Thank you everybody for coming here. It's so great to see so many familiar faces but also some new faces. And this year is actually a very special year. Believe it or not, it is the 10th anniversary of the annual gathering.
So I wanted to start with a question. Who was there for the first one 10 years ago? Stand up if you were here 10 years ago. I know there's a few of you. Come on Andrea, you can stand up. Wow. Well, a special round of applause there.
And that's not the only thing we have to celebrate. I met one of the participants just a few moments ago over coffee and he proposed to his then girlfriend, now fiance yesterday and I'm happy to say the result was a positive one.
So that's fantastic news. So as you can see I have my guest here Ernest Garcia, the founder of Carvana. We're going to speak very shortly but let me just quickly start off with the schedule for this morning and then we'll get straight to it. So I'm going to speak with Ernie for the next hour or so. We'll probably aim for about roughly 45 minutes of mountainside chat. I should emphasize these things often get called fireside chats and as you can definitely see there's no fireside here but there's definitely a mountain. I'll see if I can, I don't know if this is going to work but it's a very snowy day. Do you get that? Yeah. Well, so this is definitely a bonafide mountainside chat. So that's going to go and then after about 45 minutes I'm going to open it up for questions for you guys. Also, I think there's a few people watching on YouTube, so we may get a few questions there. Then, we've got half an hour coffee break where we can all connect with each other and then we're going to come back at 10:30 where Andreas, my partner will join me on the stage and we'll take our annual Q&A, which is a highlight for both of us in the year. And then 12:00 we're done with the formal side of things. And, you know, we're in one of I think one of the most beautiful spots in the world here. It's a fantastic day. Lots of snow. So, we're going to head up the mountain. There's different hiking groups. The hiking groups are being coordinated, I think, by Andreas and Joel. And there's also a skiing group. Well, there's a skiing group and we have our resident powder hound here, Nate Chzley, who's going to be coordinating that. So, thank you, Nate. So, we do have a room reserved for us at the troops, which is the first stop if you go up with the cable car. But please don't feel like everyone should go there at the same time. We're a very large group. I think about 250 people. So the restaurant will be totally overwhelmed if we all go once. So some people probably want to get away immediately after we finish. Some people maybe want to hang and talk a little bit. So that's totally okay. So hopefully we don't all arrive at the same time. And ditto. I think with the skiing and also the hiking, you know, probably unrealistic we can all go together. So, feel free to break up into groups and, yeah, whatever makes most sense for you. And then at 6:00, we're going to be back here. All these chairs will be taken away. Instead, there'll be cocktail tables and a buffet. And many of you have been here before. You know the drill. In the evening, we all hang out with each other and there's not a lot of opportunity. Everyone speaks to everybody. A lot of great connections happen and for me that's one of my favorite parts of the weekend. Then tomorrow is the emerging manager day. The backstory there was, you know, when I started my fund 20 years ago, I wrote a little memo after about 5 years explaining how I did things and a result of that I sort of got a lot of young men and women reach out to me asking about how I started my fund and looking for advice for themselves and what I realized was rather than try to speak to everyone individually it was much more valuable for them to bring them all together here and they'd not only get to hear from me and speak to me but also more importantly probably connect with each other. So that's the idea of the day tomorrow. But I'll talk about the schedule at the beginning of the session tomorrow. So Ernie, welcome.
Awesome. That you made it here and the place I wanted to start with our conversation is, tell us a bit about your background. You know, where you came from, a little bit about your youth, and maybe some formative experiences.
