was scalable and repeatable. He had no legacy position as the stock fell from approximately $250-$300 into single digits.
2. The thesis has moved from survival to scalable incremental margins
Chan divides the rerating into three phases. Phase one was the distressed wager that Carvana would survive, carrying shares roughly from $4 to $40; phase two began with the debt restructuring and efficiency program, then ended as Carvana surpassed CarMax’s profitability around the first or second quarter, supporting a move from approximately $40 to $125.
Phase three is “the lowest IRR, but probably the biggest, longest phase”: determining how fast the company can grow, how much market share it can capture, and what economics attach to that growth. Even bullish models often assume deteriorating unit economics as volume accelerates; Chan thinks “that’s a mistake” and expects positive incremental margins.
Sell-side expectations called for low-20% unit growth, while early production data from the preceding three or four months suggested something closer to 50%. Chan stressed that the evidence was preliminary, but inventory was expanding and “the machine is kind of ramping up.”
Walker’s pushback—worth keeping—is that investors feared precisely this setup before COVID: Carvana pursued hypergrowth, “got ahead of their skis,” and suffered sharply negative economics. The unresolved trade is whether the repaired machine can accelerate without recreating the behavior behind its near-death experience.
3. The 99% collapse required three kinds of leverage to break at once
Chan’s closest analogy was American Tower around 2002, when the stock fell from approximately $60 to $0.60 before recovering over the long term. His “potion for massive volatility” combines high fixed costs and operating leverage, heavy financial leverage, and capital-market dependence just as growth slows and liquidity tightens.
Used-car demand itself was comparatively stable: annual industry volume is around 40 million units, plus or minus 10%, and the 2022 contraction was roughly from 40 million to 36 million. At sub-1% share, Carvana did not need industry growth; the damage came from consumer retrenchment, sharply higher capital costs, excess inventory, the ADESA acquisition, and impaired channels for recycling loan and inventory capital.
Before the crisis, Carvana filled infrastructure gaps with expensive vendors for paintless dent repair, logistics, middle-mile trucking, drivers, and other functions. It was therefore not yet operating the fully integrated model its architecture implied. The 2022-24 response pulled infrastructure, labor, workflow, and proprietary technology in-house, producing much of the subsequent unit-economics improvement.
Chan does not claim macro immunity: another shock could slow growth and hurt results. His narrower claim is that the previous “negative leverage on every front” is harder to recreate because organic cash flow is stronger and the company barely uses its floor plan—under $100 million drawn against roughly $1.5 billion of capacity last quarter.
4. Replicating the network means losing money before matching the service
Carvana was “born with advantages” unavailable to Vroom, Shift, and other challengers: DriveTime facilities, loan-servicing infrastructure, subprime-origination expertise, and inherited operating processes. Those advantages mattered before Carvana had national scale; its current brand, delivery density, and installed network make a fresh attack still harder.
Chan walked through the entrant’s bind. Buy cars at wholesale auctions and the challenger pays auction and acquisition fees; outsource reconditioning to a provider such as Manheim and it adds variable cost; require pickup and it remains local; offer home delivery and it must build logistics before having density. Vroom and Shift sometimes took around 20 days, versus Carvana’s roughly three-to-four-day average.
Even a deep-pocketed competitor would face physical constraints. Chan said Carvana’s Rocklin facility likely took about seven years to develop because California permitting and zoning are difficult and communities resist these sites. Before ADESA, this produced an “almost agricultural footprint”—several Ohio facilities, little coastal production, and expensive long-haul shipping toward demand.
His verdict was not that Amazon-sized capital could never enter, but that the funding proposition is unattractive after Carvana “crossed the chasm.” Asking a new company to prebuild national infrastructure and compete head-to-head now resembles building fulfillment centers to challenge Amazon: “Good luck.”
5. National branding sits above a deliberately regional inventory system
Walker challenged whether national scale matters in a historically local business with expensive last-mile delivery. Chan conceded the core point: buyers usually want a regional vehicle quickly, not one hauled from Florida to California, and the advantage is “more about what can you do regionally or locally” than exposing every VIN nationwide.
Carvana reconciles national selection with regional economics by charging customers for long-distance shipping, nudging them toward nearby inventory. Every vehicle is a “unique VIN,” unlike replicated fulfillment-center SKUs, but 200 locally available Toyota Camrys may substitute well enough for a nominal national pool of 1,000. ADESA added hubs, parking, and local production capacity to support that regionalization.
National advertising still builds awareness more efficiently than hyperlocal dealer campaigns, while the inventory and delivery algorithms operate locally. Chan’s honest non-answer on how much spending creates durable brand versus reacquires infrequent buyers was simply, “I don’t know.”
His advertising “aha moment” was instead denominator-driven: Carvana’s ads acquire sellers as well as buyers. It sources about 80% of inventory from consumers and buys more cars than it retails, yet expenses are commonly divided only by retail units sold. Counting both “sell to Carvana” and “buy from Carvana” transactions reveals operating leverage obscured by reported per-retail-unit metrics.
6. Accounting definitions explain part of the GPU gap; integration explains the rest
Walker noted that Carvana’s EBITDA per unit appeared to exceed some peers’ gross profit per unit. His comparisons included Lithia at roughly $4,000-$5,000 of gross profit per vehicle and CarMax around $2,300 on a used vehicle plus approximately $650 of wholesale gross profit—raising the question of how Carvana could earn more after SG&A than competitors before it.
Chan’s first answer was definitional: Carvana excludes middle-mile logistics from gross profit whereas CarMax includes it in cost of goods sold. Carvana also adds wholesale revenue and profit to GPU, then divides by retail units sold even though wholesale activity is not generated by that retail sale. Shipping, wholesale, and SG&A therefore require line-by-line rebucketing before comparison.
Genuine advantages remain after adjustment. Carvana acquires directly from consumers rather than paying auction economics, runs inspection and reconditioning facilities processing roughly 40,000-50,000 vehicles annually versus about 5,000 sales per CarMax location, and uses proprietary technology and assembly-line scheduling for high-volume production of nonstandard items. Chan estimates these integration gains exceed $1,000 and may approach $2,000 per unit.
Financing supplies another major delta. CarMax Auto Finance originates roughly 45% of CarMax volume, mainly prime loans, while lower tiers involve third parties and sometimes payments to place difficult credit. Carvana originates across the spectrum, with financing attached to about 80% of transactions; its DriveTime heritage creates comfort in higher-margin subprime lending that competitors may avoid.
7. The short case combines combustible optics with testable economics
Chan understands why Carvana attracts prominent shorts: Garcia family control, Ernie Garcia Sr.’s savings-and-loan-era legal history, super-voting shares, subprime exposure, and related-party dealings with DriveTime affiliates create an unusually potent collection of “buzzwords.” Carvana leases some facilities from DriveTime, Bridgecrest services its loans, and DriveTime pays commissions on vehicle-service contracts worth roughly $400 per unit.
The narrative itself has flipped. During Carvana’s losses, bears argued Garcia Sr. was overcharging it and transferring shareholder money to DriveTime; after profitability surged, they argued he must be undercharging Carvana and selling his stock to fund losses at DriveTime. Walker characterized the reversal as “really 4D chess. How can it be both?”
Chan’s quantitative rebuttal was that doubling every related-party cost would reduce EBITDA by only about 2.5%. Walker’s fair pushback was that Carvana should eliminate the arrangements if they are financially immaterial yet create constant suspicion; Chan agreed he would prefer arm’s-length counterparties, particularly where the market offers no clean external check.
Bridgecrest loan servicing does have a check: it services Carvana-originated loans even after buyers such as Ally acquire them. If servicing were deficient or overpriced, Chan argues those buyers would not keep renewing flow agreements, while ABS investors would reflect the problem in pricing and demand.
