Rivian Automotive, Inc. (RIVN) Earnings Call Transcript & Summary

Okay. I don't think you need much of an introduction, but we're very pleased to have Founder and CEO of Rivian, with us today. RJ, thanks for coming.

Yes.

You said you've been -- as we were sitting down, you're like you've been crazy busy. What you've been crazy busy doing and what message do you have for us today to kick things off?

Well, it's just an incredibly dynamic environment from both the demand side and watching. There's -- just the sentiment around electric vehicles and the tone on electric vehicles, in particular, I think as they become more politicized as part of this election cycle, has certainly been there. But for us specifically, we've been -- we just went through a major relaunch of our vehicle with what we call Gen 2 so mass upgrade to our network architecture, significant cost down across the vehicle. We changed out well over half the suppliers and the parts of the vehicle. So while the car looks very similar from the outside, there's not a part that hasn't gone untouched, big consolidation, the number of components, ECUs, parts and suppliers in the vehicle. And so the -- with that, we've had a couple of supplier issues of recent that have been challenging. And in particular, a few issues around our in-house motors with some of the components that have been painful and a reminder of just how a multi-tiered supply chain can be difficult, a reminder of what we saw in 2022, in the beginning of 2023. So we've been working through some of those issues, which is impacting us in this quarter. But along with that, we also continue to see R1 be of the top-selling vehicles in its price categories. So over $70,000, we're currently the best-selling vehicle in a number of states, over $70,000, in the EV space, the R1 platform is -- continues to be a top seller. But the size of the market, over $70,000, is smaller than we would like and certainly with the macro that we're living in and experiencing, we do see a need for our R2, for a product that's priced under $50,000, something in that $40,000 to $50,000 range. So boy, do I wish R2 was ready to be selling to customers today. But there is a tremendous level of excitement. And I think -- I've said this before, there's a lot of latent demand for EVs that's hard to see, but customers that want a form factor that's not available, are looking for a brand and brand story that's different than what they have today and want something that's clearly sub $50,000. And there's just not a lot of great choices. Of course, the Tesla products are outstanding. Model Y continues to dominate. But there's -- outside of that, there's not a lot of really highly compelling products and so we see the R2 as being significantly different than Model Y intentionally, not to say Model Y is not a great vehicle, it's just answering a different customer use case than what an R2 does. But the similarity of price points will create, we think, another viable choice for a lot of customers.

Okay. I want to double click on demand and definitely some of the things you're doing with the R1 relaunch, if you will, it really is.

Yes.

I'm not aware of any product that has undergone such an extensive reengineering, I don't think you are either, you're doing it. But just the issues with the supply and the motor, is this something that kind of threatens the goal of gross margin positive by year end you think?

Well, we're working really hard to solve it. And trying to resolve that, we remain very confident in our gross margin positive by Q4 and ultimately, we need to be able to achieve the volume to absorb the fixed cost of our plant. But we remain hyper focused on that. These supply chain issues are frustrating especially when they're a level or 2 removed. And in R1, we've gotten to a point of stability, and then we inserted a lot of new suppliers into the system, which, of course, have the benefit of saving us a lot of cost. But also have the challenge of bringing up a new supply base.

Okay. So supply is a bigger concern right now than demand? Is it fair to say? In the very short term -- real time.

Like today, certainly. I think in the long term, it's actually -- in the short term, certainly; in the medium term, all of you are looking at us through the lens of R1. So you're thinking about Rivian, our ability to be profitable through our flagship products. But really, in the long term, our story, us as a company is going to sit really entirely with the success of R2 and R3, and not too different from, of course, Tesla's success with Model 3, Model Y. We don't really judge Tesla anymore on the Model S or Model X. Their flagship product still exists. They're important for Tesla as a brand, but they certainly have very little bearing on the financials. And we are working very hard to have a very similar outcome where R2 is the dominant vehicle within our portfolio from a volume point of view. And so as we think about what R1 represents for demand for 2025, it -- again, it links very heavily to macroeconomic conditions that we're going to see. And with a vehicle that has an ASP of close to $90,000, there are limits to how many people buy that -- buy vehicles at that price point relative to something that's half the price with our R2 or R3.

All right. So you mentioned EVs, demand seems to be getting politicized. So how is demand then? Is it -- how much of it is politicization? How much of it is just wealthy families that wanted an EV, got one. I mean we see them all over Westchester. You see them in this part of the world, they're beautiful, and customers love them. But finding that next 1% is a big -- is tough. How much of it is just that kind of the asymptote of that price point? Or is there something else?