My youth. I got in trouble a lot. Okay. I'll try to give a quick version. I grew up in Phoenix, Arizona. Whoa, there we go. I grew up in Phoenix, Arizona. I think sports were a big part of me growing up. That's kind of what I thought I was going to do until I realized I wasn't that good. Then I went to school. I didn't consider myself like a smart kid, so I got into Stanford, which blew my mind. And you definitely consider myself like a jock. And I did okay on the test, and then I got into Stanford. I thought that was amazing. And went there. That was an incredible experience that I'm extremely grateful for. I think I met a ton of people and it gave me visibility to a lot of really ambitious people that were going to do ambitious things later on which I think was influential and helpful. I went and worked in New York for a couple years. Mike McKver who's sitting out here in the audience somewhere runs capital markets for us and we worked together. And I loved that. That was in 2006, 2007. Where I worked was a place called Greenwich Capital that at the time I believe was the second or third largest kind of processor of mortgages into the securitization market that was out there. And so it was a really cool place to sit and watch 2007 and 2008 happen. Right around that time I actually had some family reasons to move back to Phoenix. My dad had started a company before. My dad was always pretty entrepreneurial and I think my story inside of my own head was that I was going to do my own thing and I would never work for my dad because I kind of knew that when I met someone who worked, you know, for their family, at least there was a little bit of an asterisk in my mind about how much they had really achieved themselves. But in Phoenix, there's not a lot to do and I had to be back there. And so I had to swallow that pride and work for his company which is called Drive Time, which is basically kind of a subprime automotive retailing finance company. I'd kind of grown up around that business, you know, working in it from the time I was 16. But there just weren't a lot of really interesting companies in Phoenix. So, I started working there and started another company, you know, in an effort to achieve my self-identity. That was totally separated. It was a company called Ludarang. And it was kind of like a Yelp meets Groupon type of concept at the time. And we made it absolutely nowhere. I ate through all the money that I had earned so far and was down to zero. And then launched Carvana inside of Drive Time and so far that one is working but we've done our best to make it look like it wasn't a couple times.
Let me stop you there because we're definitely going to get to Carvana but there's legendary stories about your sort of competitiveness and ambition. Where did that come from?
I think the root of that is probably insecurity but I don't know. Like I said, I grew up playing sports and I was definitely like a kid who cried every time I lost. Like to an embarrassing extent and for an embarrassingly long time, including high school. If we lost in football, I would sit in the back of the bus and everyone knew they weren't supposed to talk to me because I would want to fight them. But I don't know. I don't know where that comes from. That was just there.
Well, we have a common friend, Cliff Sosen, and he told me a story once how he challenged you to a sort of a chin-up competition, and he started off with 15 and then he said you started and by about the seventh, you were sort of really struggling and he's like, 'Okay, I've definitely won this,' but somehow you managed to make it to 15.
I'm not going to lose to Cliff, no matter what, not going to happen.
So, not everyone here will be as familiar with Carvana. So why don't you tell us what Carvana does and the problem they solve for people?
Sure. So I'll try to be brief on this one too, but I'll probably struggle. I think buying a car, especially in the US, is definitely viewed as a tough experience. And I think the root of the difficulty of that experience is that I think the way that industry has evolved is it's an inherently very complex transaction. You're buying a car that is variable in its price and its features. You're probably trading in a car. That's a complicated thing to get right. You probably need financing, you probably want a warranty. There's regulatory components. And so the way that industry has evolved over time is basically the OEMs hired a bunch of franchise dealers. These franchise dealers kind of outsourced many parts of the technology stack to 30 or 40 different third-party software providers. The OEMs require that the dealers make certain investments in their facilities both the real estate and facilities themselves. And then because they're all using the same third-party software providers, they all have to hire the same people. And so that evolution to handle the complexity where it got broken into I think many different horizontal slices led to highly undifferentiated underlying cost structures. And so when you kind of take highly undifferentiated businesses they either have to have, of which there are many, there are tens of thousands of dealers in the US, you would expect for them to either have very low profits because they're all selling a somewhat commoditized product with the same underlying cost structure or they need to find some way to take that transaction and take the components of the transaction that a consumer may not fully understand and find ways to monetize that. And I think dealers have gotten very good over a very long period of time having a customer come in and figure out sort of like are you elastic to price? Are you elastic to rate? Are you elastic to trade in value? Are you willing to buy a bunch of different products? And they're good at putting that basket together in a way that takes four hours and makes the customer feel like they don't really understand what's going on, but also leads to profits that are pretty good in retail despite how competitive it is. And so in our mind, that was the root of that problem was very deep. There wasn't a way to make the customer experience better unless we actually fixed the underlying cost structure and built a different kind of experience. And we also thought that the experience hadn't evolved, hadn't really adopted modern technology and adapted to modern customer preferences because it was so hard and it was so kind of ingrained in all of these different systems. And so we thought if we, you know, build a more automated process that's fully deterministic and that leverages all of these cool online tools and reduces real estate expenses but then gives customers a 7-day return policy instead of a test drive that like the math of all of that should work. And I think we didn't appreciate just how hard it would be. But I do think the premise was right. And I think as a result, we've been lucky enough to be really successful. And I think by most measures, we're now enjoying outside success. And I think we're doing it in a way where there's not a lot of other businesses that can easily compete with us because we've been investing in it now for 12 or 13 years super aggressively to try to build that system that we didn't appreciate just how hard it would be.