8. Subprime is uncomfortable, but loan vintages—not adjectives—decide the argument
Chan accepts that Carvana over-indexes to nonprime and subprime borrowers, aided by Bridgecrest’s servicing capabilities. What others call “unsavory”—loans at 22% or 23% interest with perhaps $4,000 down—can still be a viable product for someone with poor credit who urgently needs transportation; “it doesn’t mean the terms are attractive.”
The workflow broadens access: after a soft credit check that does not affect the customer’s score, buyers can “shop by monthly payment and down payment” across the entire inventory. Traditional dealers reverse that order—select and negotiate a car first, then disclose financing and the resulting monthly cost.
Chan’s empirical test is vintage performance. Carvana’s prime loans performed better than CarMax’s, in his analysis, while subprime results were broadly consistent with comparable paper. Recalling a 2023-N1 cohort, he estimated expected lifetime losses had risen from approximately 17.5% to roughly 22%-22.5%; Carvana tightened underwriting later in 2023, and subsequent curves returned closer to historical trends.
He called it “an arrogant position” to claim a roughly $400 billion, liquid ABS market repeatedly buys and trades Carvana paper without recognizing supposedly obvious defects. On former-employee accusations, Chan’s dozens of ecosystem calls contained both praise and criticism; fired employees and stock-owning alumni each bring incentives, so he tries to “take it all with a grain of salt” and verify the claims independently.
9. Mature-market saturation is the real risk; platform optionality is the upside
Chan’s clearest failure condition is mature-market share stalling around 5%-6%. If Atlanta and other seasoned markets flatten there, Carvana might still reach 3 million units—about 7.5% national share—and remain highly profitable, but the right-tail case would collapse as a growth stock “slams into a wall,” inviting severe multiple compression.
Autonomous transport and private-ownership erosion are lower-conviction wild cards because Chan believes Americans remain attached to owning transportation assets. EVs appear manageable: the Tesla Model 3 was reportedly Carvana’s most popular model the prior year, EV reconditioning can be cheaper, and Carvana avoids vehicles likely to need the costliest repair—a battery replacement. Rural service also works if an extra $200 delivery cost sits against roughly $7,300 of GPU.
The valuation case starts with unused capacity. At approximately 400,000 units, Carvana may utilize only 15% of infrastructure expandable to more than 3 million units with another roughly $1 billion of capital. With $7,300 GPU, $2,300-$2,400 of variable costs, and about $5,000 incremental EBITDA per unit, 3 million units imply $15 billion of EBITDA against the discussed $55 billion enterprise value.
Capacity is not destiny, and Walker pressed why a truly superior standardized platform would eventually stop at a minority share rather than take 50%-70%. Chan’s answer was optimization, not monopoly: some customers demand test drives, distrust reconditioning, dislike shipping or return terms, or accept a dealer’s “knife fight for four hours” for a lower price. Maximizing free cash flow per share need not mean serving 100% of transactions.
The longer-run option is marketplace expansion. Chan estimated peer-to-peer activity at roughly 15 million-16 million of the 40 million annual transactions, while Hertz is already placing fleet inventory through Carvana. Walker characterized the arrangement as EBITDA-per-unit neutral; Chan emphasized that Carvana captures retail, financing, and other services with light capital intensity. Chan likened third-party inventory to Amazon’s transition beyond first-party retail; future OEMs could likewise access national demand without building hundreds or thousands of dealers.
That optionality underpins Chan’s strongest claim: “Growth is more of a choice for Carvana than it is for any other business I’ve looked at.” As density increases, Carvana could shorten delivery, add pickup locations, pay more for supply, price more aggressively, or trade customer surplus across financing, convenience, and vehicle price—though “what should happen and what does happen rarely matches,” even over long horizons.
Full transcript
With me today, I'm happy to have on for the first time Aaron Chan. Aaron is the CIO at Recurve Capital. Aaron, how's it going?
Very good. Thanks so much for having me. I'm really excited to have you.
I thought we were going to have you on for another name I was researching last summer, but I'm super excited to talk about this one. Before we get started, a quick disclaimer: Nothing on this podcast is investment advice. Please consult a financial advisor, do your own work, and all that type of jazz. The name we're going to talk about is kind of controversial on FinTwit, so people should keep that in mind as well.
Aaron, the stock we want to talk about is Carvana. The ticker is CVNA. I'd be surprised if any listeners weren't at least a little bit familiar with it, but let's hop into it. What is Carvana, and why is it so interesting?
Carvana is the largest e-commerce-only used auto retailer in the country. They launched in 2012, founded by Ernie Garcia Jr., and kind of spun out of DriveTime, his dad's company.
It's unique in that it's a fully vertically integrated retailer, which means it has financing, logistics, and delivery operations. All the reconditioning and inspection work is done in-house. That makes it a pretty interesting vertically integrated machine, which, as an analyst, elevates the complication and complexity of analyzing the business.
I think there are a lot of misconceptions about the company that arise from that vertical integration and the extent of it, as well as from comparing it with others in the industry. But it's interesting because it's a scaled company. They've spent a lot of money building out this custom-built architecture and infrastructure. They have 1% market share in a large, stable end market.
The kinds of businesses I get drawn to are ones I call disruptive companies in nondisruptive industries. Find a disruptor in a pretty sleepy industry where things have been done the old-school way for a long time. If someone's building something interesting, different, and new that has a great business model attached, that's an interesting place for me to be fishing. I think Carvana ticks a lot of those boxes, if not all of them.
It's a sleepy old industry. There are 40,000-plus used car dealers—independent used car dealers—out there. Everyone hates the experience. My mother-in-law bought a car, a new car actually, a few weeks ago. It took her 5½ hours. It's just a terrible business.
She bought a new car or used car, and she didn't go through Carvana?
No, she bought a new car.
New car, okay. I was going to say, stop the podcast, short everything.
I know. You'd think I'd have some story with the family.
But she was talking to the finance guy as she was going through the process, and she said, “I thought you guys fixed this already. Why is it still taking so long?” He said, “Nothing has changed. You just do this so infrequently, you forget how bad it is.”
The way I think about it is that it's a little bit like Netflix and how disruptive the experience is from a user perspective. You can find a car, you can purchase it, and it'll show up at your door. The whole process, in terms of the number of man-hours required from the customer, can be under an hour. It's 10 to 20 minutes to find a car and check out, and it's another 10 to 20 minutes to receive the car. That's very disruptive in terms of the user experience relative to what the rest of the industry does.
The difference between it and Netflix is that you use Netflix every day, and you buy a car every 6 years. People forget how bad the experience is, and they may not have the same kind of memory or bad memory about it.
You hit on one of my questions I was going to ask later, and we'll come back to it. Let me start with the first question I like to ask every guest.
The market is a really competitive place. Obviously, you've got a position here, so you think this position delivers risk-adjusted alpha. That requires a different view. What is your differentiated view on Carvana from what the market is pricing in here?
I came into the name during the 2021–2022 drawdown. I didn't have a legacy position in it, but I'd been studying it for years and just hadn't reached the critical mass that I wanted to convince me that it was scalable and repeatable, and that the growth story would be really strong.
After 2022, I think about it as a 3-phase exit, or narrative change, and I think we're in phase 3.
Phase 1 was that it wasn't going bankrupt—a mispriced distressed situation. That probably took the stock from around $4 to around $40.
Can I just pause you for 1 second, Aaron? For those who aren't familiar with Carvana, in 2021, the stock was hitting $250 to $300 per share. As Aaron is saying, by late 2022, it was in the throes of distress, and the stock was in the single digits. I think short sellers were dancing on its grave, and we'll probably talk about short sellers at some point.
Just so people have the background: This was a high flyer that stumbled for a lot of different reasons. I think Aaron is just saying that he came onto the stock after the stumble, when he mentioned $4.