I mean, just to be clear, EV demand is still growing. It's growing at a much smaller rate than it was in '22 or '23. So the growth rate has slowed, but it's still a growth story. And so I think the question that the industry is challenged with today is how fast we'll get to the end state, which is 100% of new vehicle sales are electric. And I would say unequivocally and without any hesitation that, that is the end state that hasn't changed. And so what we're witnessing is companies that are -- were planning to transition sooner are now dramatically pulling back. I think that's bad for the world, but I think it's actually -- on a comparative basis, it's actually good for Rivian and that there will be a relatively small number of true pure-play focused companies, of course, Tesla being the existing leader but ourselves being there as well. And so I'm surprised at how quickly large and common OEMs have pulled back, and I'm surprised at how much emphasis has gone into ICE and hybrid. Of course, that's optimizing for profitability in the short and medium term. But I do think it is a challenge. And it's a challenge for the bigger OEMs because they have a large denominator. They have to capture, they have to continue to sell similar volumes what they've sold in the past. So I think we're going to see continued presence of EVs, non-EVs, both in hybrid and pure ICE form. But by 10 years from now, I think we'll see very significant penetration.

Okay. So China is not a market that you participate in right now. However, many in the industry argue that China is now the benchmark really to use in terms of BOM cost and design and architecture, industrialization. Even some argue even more so than Tesla, that Tesla is now looking at China and seeing how they're doing things and look to incorporate. So as your team -- tell us what your team has done to help benchmark some of these things. I don't know if you maybe bought -- have you bought an SU7? Some call it the SU7, are you're looking to taking it apart and seeing how they design some things. And even though they're different customers and different ethos of the brand, what are you learning?

Yes, it's a great question. Yes, we're seeing in China, there's an enormous level of competition amongst a lot of different brands and both new companies, existing companies in the EV space. And of course, BYD, chief among them. We've seen some pretty impressive -- I mean, I should say, very impressive vehicles from a cost point of view and a price point of view. And so lots of people have taken apart those cars and looked at what's in them. And in many cases, there's nothing -- there's not some singular magic wand. There's not -- you don't take it apart and say that this is wildly different than you expected. But it's a cost advantage that exists across every nut, in every bolt, in every wire and every panel in the vehicle. And supplemented with the companies who are competing in an environment where they're operating at 0 gross margin and planning to do that for a very, very long time. And so -- or negative margin and supported in some cases, through other means. So you have the landscape that's just very aggressive competition. And with that as context, we have made the decision to not enter China for a lot of reasons beyond just cost, but also the geopolitical, of being a U.S. company trying to enter China. I think that window is open for Tesla. They've executed that. It's a huge part of their cost structure today, but it's not something we're going to pursue. But it is an important question to say, will those Chinese products be products we compete with here in the U.S. or for that matter in Europe. And the tariff structures that have been put in place today is 100% tariff on vehicles, full vehicles from China. There's roughly 30% tariff on batteries from China and then a much smaller tariff on components. And what we've seen happen over the last 18 months and really, I'd say even more over the last, call it, 2, 3 quarters, is a lot of Chinese entities are recognizing that the way to access the U.S. market and, to a lesser degree, the European markets is not necessarily with full vehicles but through component relationships, where the tariff structures are lower, where the geopolitical challenges that exist at the respective top of house for both countries are not really felt. And as I described, when you take apart and look at the cost structure on some of those Chinese products, what's helping them is that it's not necessarily the part design, there's efficiencies of design, no doubt. But it's that every single component through their tiered supply base is 20% to 30%, sometimes 40% cheaper than what we would have for a part or a component that's sourced in a western market. And so really designing and engineering our supply chains to identify where we see opportunities to leverage the much lower cost basis that exist, that's been built out in China, and that's on mechanical parts. That's on -- often, we try to put it to parts that are relatively fungible to move. We think they're not things where we build a lot of technical dependency but that often carries pretty significant costs. And so we've really been super intentional on designing our supply chain with R2 very differently than how we approached R1.

All right. Let's talk about R2. I want to talk about R2. I want to move into like into tech and in the Volkswagen relationship as well, and we'll open it up. R2, it seems to involve -- you're talking about $40,000, $50,000, kind of sub-$50,000 price point for that to be profitable, you're going to have to get your BOM cost down, volume adjusted, like 60%, 70%. I mean -- how do you do that?