Yeah, I definitely want to dive deeper into some of those challenges you had to solve to sell a car online. I want to ask first though about the origin story which is I guess a typical question to ask in this sort of thing but I'd like to dive deeper and like when was the exact moment you came up with the idea for Carvana? Was it something you always in the back of your mind wanted to do? I guess you grew up in a car family. Maybe it was something at the time when the web was getting bigger. So maybe it was something always in the back of your mind or was there sort of like an apple hitting you on the head sort of moment.
I suppose so. I think the origin was a desire to do something that I thought was going to be like an interesting business that we thought we could find success at and that kind of led to Ludarang first and then...
Exactly when would this have been though?
When would that have been? That would have been 2010 give or take, 2009. So after the first startup it failed or around...
No so sorry Ludarang was around 2009 and basically we didn't even get all the way through Ludarang when we started Carvana we were like halfway through Ludarang because at that time you know again like that combination of like insecurity and arrogance you think you can do everything at once and so Carvana was like another idea that was popping into my mind at that point in time because I felt like I didn't understand why automotive retail was such a disliked experience and it didn't seem fundamental to the product and I was you know working in Drive Time at that I was basically, you know, going to Drive Time at whatever it was, like 8 in the morning, and then I was leaving at like 4:00 or 5, and then I was working on Ludarang until 2 a.m. I didn't have kids at that time, so I could just kind of work all the time. But was starting to realize that Ludarang wasn't working. And I think a big part of why it wasn't working is because we didn't know anything about that business. And so, you know, I think when you're starting a business, you sort of want to not know too much because then you think that all the ways things work are the ways they have to work. But if you know nothing, you spend all kinds of time trying to figure out things that are obvious to other people. And it took us a long time to figure out obvious things.
There's a sweet spot of not knowing too much, but not...
I think there is. And so anyway, so we realized that, you know, Ludarang wasn't working great. Carvana was something that was another idea that had popped into, you know, my head. And we wanted to try something that we sort of knew something about. The CEO of Drive Time at the time, because I was constantly throwing big ideas out to him, he asked me to go to an auction, which is the way that kind of dealers buy cars in the US, and he's like, 'Will you please find a way to save $1 million?' At the time, Drive Time was selling, you know, 50,000 cars per year. And he's like, 'I just want you to find a way to save $1 million. I don't want a bigger idea than that. Just like save a million dollars with high confidence and then let's implement that.' And so I went to the auction and I won't bore you with all the details, but watching the way the auction functioned made me like much much more confident that selling cars online should work. And so I came back and I put together...
Because it was so dysfunctional or what made you think that?
Because the dealers that were buying cars were sitting there and kind of maybe opening the door and feeling the paint line and then walking away and then they were paying $20,000 for a car they were going to sell to a consumer. And the fact that it took them all of 13 seconds to get the information necessary to make that kind of investment and then they were going to sell it to a consumer made me realize like so much of the information gathering that people value is not actually that valuable and you can also massively mitigate the need for it if you recondition cars. And so I went to...