I wish my basis was at $4, but I think I bought it at $3.90. There was peak fear. The capital markets were freezing up and seizing up. This is a business that requires a lot of capital markets activity just to cycle the capital as they originate loans, buy inventory, and do all these things to keep the machine running. There was a lot of fear around all of that. I'm sure we'll get into it.
That was phase 1, and you could have made a lot of money just betting that it wasn't going to go bankrupt.
I think phase 2 was proving that this is a good business model and the best business model in the industry. That phase began when they did the debt restructuring and fixed the balance sheet to give them room to implement all the efficiency measures they wanted to put in, roll out all the proprietary technology they'd been working on, stabilize the business, and improve the unit economics.
Then they surpassed CarMax's profitability earlier this year, in Q1 or Q2. I think that was the end of phase 2, which took the stock from $40 to $125. Now we're at roughly double that price, plus or minus.
We're in phase 3, which has the lowest IRR but is probably the biggest and longest phase. The question is how far you can take this model and how fast you can grow.
When you ask what's missing from the market's view, I think there's still a lot of confusion around the scalability of the business model. Even very bullish people don't model positive incremental margins from here. They model declining unit economics as the business scales. I think that's a mistake.
I think that's a big one. Then there's a debate about how far this business model can go. How high can market share go? It's at roughly 1% today, plus or minus. They're growing fast. They've turned growth back on. They're growing inventory. The machine is ramping up.
The question is: As the machine ramps up, the street's sell-side expectations have low-20% unit growth for this year. The early data suggests something much higher than that—more like 50% so far. It's still very early, but we'll see.
You could look at their production rates over the last 3 or 4 months and get numbers that are much higher than what the street is modeling. So it's a debate about 2 things: How fast can they grow, and what are the economics attached to that growth as they accelerate?
What everyone fears is that they went through a hypergrowth phase pre-COVID, and the stock fell 99% because they mismanaged the growth. They got ahead of their skis—not just a little bit—and the economics went strongly negative as they hit the gas on growth. People fear that that's going to come back again as they scale.
Let me ask a question on that. One of the things that has scared me—and I think my first notes on the company were from 2018 or 2019—I've loosely followed it for a while. I think I had friends who pounded the table at $300.
I've had friends who pounded the table at $3. One of the things that's always scared me here is that it went from $300 to $3. I've seen stocks go from $100 to $30 on no fundamental news, but it went from $30 to $3 because they were having real issues.
I guess one of the things that's always scared me is whether that was because they got over their skis and did the ADESA acquisition, which stretched the balance sheet, or because this business is a lot more cyclical than people think. Yes, it's selling used cars at its core, but maybe there's something about the model that's just hugely cyclical. Right now it's working, but in 10 months we could be saying, "Hey, interest rates ticked up. We'll talk about the consumer, I'm sure—consumer sentiment got a little bit worse—and this huge fixed-cost infrastructure means they're really laboring under it, and the stock's gone from $250 to $25 again," or something. Does that make sense? It's just hard for me to marry this category-killing company with the near, near-death experience they had in '21 and '22.
Yeah, it's very rare—the dive down and the bounce up is a very rare thing to happen in the public markets. You know, the base rate for a stock that's down 99% is that it doesn't come back. This is the exception that proves the rule, right? Ninety-nine percent of the time, they don't come back.
To me, the closest analogy I've come up with is American Tower in 2002. It went from $60 to $0.60, and then it's bounced—you know, what is it?—a couple hundred dollars a share now. I think that's right, yeah.
What you need—the combination, the potion you need to create a 99% drawdown that isn't existential—is super-high fixed costs, a high-operating-leverage business model where growth slows at the same time as high financial leverage, and dependence on capital markets along with a fear of liquidity tightening. If you get those 3 things combined, you have a potion for massive volatility, because the equity just starts getting priced for maximum fear.
Having lived through '07, '08, and '09, you saw some of that back then. 2002 is a little bit before my time; I started my career in '04. But the first industry I studied was the cell-tower industry, so I was very close to studying that historical period in '04, after they had bounced out of the trough.
I think Carvana diving this much, or drawing down 99%, is more a function of high operating leverage, high financial leverage, and capital markets seizing up—all happening at the wrong time. Growth slowed at the same time that capital got more expensive. The consumer pulled back, and the traditional channels for them to cycle their capital as a dealer started to either get more expensive or become a little more difficult. Then you have peak fear surrounding something, and it never should have gone down as far as it did. But equity guys can freak you out pretty easily.
What's interesting going forward is that those characteristics aren't really present anymore. From a macro or industry perspective, used cars are pretty stable. It's 40 million units, plus or minus—call it 10%. I think in 2022, during that drawdown, they went from roughly 40 million to 36 million.
So, you had a 10% drawdown in units, but we're talking about a company that had less than 1% market share at that point. It shouldn't have necessarily impacted them. Right now, I think Cox Automotive estimates this year's growth at 2.5%. The industry growth rate doesn't really impact—or shouldn't really impact—Carvana. It shouldn't be a governor on Carvana whatsoever at 1% market share.
It shouldn't have been back then, either, but you had the cost of capital spiking aggressively as interest rates went up. You had consumers pulling back, and you had the company overbuying inventory and buying ADESA at the same time. You had this potent mix: expensive capital to buy ADESA, expensive capital across the entire capital-market complex, and liquidity being pulled out of the system. That was the combination of factors.
Now, the question, to your point, is what happens from here? Why can't that happen again? I think what they did after 2022 was focus on the internals of the company—the efficiency and the unit economics. They were growing at any cost previously and using a lot of third-party vendors to do certain functions, like paintless dent repair, or even logistics and the middle mile of the network, using third-party trucks and drivers and just filling in gaps with third parties, which was much more expensive. That wasn't a fully vertically integrated way to run the business back then.
Now, they've pulled everything back into first-party infrastructure, labor, workflows, and processes. I think they've fixed a lot of that, and that's where a lot of the gain happened from 2022 to 2023 to 2024. I think they won't go back to that anymore. They know they've learned their lesson there.
I also think the cash flow of the business is now so strong that they don't have to worry so much about capital markets. If you look at their floor plan, this is an auto dealer that basically doesn't use its floor plan. That's a pretty fascinating thing. I think they have $1.5 billion of capacity on the floor plan, and it's under $100 million of utilization on it last quarter. So, they have plenty of sources of capital if they need it.
But I think the organic business is now on such strong footing that if they got hit by a wave of disruption in the economy or anything like that, they could absorb it a lot better now. It doesn't mean they won't be impacted, and it doesn't mean growth wouldn't slow. But to have that kind of negative leverage on every front, I think, would be a lot more difficult to see from here.
Let me ask a different question. As you and I are speaking, the stock price is around $250, with a $50 billion market cap and a $55 billion-ish EV. I'm just looking at the balance sheet: $2 billion of equity capital in, $5 billion or $6 billion of debt in. So, invested capital is $8 billion; total liabilities are $7 billion. If you want to call it $10 billion, whatever.
That's a lot of value creation that the market has given them, right? $10 billion of invested capital or less to $55 billion of value. When I look at this, Carvana is online used-car retailing. As you said at the beginning, there has been no one for the past 50 years who's been saying, "Hey, the used-car-selling experience can't be improved."
But when I look at Carvana, I say the market is signaling something. The market is a signaling function. It's signaling that they're going to produce a lot of cash flow. I think you published a rebuttal to the Hindenburg piece, which I should have mentioned earlier, that has them doing about $2 billion of free cash flow in 2027 or 2028, in the mid case.
That's the equity capital invested here. The market is signaling that someone should come in here and create a competitor. But to my mind, the used-car dealers haven't really tried. Vroom went bankrupt. There was 1 other competitor that shut down. CarMax hasn't done anything nearly like the economics here or the integration here.
So, I guess my question is: Why is no one copying the Carvana model, and why is Carvana's model so moaty? I've seen high-growth, high-fixed-cost businesses before, and when you've got the signal that says, "Hey, we'll reward you with 5 to 10 times the invested capital in market cap," that eventually does attract people. What is so moaty about it that prevents that?