Yes. This is -- there's a few things here. So the R1 BOM, what we launched with the bill of materials on the vehicle that we first put into the market, and that was a bill of materials that we largely negotiated in 2018 and 2019, just to wind everybody's clocks back. That was like a moment of peak auto. The auto industry was thriving. Everything was going really well.

Good old days.

It was the good old days. So it was a hard time to negotiate with suppliers. To go to a supplier, imagine in 2019 before anyone seen or driven a Rivian vehicle, before any of the success we've had on the customers' side had happened. This past year, U.S. News & World rated us Top Brand, J.D. Power put us the highest level of brand appeal 2 years in a row, the highest level of customer rated satisfaction from customers, not from us. So all that wasn't there. And you go in and talk to a supplier and say, "We're going to electrify an SUV and a truck." And they say, "You're crazy." Say, "We really need parts." And say, "Okay, we'll sell them to you." But you can imagine there's a significant risk premium. And that risk-adjusted bill of materials that we had to launch with, and it took us -- we anticipated that we'd be able to drive that down very quickly post launch. We'd say the product is successful. We're doing well, to stay a supplier to us, you'll need to bring your BOM costs down. What we didn't anticipate was essentially, the supply chain crisis that happened with COVID, and we went from an already inflated BOM, where we had to pay a lot to get suppliers to work with us and these contracts were negotiated in 2018, 2019, to then having to beg those very same suppliers to only not lower the price, but to not charge us extra just to give us parts in that really complex environment. So we had -- you could argue a perfect storm that made it very hard to take cost out of our bill of materials. And so that ultimately led to what Adam referenced before, and I referenced, which is what we call this Gen 2 vehicle, where we said, "Look, either the cost comes down on these parts to what's reasonable or we're going to move suppliers." And we had to move a vast majority of our supply base to new suppliers that were willing to be part of our long-term story. And were working with us at price levels that correspond to the risk profile that we have today, which is to say, much lower. And with all that, we also made changes to design, and we've talked about this now a few times, we're able to bring the bill of materials down quite a bit. We spent time on this at our Investor Day. But that's at R1. As that's been happening over the course of last year, we've been negotiating pricing for R2. And what's been remarkable is the R2 BOM is -- the R2 bill of materials is dramatically lower than even the lower BOM in R1. And it's because we're starting these negotiations from a clean sheet. We're developing components and systems from a clean sheet. And so the combination of the design of the vehicles benefits from being a true clean-sheet design relative to our first vehicle. The features and the content of the vehicle have been adjusted and of course, the size of the vehicle has been adjusted for the price point it's operating in. And then the leverage that we have, as I just described is night and day. And I've described this before, but lots of ways to try to communicate just how different this is. But on the R1 initial sourcing for -- like to get a supplier to work with us, it would have usually involved some version of me flying to a building in Southeast Michigan, somewhere into the Detroit area, waiting in a lobby to start meeting 15 minutes late with like a senior manager of sales at a supplier. Those very, very same companies today will fly their CEO to Normal, Illinois to meet with us to try to be a supplier. So it's just like the leverage we had is so different. And it's surreal for me because it's like man, just a few years ago. we were like begging to get a meeting to meet some person 6 levels down from the CEO to sell us a few parts and now the CEO of that company, who was also the CEO then, is flying out to the middle of Illinois, to come to our plant to pitch us on being part of R2. So that translates to a very different cost structure in R2 relative to what we saw in R1. Now, all that being said, of course, the announcement of our deal with Volkswagen is I think everyone in the room, at least, maybe heard of this, but just a high-level point, it's a $5 billion deal, whereby Volkswagen as a group is adopting our electrical architecture and our software stack across their portfolio of brands and products. That statement of confidence and statement of endorsement has led to a further sourcing tailwind for us as we complete a lot of negotiations on our bill of materials for R2.

So if I were to adjust -- if I look at your stock at whatever it is, $12 or -- today. And if I were to NPV, the cash you're going to get from Volkswagen, adjusting for the share count changes, your enterprise value is roughly $6 billion, maybe $7 billion, okay. So the stock market is looking at your business and making some -- solving for, wow, this is going to be really tough. And you have many, many people, including some of the investors in this room who are asking the question, is Rivian better off being a kind of a stood-up Tier 1 supplier for advanced architectures and technologies and software, software-led, true software-defined vehicle and designs, which -- were proven, we're going to talk more about it, or -- versus being an OEM? What do you say to them? Why is it important for you to be an OEM because like they're asking questions about capital, the capital commitments now. And then when you throw in the AI commitments, whether it's on-prem or not on-prem, this is -- what do you say?