Super interesting because I always assume the sort of the idea came from more the sort of the customer-facing buying a car side of things but it sounds like it came more from the idea of oh there's a lot of inefficiency in just the way a car is sold. Sorry, border option. Yeah.
Yeah. Well, I think it was both. So, it was those two ideas coming together. It was why is buying a car such a bummer and then that was that whole thought process of how the industry had evolved and how you could fix it, but you had to build something fundamentally different. And then I think it was going to auction and seeing thousands of cars trade in this like very simple way very quickly. And then also looking at that, it's like kind of fun from a markets perspective and saying there's $20,000 cars that are trading with basically $500 of fees between the wholesaler and the purchaser of the car with probably $500 to $700 of additional expenses just to get the car to this place. So you've got like a 5% transaction cost for something that people are looking at for 2 seconds and buying. And you're like that would never work in any other market. And so the whole system just felt really really inefficient. And so then we put together a deck or I guess I put together a deck that was like 20 pages long called online auto sales and showed it to the CEO and asked for $50 million and he cussed me out. But he did ultimately give us the money.
Yeah. I mean just jumping ahead a little bit here but it sounds like right from the beginning the idea of data and the intransparency and efficiency was like core of the idea and that was you know 15 years before AI came along. But I mean that must be an incredible tailwind for the business. I think...
This is jumping way ahead. It just seems to fit here. Yeah. No. Yeah. Well, I think I'll try to put that in a slightly different but like if you think about like what is the purpose of physical retail versus online retail as a general matter and I think as a general matter it's about information collection. It's like the reason that you want to physically see something. You see it, you touch it, you smell it, you can like, you know, see how heavy it is in your hand, whatever it is the thing that you're buying. But that's like the actual value you get from going and you know physically going to a store versus e-commerce which in general would be more cost effective. And so I think in auto it was interesting to just think about what is the actual information that consumers getting. And I think consumers tend to value the idea of a test drive and it feels like a responsible part of your diligence that you should do when you buy a car. But I bet you if we surveyed everyone in here if we like did a blind drive test of 10 cars they couldn't pick them out. And so a lot of that was perceived value, not real value. And so we kind of believe that if we could find other ways to replace that information, whether it's 360 car spinner or, you know, condition reports or a 7-day return policy or whatever it is, other ways to convey that information that would give people confidence, you could do it way cheaper than the traditional way.
Yeah. So a few times you've said we and I know you have an incredibly loyal and core team of people around you. You know, who are they and when did they sort of start entering the picture here?
So, I think I could expand that we pretty wide, but I'll say the first couple were Ben Houston, who's our COO. He's someone I went to college with. And I think maybe it's interesting because he grew up in a house, this is like more origin story stuff, but I think it's interesting to think about how people are formed. He grew up in a house. He had three brothers. His parents both work two jobs and his parents are great and they're like fun to be around and they very clearly love their kids a ton and they also were never around because they were working all the time and so they basically kind of like raised themselves and I think it's super interesting to look at that because he's probably the most effective human being I feel like I've ever met in my life as far as he's like really smart and effective in a boardroom. He can think through any problem. He could, you know, he'll kick your ass in a bar fight. And he's you know you met at university anything in between we met at college and then he went on to be a lawyer because that was you know his internal story was okay I'm going to go to law school and I'm going to be a lawyer that's how to be effective. And then I called him and actually asked him to be the CEO of Ludarang and he came over and did that and then transitioned over to Carvana afterwards but he's an incredibly effective person and Carvana would not be successful without him and he and I have very different approaches to the way we think about the world which has been helpful in finding the best answers because we are both very comfortable arguing as well. Ryan Keaton was one of our other co-founders. Met him actually through his wife that worked at a third-party marketing company and we were kicking off the idea for Carvana and she was like you should meet my husband and now he runs brand for us and has been around since the very beginning. Mark Jenkins came over relatively early. Mark Jenkins, CFO...