We've seen people try, all right? I think your point on Vroom, Shift, and others failing is pretty telling. I think Carvana was born with advantages that others didn't have. It was born with the advantages of being able to use DriveTime's infrastructure and processes.
The initial IRCs were DriveTime facilities that they leased from them. They got to leverage the DriveTime expertise on loan servicing, subprime origination, and things like that. Spinning out of a company that has a lot of experience, infrastructure, and know-how gave them a pretty nice head start on vertical integration.
I also think the mere existence of Carvana at this size and scale—a national brand with a national ad market—makes it much harder for a new competitor to come in. How would you or I try to enter this market? Let's go buy cars at wholesale from auctions. If we want to do capital-light, let's bring them to a third-party reconditioning center, maybe pay Manheim to recondition the cars to retail quality.
We’re paying variable costs on acquisition. We’re paying auction fees and variable costs to recondition the car. How do we deliver to customers? We have them pick the cars up. If so, we now have a pretty narrow market. We’re a local player, not even a regional player.
Then we have to market only locally using whatever media tactics we have to address a 50-mile radius, maybe. It’s a pretty narrow market—maybe a 200-mile radius if you want people to pick up the cars. If you want to deliver a car to end customers, you have to figure out how to take those cars and deliver them to people’s homes within a reasonable time frame.
If you look at Vroom and Shift, the delivery times are crazy, right? Sometimes it can take 20 days to get a car. So, matching Carvana’s service level—which is almost Prime-like in its delivery times—is not quite there. It’s probably 3 to 4 days, roughly 4 days, on average right now.
If you’re coming to the market as just a local guy, using all third-party reconditioning and outsourced labor and infrastructure, it’s not only way more expensive for you, but you can’t even match the service levels Carvana is offering today. Now, does that mean someone like Amazon couldn’t come in and spend $20 billion to build all the infrastructure?
They could say, “Okay, we’re just going to acquire land,” which you probably can’t do because you can’t get it zoned very easily. If you look at the Rockland IRC that Carvana built, I think it took them about 7 years. That’s because it’s California, number one, and it’s just hard to get these places zoned and permitted because nobody wants one of these facilities locally.
That’s why Carvana’s production footprint was so unusual. They had 3 facilities in Ohio because they could get them zoned and permitted there. They had nothing on the coast before ADESA, right? It was this weird production footprint, almost like an agricultural footprint, where they were producing in the middle of the country and shipping out to the coast. It’s very expensive to do it that way.
Now they’ve kind of filled in the network with ADESA. But as a new competitor, how do you get turnkey capacity to compete nationally if you want to do it efficiently? How do you offer service levels that are competitive with Carvana? And how do you make money doing that?
Everyone else was eating through losses. Carvana is the one that crossed the chasm. That doesn’t mean someone like Amazon or speculative capital couldn’t come in, but I feel like the barrier is: who’s going to fund it? Who’s going to want to compete against Carvana head-to-head at this point?
It’s like competing against Amazon, excluding Walmart. Some random new company comes in and says, “We’re going to build fulfillment centers and compete against Amazon.” Good luck.
Just real quick: they’ve got the national scale, right? You mentioned that, and obviously it gives them some advantages. Whenever I watch the NBA, guess what I’m getting hit with? Dax Shepard and Kristen Bell pitching Carvana, almost trading in your used car like a shot.
There is some advantage there, but it does strike me that the used-car game was a local game for 100 years. Most of the costs in the used-car game, if you’re going to deliver, are in the last-mile delivery, right? This is a really heavy thing. You’ve got to have one truck—I think Carvana has 1 or 2 cars per truck. You have to hire the driver, and they have to bring it.
They’re competing against the hidden cost of the old model. I would go to the car dealership, buy and pick up my car, right? That was a cost that was kind of hidden there.
My question is: how much of an advantage is the national level? How many times are they delivering used cars in Florida and matching that with someone in California? It does seem like this is a local game. That doesn’t mean it can’t win, but it seems like a local franchising game, where most of the advantages are at the local level versus the national level.
I don’t know if I’m quite making my point, but I think you see what I’m trying to say.
No, I think I get it. I’ve thought about this a lot, actually. When I first picked up Carvana in late 2021 or early 2022, before they announced ADESA, I was digging into it and thought there was too much long haul in the network.
You can’t offer national inventory with all these long hauls, given the nature of their production footprint. It felt really inefficient and really difficult to scale economically.
I think what they’ve done since is 2 things. Number 1, they got ADESA, which helped them regionalize a lot and establish hubs, parking lots, and local production to some degree. They’re ramping that up.
Number 2, I think they—and Amazon, in parallel—regionalized their inventory a little bit better. How do they do that? I think you can buy nationally if you want from Carvana, but you’ll have to pay for the shipping. So, they economically push people toward the local inventory.
You don’t need all 1,000 Toyota Camrys that are in inventory, or whatever the number might be. You may only care about the 200 that are in your local market because that’s enough. There’s enough replication of similar makes, models, and years that you don’t have to open up the entire national inventory for that.
So, I think you’re right: it is a regional business. Most people are looking at what they can get regionally and quickly. Nobody wants to wait 8 days to get a car shipped from Florida to California if they don’t need to.
It’s a difficult business because you’re not replicating SKUs like toilet paper or paper towels in a fulfillment center. There’s no replication here. These are unique VINs, right? You’re managing a network or an inventory of more than 50,000 unique SKUs, but enough of them are similar enough that you can optimize the network in that way and have enough regional inventory to meet the needs of the local market.
I tend to agree with you that the advantage is more about what you can do regionally or locally, as opposed to nationally. But you listen to any kind of cable M&A call from the last decade or more, and all they talk about is how, when you get to buy nationally and advertise nationally, that’s your ad market instead of a hyper-regional, hyper-local ad focus. Your efficiency goes way higher when building a brand.
I think they can do it in 2 ways. They have the national branding, and then they have the local inventory-management optimization and delivery.
Okay, quickly on brand. They do talk about leveraging advertising. I’m just remembering the Q3 earnings call, where they showed their long-term model and talked about leveraging advertising.
I do wonder: you talked about building a brand, but we mentioned up front that you buy a used car every 7 years. I know some of the biggest bulls hope that Carvana makes the experience so good and cuts out so much cost that maybe people go from buying every 7 years to every 5 years or every 4 years. Then you actually go from 40 million to 48 million, and I’m going crazy.
Just on the brand, is it really buying a brand? I grew up in New Orleans, and the used-car dealerships advertise a lot. I always think it’s a little bit of a brand, but I think it’s mainly customer acquisition cost, right?
You buy from a car dealership in 2000, you forget about them, and then in 2006, when it’s time to get a new car, you’re open to whoever. They’re actually just trying to reacquire the customer. So, how much of the advertising is brand building versus customer acquisition cost?
The short answer is, I don’t know. The mix between the 2 is something I don’t actually know. Let’s drop it, because I don’t think it matters to the long term.
Let me talk about something that does matter. But let me say one thing on advertising that helped me in my analysis of the company. It helped me see where they were getting operating leverage when it didn’t look like they were.
The advertising and all the variable expenses aren’t just for attracting retail sales. They’re also attracting demand for selling to Carvana, which is where they acquire 80% of their inventory, right? They buy more cars from customers than they sell to customers.
That’s where their wholesale inventory comes from, which they clear in the wholesale markets. One of my aha moments when I was studying Carvana was realizing that they had gone from roughly 20% share of sell-to-Carvana inventory to roughly 80%, but they were dividing every metric by retail unit sales.