Well, I think the first thing to recognize is if we look at the end state, if we look at -- if we can debate when 10 years or a few years from now, if you're in a [ car ] state, it has to be -- 10 years from now, everything will be electric. And just to reiterate that point I just said, quickly. There's a surprising lack of appreciation for how much and how fast and how significant the change to electrification will be over the next decade in California and the states that have also adopted the California Resource Board standards, from 2035 onwards, you will not be able to buy a vehicle with an engine in it. It's like that -- 2035 sounds like it's forever away. That's 10 years away. So it's a product cycle and a half. So the scale of this change is going to be significant. The United States, certainly from a policy point of view, at the federal level is not as aggressive as Europe and certainly not as aggressive as China. But policy is going to be driving us to that. And we think consumer demand will start to shift to that at a large scale as well. And so we believe there's going to be the emergence of a few brands that build up dominant market positions in this new world. And if we can translate even a fraction of the market share dominance that we have at a high price point. We're at -- over $70,000, we're a very, very strong market share player. If we can translate that into the sub-$50,000 price range, which is where you see Model 3, Model Y, by the way, the vast majority of vehicle purchases, the average sale price of the new car in the United States is just under $50,000. We believe we'll be in an outstanding position. And certainly, we also believe that the market doesn't yet value that for a lot of the reasons we talked about. But one of the hardest things to create and one of the things that if I remember back or wind the clock back 5 years, the thing that I used to stress about and not be able to really, frankly, articulately answer how we'd solve it, is how do you go from nothing as a brand to something that people care about deeply. The idea -- like brands are a funny thing. It's -- they're ethereal. There's no -- it's like you can feel it, you can sense it, but it's hard to be quantitative about it. Why does Nike mean so much to people when I say that? Why does Apple mean so much? So the idea of creating a brand is it's a bit of magic in a bottle. And miraculously, we can spend a lot of time talking about how it happened, but we have built a brand, and we built the brand that is highly valued by customers. And I'm saying that as obviously a biased party in this. I believe it's a really cool brand. But importantly, customers are saying that. So we have the #1 rated brand in the auto industry right now in terms of repeat purchase and in terms of first-time appeal. And that's 2023, 2024 as per J.D. Power. Consumer Reports did a similar study, #1 rated brand in the premium segment. U.S. News & World Report did a similar study, #1 rated brand. So we've created a brand that has real value, now that's not to say that the business of selling technology isn't valuable as well. So what we've done with Volkswagen is an example of that. But I would say it'd be an incredibly sad miss to take this wonderful brand that currently has higher levels of satisfaction and engagement than some of the best automotive brands in the world, Porsches, your Teslas, and to not go leverage that with a product that has -- a product reach that's much broader.

RJ, it sounds like you really want to make cars.

Yes, yes.

Okay. So why Georgia then? A, do you have enough money to make that your plan A? And why is it -- when you think about various demand scenarios, whether there's demand for 100, even 1,000 Rivians or double that in the next few years, I imagine that's part of the calculus as to why you would need just physically more capacity. Or is it that just normal, just -- it's too normal of a car plant, and you need to clean the sheet. Because there's a lot -- again, there's people in the audience listening who are thinking like, why Georgia, does that really -- they're not confident that's part of the way -- questioning it.

Yes. So we had an adjustment to our strategy we talked about earlier this year. So originally, we were going to launch the R2 platform, which underpins, as you can see now, R2 plus other variants, R3 being one of them. You can imagine there's other things that will come on that platform. But we said that platform was going to launch out of Georgia. And we announced the change to the plan to launch that out of Normal, out of our existing production facility in Illinois. And there's a few reasons for that. The biggest immediate reason, which is certainly on the minds of a lot of investors, and I'd assume on the minds of all of you is capital efficiency. So it allows us to remove more than $2 billion of capital spending through the launch of R2. But the other reason is it removes -- potentially removes volume risk around any single vehicle. So now in Normal, we'll have R1, R2 and our Commercial Vans, which we really haven't talked about, but our Commercial Vans are produced under the same facility. And it gives us a level of fungibility to say if the demand for things that cost $80,000, $90,000 is low, that's fine, we can sell more R2s. And if we are selling a lot of high-margin $95,000 R1s, we can make those as well. But it removes long term, any risk around the overall macro conditions that exist for a flagship product, and it sequences importantly, when we turn on Georgia. So Georgia is going to be built across 2 phases, ultimately getting to a nominal capacity, meaning, I call it baseline capacity without really pushing over time, of 400,000 units with 2 chunks of 200,000 units each. And so that's a decision we can take as we see the market evolving, but it allows us to launch R2, get to positive margins, positive cash flows of business with our Normal facility. And then derisk sort of managed or risk weighted in terms of how fast...