He was an academic. He got his PhD at Stanford. I met him there when he was doing some research and was using Drive Time data to do that research and then he went and was a professor at Wharton. But his brain power and throughput is like completely unmatched. And so I basically begged him for years and years. I don't even care what he did but I begged him for years and years like please come over and do something with us. So I think he's also like pretty non-traditional. You know, we've had lots of short reports over the years and some of them have pointed to him as being a non-traditional CFO. Like that's a problem. Which I...
I'm a non-traditional fund manager. So that's...
Yeah. No, I don't. The idea that you're supposed to have some like very traditional background I think is a misplaced thought. But anyway, so he's someone else that was around from the very beginning and there's many more.
Yeah. Is it fair to say a lot of that core team cut their teeth on the first startup which didn't work out but it kind of bonded you and like...
I think Ben and I did and then Ryan and Mark were both kind of, you know, well actually no Mark was there too. Ryan was new to Carvana but yeah so yes I think that is fair.
Yeah. So you know I guess some people might say hey what's the big deal selling a car online? Anyone can put together a website and put a bunch of cars up there. But there's a lot of hard problems you have to solve and I'll let you of course respond to that, but I don't know if Claudine is in the audience here but she sent me an email a couple of years ago. I think at the time you were having like problems with the DMV registrations and she sort of asked me Rob what's up with this DMV stuff all these cars Carvana are saying they're not registered what's happening there and I don't know what ever awakened the belief in her that I understood DMV processing but I appreciate the trust my investors have in me but anyway tell us what it takes to sell a car like what are the different steps?
There's so much. I mean I'll try to make it concise but buying a car correctly is hard. Cars are highly variable products with, you know, hundreds of features that describe those cars and then there's a bunch of those that are hard to measure, like condition from afar. So, I think buying the car is hard. Then you've got the whole reconditioning process. We put about $1,000 of parts and labor in every car. And so, the car goes through an inspection process. We figure out what needs to be fixed. That includes mechanical things. That includes cosmetic things. You got to build a process to do all that. And it's not like a normal manufacturing process because it's not every widget doesn't need the same exact set of steps. They all need different steps. So, it's kind of like a pseudo manufacturing process with wells where you pull cars out and do individual things. That's a complicated process to try to build. You have to build a transactional website because, you know, consumers are starting from a place of I need to test drive a car. Like, I buy a car once every 5 years. I need to make sure I get this right. They don't want to make a mistake. So, you need to build a website that makes them feel confident, and makes them feel like you're institutional enough to where they can trust the 7-day return policy, because a 7-day return policy is just a promise from a website. They have to believe that that's real. You have to build out financing, that means credit scoring, pricing, loan structuring, figuring out like the combination of down payment, monthly payment that makes sense for that customer. You do have to do title registration. Title registration ends up in theory overlapping a lot with finance because a lot of what title registration is is you're basically just trying to put a database entry in a state database that says this VIN belongs to this identity. But unfortunately you know many of these state databases are very tired...
Or even manual in some cases.
Many times they're manual. Many of the document requirements are very very burdensome. You know in certain counties they'll require if you want to put your spouse on the title you need to have like your marriage license. No one even knows what a marriage license is but you have to have a physical copy of that. So it can get hard because you have to have all of these rules in a system and you have to pull down all that data from a customer and then submit it to the DMV who accepts it or rejects it and you have to handle all that and then you have to have last mile delivery to the customer's door which is basically just a logistics network except for this logistics network is moving multiple thousand pound items...
But you can't use UPS to transport.
And you can't use third parties because they don't, third parties traditionally there is like a third-party shipping network in the US but it's dealer-facing not consumer-facing. So they give you like a we'll show up sometime between 7 and 14 days. They don't give you like a precise window which isn't consumer friendly.
Yeah. I mean it's, I mean the titling is just a tiny part of the transaction but it's like if that doesn't work the whole transaction breaks down and there's tons of things like that, right?