If you look at total retail transactions, they were effectively doubling the total retail transactions as they moved from 20% to 80%. They have the buy-from-Carvana transactions, and they have the sell-to-Carvana transactions. Because they divide everything by retail units sold, you’re missing—or at least I was missing, until I realized this—how much operating leverage they’re getting on advertising and on variable operating expenses.
They have to go pick up your car if you’re going to sell it, right? But that cost is not included in a retail sale.
And most of their sell-to-Carvana volume is not going to a trade-in directly. It's usually an isolated transaction: I'm going to sell my car to Carvana, and separately, someone else is buying a car from Carvana. But that's a last-mile visit that has to be paid for, and it's divided by retail units sold, so it never shows up. There's advertising to acquire that inventory that doesn't show up, either. So, to me, you can see the operating leverage in the business more easily if you look at it that way and look at total retail transactions instead of retail units sold.
That is a fascinating insight. It also speaks to every commercial I can remember from them, like the Dax Shepard and Kristen Bell commercials. It's them selling their car, not them buying their car, so they're clearly looking for the inventory. If I sell a car to Carvana for $15K, do they offer, “Hey, $15K in cash, or $15.5K if you use it to buy another car from Carvana?”
No.
Okay. Okay.
Not that I'm aware of. When I traded my car into Carvana, it didn't matter that I was buying one. It wasn't like that.
I'm just curious. Quick question on the economics. Again, I'll link your rebuttal report to Hindenburg, and we'll talk about that in a second. You have $4,250 of EBITDA per unit in 2027 or 2028 on 1 million units in the medium term. The company, I think on their Q3 call, mentioned, “Hey, we think we've got the infrastructure right now in place to grow to 3 million.” So, obviously, there would be more operating leverage before that.
But I look at that $4,250 number and look at some auto companies, and Lithia Motors, ticker LAD—a big, mainly new-car retailer—their gross profit is $4,000 to $5,000 per unit, and that includes new cars, which obviously are going to have a larger profit than used cars. CarMax, I'm not crazy familiar with, but I pulled up the 10-K and the 10-Q, and they're—again, this is gross profit, which is above EBITDA—at $2,300 of gross profit per used vehicle and maybe another $650 of wholesale gross profit. So, let's just call it $3,000. I'm not sure how the other relates, but $3,000 in gross profit before SG&A, right?
So, I look at your numbers. Even right now, I think Carvana is doing EBITDA per unit above the gross profit per unit of a lot of their peers. Obviously, it's naturally scaled, but we've talked about how a lot of these costs are pretty fixed and are going to be pretty similar for Carvana versus others. So, how are they getting economics so much better than peers?
I think there are 2 main points to talk about on that. Number 1 is a definitional difference. Their gross profit is defined differently. They exclude logistics costs from their gross profit; CarMax includes them. So, logistics for Carvana—which is middle-mile logistics—is embedded in the cost of goods sold for CarMax, but it's not for Carvana. You have to bucket and rebucket the costs and adjust for the definitional changes and differences between the companies.
They also include wholesale, and it's transparent. I'm not saying they should or shouldn't, but everyone should make the adjustments as appropriate. They add wholesale revenue and gross profit to the GPU metric, which technically has nothing to do with a retail sale, but they put it in. Again, when I talk about dividing by retail units sold, just as there's something to discover about the cost structure when you look at total retail transactions, there are other things to discover when you're dividing things that don't relate to the sale by retail units sold, like wholesale.
I think that's one avenue of difference, and I've tried to compare apples to apples. I thought J.P. Morgan did a pretty interesting job around this Hindenburg situation earlier this month, comparing CarMax to Carvana line by line through GPU, shipping fees and shipping expenses, SG&A per unit, and things like that. So, I think if you look category by category, it makes sense.
However, Carvana is still more profitable. The second point is that I think their vertical integration—and how deep the vertical integration is—generates over $1,000, probably closer to $2,000, a unit in efficiency gains. If you look at some of the things related to the processing of a unit, not having to buy from auction because you're buying from customers, that's better margins for you than for the others. You'd have to compare like for like to understand that relationship and how they compare.
They have their own shipping and logistics, so their IRCs are more efficient because they're much higher volume. Those are doing 40,000 to 50,000 units a year. CarMax per location does about 5,000 units of sales, and they do most of their work on-site. So, they've arranged the infrastructure differently for higher volume and higher scale.
There's more proprietary technology. It's much more of an assembly-line inspection and reconditioning process. I visited a few of their facilities, and it's impressive what they do. It's much more automated, and the processing times are very scheduled. If you were doing high-volume production of nonstandard items, this is probably what it would look like. So, I think they have efficiency gains there.
Then the third bucket—or the second bucket within the higher-efficiency, vertical-integration part of the delta with others—is the financing operation. Their finance gross profit is higher than everyone else's. If you look at CarMax, about—I think—45% of their volume is originated by CAF, their captive CarMax Auto Finance. That's where they have the full stack of profitability. It's mostly around prime originations.
Then they use Tier 2 and Tier 3. Tier 2 is where they get paid a little bit to originate or flip the loan to a third party, or—sorry—a third party pays them to originate the loan on their behalf. For the tail, Tier 3, they're paying a third-party lender to take that loan just to make the sale.
Carvana is a full-stack, vertically integrated lender. They originate all their own loans. About 80% of their transactions have a loan attached, and there's a mix of prime and subprime. They go across the credit spectrum. I think because of their heritage with DriveTime, they have comfort in subprime that others don't, and the margins are higher in subprime. So, I think that explains the bulk of the difference.
Again, there's a definitional difference, and there's a different business mix for each of these companies. If you adjust for the mix differences, it's not that hard to get to Carvana's level of profitability. So, you have the mix differences, and then you have the definitional differences. You combine them, and this is how you get to Carvana's numbers.
Perfect. Okay, that makes total sense. Well, let's just go to this, because I think we'll address it. I want to talk valuation second, but let's go to the elephant in the room.
I don't know if there's ever been a company that has attracted more short sellers than Carvana. When I was prepping for this, Hindenburg—the reason you and I connected is because you published the rebuttal report to Hindenburg—published a piece at the beginning of 2025. Kerrisdale, one of the most high-profile short sellers out there, published a piece at the beginning of 2024. I searched Jim Chanos and Carvana, and sure enough, I think in 2022 he said he had a short in Carvana. Those are 3 of the most famous short sellers out there. I didn't look at Muddy Waters; that would probably be the 4th. And then there was also a Spruce Point one, wasn't there?
No.
What hasn't there been at this point? This company is a flytrap for investors who love quality compounders and revolutionizing an industry—exactly what you described up front, right? You and several other high-profile bulls I know make these types of investments in things that can just explode and capture a huge piece of the industry.
And then it's also a Venus flytrap for short sellers who focus on all of the issues that I think we've addressed so far. They talk about accounting issues. You hit subprime; subprime is a huge mention with these guys all the time. So, we could go line by line through the Hindenburg report—you basically went line by line through a lot of it—but I just want to ask overall, at a high level: why are so many short sellers attracted to this company?
Honestly, I don't get drawn to this much controversy normally. It's been fascinating to go through this experience. People love to hate Carvana, and I don't know why. I think it's a couple of things.
Number 1, the Garcia family controls the company, and Ernie Garcia Sr.'s legal history has been easy bait.
Can you describe the legal history for people who don't know?
Ernie Garcia Sr. got involved in the savings-and-loan scandals, I think—or one of them—in the late 1980s or early 1990s, if I remember. That's obviously not the kind of history you want from the founding controlling shareholder. So, I think a lot of people just assume that there's something nefarious going on because he controls it. He has super-voting shares.
I think a lot of people make the easy jump—and the short sellers are obviously much more sophisticated in this—but they're like, “Hey, controlling shareholder with a legal history plus subprime. Could you put together 2 better buzzwords?” Plus, there's the related-party issue because of the heritage of the company and the fact that it's spun out of DriveTime.