It seems like the R2, just SOP and ramp in Normal is -- opens up, but does it -- changes the probabilities of...

Yes, totally. And to your point, we're seeing -- and the amount of excitement we're seeing for R2 is really high. It's humbling. It's like people, especially right now, we get so much inbound from customers that are saying, "Hey, can I get an R2 next summer?" And I'm like, "I wish we could get it to you by middle of '25."

I want to move to your autonomous tech, which you've been increasingly moving in-house. And you've explained how it's either farm everything out to someone to do everything or do it all yourself, in between thing, which most OEMs are in some form is not optimal and you've decided to more vertically integrate.

Yes.

And in addition to just kind of emphasizing why that was so important to you, I guess, specifically, we've discussed this before but for this audience, there's many in the AV community and the robotics community that say that large language models and this latest AI, turbocharging of machine learning is transforming the way the -- this embodied AI work, learns. Do you agree and how does it change your capital? The amount of capital that you allocate towards compute model training or other aspects that you might need to just stay in the game?

Yes. Really quickly, just to provide background on how we've approached this. Our Gen 1 vehicles, we had a compilation of systems that came together, including a sourced front-facing camera that provided both processed information that would tell us object -- detect objects, classify the objects and provide the vectors for those objects. The world of using multiple systems coming together with a late fusion of information is the wrong architectural approach. And we deeply believe to do this right, you need to control the entirety of the perception stack and you need to use information as early as possible. So it allows and enables an end-to-end approach in terms of how you're training the models, going from cameras all the way through to the controls. And to your point with what's happened in LLMs, the world has changed so dramatically. It's hard to fully appreciate how we approach self-driving today, how different that is from the way we looked at it as an industry 4 or 5 years ago, where a true end-to-end approach with a heavy emphasis on training, not in the vehicle, off-line from the vehicle training, how different this looks. And so architecture, what's now in Gen 2, we have more megapixels than any other vehicle on the road, we have 55 megapixels, incredible [ light ] performance. We have a new compute stack that's 10x more powerful onboard from what we had in the past. And of course, the features we're now seeing with the Gen 2 vehicle are like the very beginning of the potential for this platform, where previous systems, we have late fusion, you end up [ asymptoting ] to capability basically, it's going to get a little better than what it launched with, but it can't learn. And so then the question which you alluded to rightly, is how do we get all that training? So where are we going to get all these H100s from? And...

Do you need them?

Well, it grows with the size of the fleet. It grows with -- [ so we need ] a lot, and often, it's -- I think, for some reason, we have this conception that the H100s care who owns them, they don't. They run the same way. So whether we build it ourselves and we own our own data center or whether we're renting them from somebody who also owns a data center, it really doesn't matter, provided we can move information freely between that location. And so we're -- we've not announced any of this in terms of our details, but we, of course, are building a very robust way to access a lot of training that allows us to, in a low-cost way, make sure that all the things that our vehicles are seeing, discrete events, heart events, the safety cases we're building around those events are used to train the fleet and the rate of progress here is it's like night and day to what we saw 2 years ago. So it's hard to fully appreciate, I think we'll see in 5 years...

I feel like to tease the audience for a conversation. Another conversation we should have on this ever-growing and changing surface area between AI and your business and think how Volkswagen can play a part of that, not just as a capital provider because anyone could provide capital. But to help kind of integrate you more and get you closer to those photons and get them closer to those photons and then how they learn...

Well, this isn't a statement about Rivian, it's just an interesting statement for all of us to think about if developing a robust automotive foundation model takes, call it, $3 billion to $5 billion of training to get to a really solid place, do we think there's going to be 25 different companies building those foundation models? The answer is probably not. So then the question is then who are the people that are going to build it and what are the partnerships and relationships going to look like? And...

And that's cool...

So there's a lot, I think, that we need to think about as an industry. So that we're not spending $200 billion training 25 models. Sort of like LLMs, the hyperscalers built the big models and there's lots of people that leverage those.

Well, thanks for your perspectives as always in your insights. RJ, thank you. Good luck.