Yeah. There's a lot to it. That's where I think it's like I do think when you're tackling these problems sometimes it's better if you don't understand how hard they are. Like you need to be right that it's ultimately solvable, but I think if you appreciate all the things that you have to do, you would be less likely to try.
Yeah. So, just for Claudine's benefit, is the titling solved now or is it a never-ending story?
I would say like just like fun facts, there's one state that actually puts out public data about how often like title work is submitted on time across all dealers, and that's about 70%. So that means like a 30% error rate. Our error rate is far less than 10% of that error rate and it was approximately equivalent to that error rate at the time that we were getting blown up in the media. So we were never that bad at it. We were approximately as bad as all other dealers. We are now way way better than other dealers at it. But it is a hard process because it can be county level.
Yeah. I mean for, I mean as a business person there's kind of like there's also like a sweet spot you want but you want problems to be hard because otherwise everyone can do them but like not so hard that you can't get it done. Which are the problems which you think are just basically going to be, so you know all the different parts of the sort of the value chain of getting a car sold which are the parts which is like it's solved it's easy and which are the parts where it's basically there's always going to be room to improve it?
I think the more a process looks like something that just goes into a machine the easier it is to get to a place where it is largely solved and scalable, the more the process requires organizing people or moving physical things, the harder the process fundamentally is and the more work you inherently have to do. Because even a process where you're moving things and organizing people in theory as you continue to automate more and more of the decisions you make it look more and more like a machine but it's sort of like inherently further from a machine. So I think things like pricing cars super scalable like very efficient today. I think things like reconditioning we have a ton of work to do for a long time to keep making that more scalable and more efficient.
Yeah. So and which part would you say is the most difficult today? I mean, there's so many different parts of the business with, you know, shipping the cars in, reconditioning them, doing the financing, underwriting the customers, getting them delivered, all that kind of stuff. Which is the stuff where you think there's the most sort of opportunity to improve?
I think there's opportunity to improve everywhere. I think in terms of difficulty like the things that make scaling the hardest I think one is reconditioning and I think inside of reconditioning it's mostly cosmetic work more so than mechanical work. I then think it's last mile logistics then long leg logistics and then basically the sum of we call customer care which is verifications for finance, registration, answering customer questions things like that.
Yeah. Do you think any other company will ever be able to replicate this? I mean, I guess now you've shown it can be done there would be a temptation for people to try and copy it. On the other hand, it's like it was...
Even you didn't almost didn't make it. Yeah. I mean, will any company ever do it? Yeah. I mean, like to me, it's that there's like this funny thing where if any group of people can do it, then another group of people can do it. So I do think that what you want is you want there to be the hardest set like sequence of things possible behind us if they want to make sure they understand how hard it is. So they can like look back at our history and they can say that looks pretty hard. But also like as we find more success they look at our success and they say I want that and so that becomes like a bigger draw to take on the hard thing. So I think we just got to keep running fast. I don't think you ever build anything where if you like stop moving no one's going to catch up to you. So I think you just always got to run fast.
Yeah. I wrote a letter a few years ago talking about how you know we as investors love to find companies with competitive advantage and that's a great thing but the problem is like the whole point of having a sort of like competitive advantage. You get to sit back and people can't touch you because you have this wonderful moat around your business. So you know in a way you don't want the moat to be too wide because you want people to still be trying to improve the business. Right.
Yeah. Or you have to have that insecurity like I think it's funny because I think especially like investors everyone in this room is probably a pretty rational person so much of what actually makes the world move forward is like emotional things that you can't quite articulate and so I think that like even that framing it's completely right but it's highly rational it's like okay you know we kind of did it we're sitting in front but I think if you just have like an emotional need to go tackle problems and an emotional need to win an emotional need to keep running past then like that will power you despite the fact that rationally maybe you're in a good spot.
Yeah. So should we talk about the crisis here?
There were several crisis years. You just got to see one of them.