Carvana uses a lot of DriveTime infrastructure, and this is disclosed: they lease facilities from DriveTime. DriveTime also services loans, or affiliates of DriveTime do; Bridgecrest services the loans that Carvana originates. Then there are vehicle service contracts, or VSCs, which are about $400 a unit of gross profit for Carvana. DriveTime pays Carvana a commission to originate those VSCs.
When you have legal history, subprime, and related-party transactions between the companies, you wonder whether these are arms-length transactions. These don't seem like arms-length transactions between the 2 companies. Therefore, which one is subsidizing which, if at all?
I've heard things like, you know, the funny thing is that the narrative changed. In 2021 and 2022, when they were going through all the negative EBITDA, the argument was, “Oh, Ernie Garcia Sr. is screwing Carvana. He's overcharging for everything because he's plowing money straight from Carvana shareholders into DriveTime's pockets.”
Now, even though the terms haven't changed, Carvana is so profitable that Ernie must be undercharging and not giving them market rates, and he's selling his stock to fund the losses he's taking at DriveTime. It's really 4D chess. How can it be both?
In your short report, I always had related-party transactions in my head. I think what I liked most is that you actually laid out the math and said, “Look, at this point Carvana is growing so big that even if you make reasonable assumptions about the related-party transactions, you're talking about a very, very small amount.” I think you laid it out as 2.5% of their EBITDA would be coming from the related party, even if you made some pretty conservative assumptions and weren't giving them the benefit of the doubt.
Yeah, I think if you double the cost of all these things that they do together, it would be a 2.5% impact to EBITDA. It's just not big enough to commit any kind of improprieties over. People have different definitions of what's big enough, I guess, but why would you put at risk—if you're the Garcias or Ernie Garcia Sr.—your tens of billions of dollars of value in the stock over 1% of EBITDA?
Well, I think that's a good question, but my pushback would probably be: Why do they continue to do it if that's the case? It's such a small thing, and it causes so much consternation.
I agree. I wish that they would just have arms-length transactions and counterparties with unrelated parties to do these things.
I do think that Bridgecrest—well, why do they use Bridgecrest? First of all, Bridgecrest services all loans originated by Carvana held by everyone. When Ally buys a loan from Carvana, Bridgecrest is servicing it. If Ally had an issue, you better believe we would have heard about it by now. They're not going to keep rolling this MPSA, the flow agreement, if they hate the servicer or if the servicer isn't doing its job.
When you have legal history, subprime, and related-party transactions between the companies, you wonder whether these are arms-length transactions. These don't seem like arms-length transactions between the 2 companies.
For the extended warranties, vehicle service contracts, and GAP waiver insurance, we don't necessarily have as much of a third-party check on those because they aren't as fluid in the market. I don't know. I wish they would, but I'm with you on the ABS side.
On the ABS side, if they were charging too much for the servicing, the ABS buyers wouldn't be getting their return, and nobody would buy the ABS, right?
Exactly. You would feel it in the market response, right?
The ones that get picked on are the ones that don't have a proper market response. I think one of my bigger points in the rebuttal piece was that the argument that they originate terrible paper that's getting bought up by related parties indirectly—first, let's just ask: Is the paper good or bad?
I think if you go through the loan performance, which is what I did in that piece, their loan performance is better than CarMax for prime, and it's definitely in line with other subprime. Were there some cohorts that performed worse than expected in the early 2023 vintages? There were, but they tightened the standards, and now they're back to trend lines.
To make an argument that this $400 billion ABS market, which is buying paper from Carvana regularly and repeatedly and can trade it in the market with each other, is so stupid that it doesn't notice that Carvana's loans are bad—I just think it's an arrogant position for Hindenburg to take, to say that this large and liquid market is completely wrong. “Don't look at the performance; just trust us that it's not good paper.” It's an absurd argument. I don't know who would ever make that argument, honestly.
I don't disagree there. A lot of the short reports do seem to focus on the subprime nature, right? I think there is some worry that if these guys are underwriting bad—really subpar—loans, maybe the environment over the past 5 years has just been really good. You've got the famous Buffett saying that the tide is not out yet, but maybe if the tide goes out and there was huge demand for lower-income workers over the past 5 years, with labor shortages and the minimum wage going up dramatically, maybe if some of that changes a little bit, all of a sudden these loans start looking a little bit worse with some seasoning.
I don't know if I'm making that up, but that seemed to be the point. All the short reports just hit subprime, subprime, subprime. So I'm trying to be generous with the point that could be made here.
Yeah, they do over-index to subprime and nonprime. That's a known risk, I guess, for Carvana. I think they have a specialty in it, probably because Bridgecrest is really good at servicing, and DriveTime and Bridgecrest are really good at servicing those loans. They have embedded history and expertise in making money off of that.
It's like arguing that Credit Acceptance must be a horrible business because of its customer base. It's funny because Credit Acceptance—this is CACC, if I remember the ticker correctly—has been a very popular short over the years, and the stock is, what, like a 100-bagger or something.
I was thinking as I was researching and preparing for this, “Look, everyone, look at Credit Acceptance's subprime, all this sort of stuff.” The stock has just kept performing for 10 or 15 years, and it's done incredibly well. I was wondering whether the same misguidedness of the shorts in Credit Acceptance was happening with a lot of Carvana.
Yeah, I think it's just a little bit unsavory, right, to be lending at 22% or 23% interest rates to the customer base that would be accepting those terms. I think people feel uncomfortable, and it feels unsavory to them, but one of the value propositions that Carvana has is not just the convenience factor. It also matches supply with demand across the credit spectrum.
It's a very different workflow versus going to shop for a car normally. Normally, you find a car, test-drive it, talk to the sales guy, negotiate the price, and then go into financing. Then you figure out what your monthly cost is. Carvana democratizes that whole process, or the whole inventory, and says, “Okay, give us a light credit check that doesn't impact your credit score. You can then go shop by monthly payment and down payment across the inventory.”
It's a very different workflow, and I think they've done a good job of getting people into cars who maybe wouldn't be welcome at a CarMax, or would be less welcome at a CarMax or another used-car dealer. That doesn't mean the terms are attractive, right? They might be super-onerous—20% or 23% interest with $4,000 down—but maybe they really need a car, they have bad credit, and this is just what the market will bear.
I think the concern would play out if their cohorts in each of their vintages were really underperforming and getting worse. I think they have pretty steady performance across the vintages. Like I said, some of the early ones in 2023 were probably a little too loose, and so the cumulative net loss expectations have gone up.
I'm going from memory here, but if I think of a 2023-N1, it was 17.5% expected losses over the life of the ABS, up to around 22% or 22.5%, maybe. So they have definitely stepped up, and I think they tightened their underwriting in late 2023. Their curves have come back down to more normal levels versus historical performance.
Plus, one of the short sellers—and this is an awkward one to ask, but it is addressing the elephant in the room—all the short-seller reports, particularly Hindenburg, but all of them, kind of—I mean, Hindenburg comes out and says it, but they allege that things aren't as good as they seem, right?
One of the quotes from a former director is, “DriveTime is like Fight Club: nobody talks about it,” even though it's the big elephant in the room.
We've already addressed DriveTime, so we don't have to address that one specifically, but it is always a little scary when you see short sellers come out and say, “Hey, we talked to 50 former employees, and all 50 of them were like, ‘This is a flaming pile of poop, and everyone should avoid them.’” So how do you mesh that with the fact that this is an experience people seem to like?
I've talked to people who bought Carvana. My friends have, and they've liked it. I don't have a car, so I can't buy one, but it is scary when you see that. So how do you mesh those 2 things?
It's a good question. I do my own calls with people in the ecosystem, former employees, and all that. It's strange to hear that kind of feedback when I've had dozens of conversations across the ecosystem and with former employees, where there are pluses and minuses in every call, right? Some people got fired, and they're really pissed, and they want to talk badly about the company. Some people left on great terms, and they still own the stock, and they want to pump up the story, right?
You have to take it all with a grain of salt. I just try to verify with my own work and not worry about what other people are figuring out or publishing. If you go to Tigus, AlphaSense, Third Bridge, or whatever, and read about Carvana, I think most of them do not agree with the conclusions that Hindenburg put out there.
No, it's one of the tough things about expert calls. You talk to a former employee, and sometimes the former employees are just ripping it down. Then you can even go talk to the manager sometimes and be like, “Hey, I talked to your former VP of accounting.” They're like, “Oh, yeah, that guy who got fired for drinking on the job, and, like, none of his numbers were correct? Yeah, he might not have been super happy with us.”
I've also seen, “Hey, it was between the CFO and the COO for the CEO job, and the CFO gets chosen.” Then you talk to the COO a year later, and he's extremely bitter. He's like, “That guy sucks.” And no, he doesn't suck. Maybe you're better than him, but for one reason or another, he won.
Anyway, I thought it was worth asking. I have 2 last questions, and then we can wrap it up. We can talk more Carvana. We can start talking cruises if you want. I've always had an obsession with the cruise line stocks.
The first question is: We talked about a ton of stuff, but there's a ton of stuff we haven't talked about. What if you and I were sitting here 3 years from now, 5 years from now, and Carvana hasn't worked for some reason? You can define “hasn't worked” however you want: the stock's flat over 5 years, the stock's down 99% over 5 years, however you want to define it.
What do you think would be the biggest thing that, 5 years from now, you say, “Hey, Andrew, I underestimated this,” or “I didn't realize this,” that caused it to underperform?
Yeah, what's interesting about Carvana is that it's not obvious where a competitive threat—a direct, natural competitive threat—would arise. We're talking about a super-fragmented industry. Normally, I'm always concerned about competition and someone undercutting or someone disrupting in some way.
I don't have an answer; it's hard to articulate why that would happen because this is an industry where, outside of Carvana, the other players aren't really set up to grow very fast. They have 1% market share, so disruption affects the field much more than it affects Carvana. I put that one to the side for the most part. That doesn't mean something crazy can't happen.
Since, like you said, most of the business is done locally, and it's a local negotiation on pricing, they can tweak their algorithms and data science to determine how to price in Atlanta versus Los Angeles and optimize the inventory to maximize yield from it.
Then it's a question of—I think the biggest question mark is whether you saw the cohort slowing in the more mature markets, with growth slowing. Atlanta is our first market, right? If you saw that curve flatten out, that would be a concern.
It's been a weird period because we went through 2022 and 2023 with no volume growth. In 2024, we're getting back to growth. 2025 is going to be a bigger test. But if you saw some of the more mature markets stabilizing at 5% or 6% share, wherever it is, and then having a hard time growing above that, I think it really truncates the upside—the right-tail upside—to the stock.
You could probably experience some significant multiple compression because you may be able to articulate upside to 3 million units, which would be 7.5% market share. But what if you can't articulate beyond that? Then you go from being a growth stock to slamming into a wall, and you have a real problem. You'd have serious multiple compression. You'd still be a very profitable business there, but you would have a real problem from a stock perspective. I think that's number 1.
The second wildcard, I guess—which I don't worry too much about—is autonomous vehicles.
Yeah, yeah, it was on my list. What happens with Chinese OEMs coming in? What happens with autonomous vehicles and robotaxis and all of that?
My personal view on that is that America, especially, is obsessed with private ownership of transportation assets, for the vast majority of Americans, anyway.
You mentioned electric vehicles. Obviously, right now I think the used-car stock is 3% to 4% electric vehicles in America, so almost all of it is ICE vehicles, but it is increasing toward electric vehicles.
Are they well-tooled? If we went to 100% electric vehicles sold in America tomorrow—which would mean that, in 5 years, 80% of used cars are electric—are they well-tooled to refurbish and sell electric vehicles, or would they have a bit of a stranded-asset problem if ICE vehicles started getting phased out?
Well, the Model 3, I think, was the most popular model for them last year—the Tesla Model 3. They've already been doing a lot of pretty high volume in the Tesla complex.
I think the reconditioning tends to be cheaper. They avoid the cars that need a battery replacement anytime soon, so they won't do a battery replacement, which is the most expensive part of reconditioning an EV. If you take out battery replacements, the reconditioning costs per unit are lower. So it's actually cheaper for them.
Would it strand capital somewhere in their infrastructure? Maybe, but I think the savings per unit would be net beneficial to them.
I actually have 2 questions, but one of them will be quick. You mentioned 40 million used cars sold in America. They've got the equipment for 3 million. You mentioned Atlanta, and it just struck me—I know Atlanta decently well. You drive an hour and a half outside of Atlanta, and you get into some pretty rural places.
Can Carvana profitably serve those pretty rural places where you have to drive, or is it really the urban and hard-suburban places where they can serve? A lot of car ownership is in pretty rural places, and I could imagine you saying, “Hey, sending a driver an hour and a half out, an hour and a half back, actually costs a lot more in our logistics costs than this.” That would kind of cap the market. I don't know if that makes sense.
I think when you're dealing with a GPU base of $7,300, if you pay an extra $200 to make a delivery, you still make money.
Cool. No, I just want to make sure. Okay, last question: valuation, real quickly.
I'm just using your Hindenburg short rebuttal, which, again, will be included. I qualified that and said this is just illustrative, but—
Completely fine.
So, completely fine. We can forget that. For their 2024 guidance, they had originally guided to EBITDA of $1 billion to $1.2 billion. They say Q3—they say “significantly above,” which, for the mathematically minded, I would have liked a number—but say $1.4 billion, say $1.5 billion, whatever you want to say, right?
The EV here is $55 billion. If I was using the theoretical in your Hindenburg short report response, which is 2027 or 2028—I can't remember the exact year—I think you had them doing—where are my numbers? You have them doing $12.50 per share in free cash flow, right? The stock is $250. So we're talking about a 20 multiple on 3- to 4-years-out free cash flow, while still growing quite quickly.
We discussed how they still have a lot of room to grow and everything, but I wanted to quickly talk about how you look at valuation because those are very high headline numbers. One of the things short reports have consistently said is they've got all this other stuff, but they've always led with, “Hey, the valuation looks really high here.” They said that when the stock was $30; they said that when the stock was $200.
If the stock was $20,000, I don't think you'd be involved anymore. I do just want to quickly talk about valuation—how do you look at fair value here?
Yeah, I mean, it's a growthy kind of company. The way I think about it is, I'll go back to my cell-tower analogy. The cell-tower business historically has had 7% to 8% returns at 1.1 tenants and 25%-plus returns at 4 tenants.
You could look at this as being at 400,000-plus or minus units on an infrastructure base that can be expanded to 3 million-plus units with an extra $1 billion of capital, and that process has already begun.
So maybe we’re at roughly 15% utilization of the infrastructure—the assets. When you’re at 1 million units, you’re still only at 33%, right? To me, one of the most interesting things I look at in businesses is where the break-even point is for the business. If it’s a cruise line or cruise ship, where’s the break-even point? If it’s an airplane, where’s the break-even point? That tells you something about the quality of the business, its scalability longer term, and the embedded or natural returns in that asset.
I think the multiple will always be high as long as they’re massively under-earning. Right now, the way I think about it is, let’s say they have 3 million units of capacity. Why should they stop there if they get there? Because if they get there, their advantages keep growing. Their value proposition will be even better.
They’ll probably have 1- or 2-day shipping. They can overpay for your car if you want to sell one, and they can undersell me a car if they want to convert me. They can play with all these different knobs in their whole vertically integrated workflow to deliver value to me and close whatever units they want to close. Growth is more of a choice for Carvana than it is for any other business I’ve looked at.
Right now, we have $7,300, plus or minus, of gross profit per unit. They have variable costs of about $2,300 to $2,400 a unit. So call it roughly $5,000 of incremental EBITDA per unit. Last quarter, as I normalize it, I think it was around $3,700 a unit. As they grow, that $5,000 might get better through what they call fundamental gains. They say there’s more to get, but they’re going to start giving those gains back to customers to accelerate growth. Maybe they’ll stabilize at plus or minus $5,000 per unit.
Let’s just run out the model and say, okay: 3 million units, $5,000 a unit—that’s $15 billion of EBITDA against your $55 billion EV today. That looks pretty cheap, right? But the question is, how long does it take to get there? Once you get there, to my prior point, what happens to growth? Does it slam into a wall?
If you’re at 7.5% market share and all your advantages just got stronger going from here to there, where should you stop? Where should that natural endpoint be? Should it be 12% market share, 15% market share? Are they going to sprinkle in omnichannel locations where they go from a vending machine and then a dozen locations in most markets to 10 of them, so you never have to go more than 10 miles? Now you’re picking up from Carvana instead of them delivering because, why not, if it’s that close? They can go a lot of different directions with this.
What I know is that this management team executes more aggressively and better than almost anyone that I’ve found. I think they’re young, hungry, and have ambitions far beyond 3 million units. That’s just what we can see today. I’m sure they’re going to greenfield some new IRCs well before they hit that 3 million to keep pushing that target higher and higher and higher.
So it’s kind of a question of, okay, you could use the cell tower analogy and say they have 1 tenant today, or less than 1 tenant today, but this asset and this business have all the embedded advantages such that it should get full tenancy. You could say this is like a triple-A office tower with just the bottom 50 floors, and you’ve got 5 floors filled today.
If I believe you—and I think I do—that they’ve got all these logistics advantages and everything, I don’t think they ever get to 100% market share, right? Because there’s always going to be a dad selling to a kid or family member or something. But why isn’t every used-car dealer in the country obsolete, and these guys are basically the ones every commercial used-car transaction is running through? Why isn’t market share, in the endgame, if I just took this to its natural conclusion, 50%, 60%, 70%? Amazon doesn’t have 100% of e-commerce, you know.
Well, e-commerce is 100,000 different things, and some you want delivered, some you might want fresh. This is 40 million units, pretty standardized.
I don’t know. Maybe I sound like a drugged-up bull here, but I’m honestly wondering: if it’s as good as you’re saying—and sometimes it’s helpful to take things to the most extreme—why isn’t every used-car dealer, and maybe in the long run every new-car dealer, dead? Why isn’t Carvana every commercial car transaction that’s happening with a consumer?
Yeah, I mean, if they offer the best terms to you and they can get you the car, let’s zoom way into the future. Let’s say they can do same-day delivery in 60% of the country, next-day in 20%, and then 2-day within the last 20% for the markets that they serve. And they can pay you a price that you think is great, or better than smaller-scale competitors out there, they should.
There’s what should happen, and then there’s what does happen, and what does happen rarely matches what should happen, even if you zoom way out. As I’ve seen a thousand times, people may want to go test-drive. People may not trust the quality of the reconditioning. People may not like the terms that are floated in front of them because Carvana doesn’t necessarily win on every dimension. They can win on a lot of different dimensions.
People may over-index to wanting to touch and feel the car, and they don’t want to go through the whole return process or pay for shipping, even though they can return it; they lose the shipping fee. Maybe they just want to shop locally. Maybe Carvana’s not floating you the very best offer because they know that you’ll really care about the convenience factor and not having to negotiate a knife fight for 4 hours over a used car with a used-car dealer. So they’re going to slightly tweak the price higher, maybe to your indifference point, such that you’ll choose to go to a local dealer instead.
There are a lot of knobs that can be turned, and they can test preferences and elasticity of demand across a lot of different dimensions because they create so much customer surplus across tangible and intangible categories. I think a lot of it is optimization, and through that optimization, they can test and learn what resonates most with consumers and what generates the highest free cash flow per share for the whole company. What generates the highest free cash flow per share may not address 100% of the market in the best way.
Perfect. It all makes total sense. I know I’ve looked at industries before where the management team will talk about capping out at 15% market share. You’ll talk to them and be like, “I don’t understand. It seems like you should be able to grow to 30% or 40% if these national advantages are true,” and they’ll list X, Y, and Z reasons, which are real reasons.
Sometimes it’s, “An organization just can’t support getting bigger.” Sometimes it’s, “There are 3 big buyers here, and they won’t let us grow any bigger. They’ll go to another player even if it’s a little inefficient.” I was just wondering why this capped out there.
The peer-to-peer market is an untapped opportunity for everyone in the used-auto business. If they make the transaction so easy and economically neutral to the buyer and seller in a peer-to-peer transaction, that’s flow that could come their way that currently doesn’t hit the retail market.
Will they ever address the full 40 million? That’s a big chunk. That’s like 40% of the 40 million. I think it’s like 15 million to 16 million units that are in peer-to-peer. That’s a big chunk of the market that they don’t naturally touch today, but maybe they get it indirectly just through the scale of Carvana’s normal retail business. So we’ll have to see.
There are also the fleets. They’re starting to address that with the Hertz thing that they’re ramping up, the rental-car fleets. I think they’re getting their hands into the whole value chain of a used-auto transaction, from the wholesale business to all the stuff with the retail customers.
Are rental-car fleets a potential competitor?
You know, rental-car companies already sell their inventory, and also at auction. They’re big auction customers. They’re all pretty poorly run, to be honest, but when you think, hey, they can do the refurbishment, right? They’ve got locations all across the country. You think about Enterprise—they’ve already got to drive out to the auction.
I think Hertz is just the one thing. At this point, Carvana's been a hot stock for 6 years. It's been a hot stock for 4 of the past 6 years. If they were going to do it, they probably would have already done it, but I was just trying to spin something up.
Yeah, I don't think they want to get into that core business of selling retail like that and going direct to the consumer—to your driveway, basically. But never say never. It's very interesting that Hertz is ramping up its activity with Carvana and getting premiums on its old inventory when it clears it out. The old way would be to sell whatever you can at the specific retail locations and then clear the rest at auction. They're finding accretion from working with Carvana, which can sell the inventory for them, and they can work out the profit share such that it makes sense for both.
Carvana gets the retail, the finance GPU, and other services attached, which is maybe $4,000 a unit, plus or minus, or $3,500. It's very light on capital intensity, too. That gets you operating leverage and very light capital intensity.
Yeah. So, from an EBITDA-per-unit basis, it's attractive, and I think they said it was neutral to EBITDA per unit. So, they're indifferent to who owns the inventory. This is kind of like the beginning of third-party inventory going on Amazon, right? It started out as 1P; now you're introducing marketplace dynamics, so there's a lot that can be done.
And to your point, if new, emerging EV OEMs come into the picture, wouldn't it be easier to attack the market and get national distribution using someone like Carvana, which has full reach—next day, same day, whatever it is—instead of opening up 100 dealers, 1,000 dealerships across the country?
I mean, you could turn on national demand tomorrow if you worked with Carvana. It's a very powerful platform for anyone new coming into the country or into the market, if it's Chinese OEMs. I'm thinking of Chinese OEMs.
But Aaron, this has been really interesting. I learned a ton. I've been wanting to do one on Carvana for a while. I'm sure we're going to hear from some of our friends who are bulls that we weren't bullish enough on this podcast, and some of our friends who are bears that we were way too bullish on the podcast. But this has been great. I learned a ton. We're going to have to have you back on for either talking dark fiber or cruise ships at some point, but I really appreciate it, and we will chat soon.
Thanks so much, Andrew. Really appreciate it.
A quick disclaimer. Nothing on this podcast should be considered investment advice. Guests or the hosts may have positions in any of the stocks mentioned during this podcast. Please do your own work and consult a financial advisor. Thanks.