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

Thanks, everyone, for joining today. I'm Toni Sacconaghi, Bernstein's IT hardware and electric vehicles analyst. And I'm really pleased to have RJ Scaringe here to -- CEO of Rivian. We were just discussing that we have mutual Italian heritage, and both of our family members pronounce our last names in different ways. And I said, "Oh, how does your grandmother pronounce it? Scaringe." And he said, "Yes. Exactly." So for those Italians in the office, we have RJ Scaringe here today. But in all seriousness, RJ has been obsessed with cars since he was a kid. He had a doctorate in mechanical engineering from MIT Sloan Automotive Lab, and he founded Rivian in 2019. As most of you know, Rivian IPO-ed in November of last year and was among last year's most anticipated initial public offering. So we're super excited to have RJ here, and we're going to plunge right in. There are question cards. We have tried to get people to use this pigeonhole technology. It generally hasn't worked. So we're passing around question cards. If you do have a question, [ Priya or Tamis ] on my team will collect your cards, and we'll try and get them in.

So without further ado, RJ, thanks again, and we'll just plunge right in. Oh, by the way, shameless plug. My idea, not RJ's. Two great Rivian, both the pickup truck and the SUV, are in front of the hotel. If you didn't notice it this morning, go check them out. Are they going to be there through the end of the day?

They're going to be there through, yes, through the day.

Okay. So go check them out there. Impressive vehicles. So RJ, just to start, like auto -- the auto market is kind of a 90 million unit market. How do you think of automobile market growth over the next like 10 or 20 years? Like let's really take the long view.

Yes. It's an important question, I think. One of the things to zoom out on and recognize as much as we think of electrification as being sort of fully underway and for us that are in the space and thinking about it, it feels like it's happened. There's around 1.5 billion cars on the planet in the global car park, less than 1% of those are electric. As you said, of the 90 million that are produced every year, less than 5% of those are electric. So we, as an industry, are so very much at the beginning of this massive transformation and the scale of what has to change to build 1.5 billion electric vehicles over the next 2 decades. We can debate whether it's 2 decades or 3 decades or 1.5 decades, but certainly over the next, call it, 10, 20 years is extraordinary. So battery cell supply chain, essentially, the restacking of who are the OEMs that are building these vehicles, what consolidations are going to happen, which companies grow, which companies shrink, which companies disappear, which new companies emerge, what the supply chain looks like, all of that ultimately contributes to, I think, a very different end market than what it's looked like for the last 100 years. And within that, of course, there is a question of how many vehicles we'll be producing every year. And in some ways, it's hard to predict. It does depend in a way on the amount of new mobility models that start to emerge in different modalities of achieving personal transportation, but we tend to believe it's going to be roughly the same, call it, 90 million to 100 million, at least for the foreseeable future. But most importantly is it's going to go from a few percent electric to 100% electric. This is really hard to imagine, but this decade, we think that's going to be very rapid in terms of that shift.

And RJ, there are 5 or 6 kind of credible new entrants in the automotive space, which is really historically unprecedented. Incredible, meaning public markets are ascribing market capitalizations kind of above $20 billion. And so you are having potentially kind of a flat market going forward. And your market, obviously, to the degree it's in EV market, it's probably going to grow 20% a year for 15 or 20 years. So you're in a different space. But if I ask you to zoom out and just think about the market ahead, you have a market that's essentially flat that's -- that will become increasingly more competitive in terms of the number of new players and, I think, more global. What happens? Is the industry more fragmented 10 years from now? Do we expect around the consolidation? And how do you see that playing out?

Yes. I mean I sort of alluded to it. But I think fundamentally, maybe paying less attention to the number of car companies that will exist and more attention to just the fact that some companies will grow, I think more will likely shrink, and I think some will actually disappear. And that's at the OEM level. That's the part that we see most visibly. And that may seem like there's a lot of change coming. I think we'll see even more change in the supply chain. And we'll see the way the supply chain has been set up between Tier 1s to Tier 2s to Tier 3s with semiconductor suppliers playing across the Tier 2, Tier 3, Tier 1 all the way up to the OEM direct. We just think there's a complete reshuffling of what that value chain looks like. And a lot of that comes down to where value is created in the automotive space. And historically, things like electronics and software have been outsourced. So very few OEMs develop those platforms themselves. They rely on Tier 1s. Companies like Bosch or Continental to do a lot of the electronics and the software stack. Our view is that the strongest players in the space at the OEM level absolutely need to control the software and the electronics and absolutely need to create highly differentiated user experiences around that vertical integrated set of technologies. And so that's a very big shift for the OEMs in terms of the skill set and talent that exists within those organizations. Fortunately, for us, we designed the company very specifically around that belief that's software and electronics really become a really driving factor for product differentiation in the long term.

Yes. I'm debating whether we should go there right now or -- so when you see kind of software as a platform and a source of differentiation, is it in terms of the attractiveness of the offering in the consumer's eyes? Or is it a differentiable perspective from the economics and profit pools from the OEM side? Or is it both?

I think it's definitely both. And the consumer side is maybe what you can see in terms of the product and the attributes, it's easier for us to understand. So in a world where an OEM like ourselves controls the software stack, it controls the electronics and all the sensors placed around the vehicle, you can create features very dynamically. So you can add features over time. You can grow self-driving capability. You can create new user interfaces. You can create new branded experiences. You can have new performance modes. We just launched a Sand Mode in our vehicles just last week through an over-the-air update. And these are all really exciting. They become almost addicting as a customer to say, well, I can't wait for the next set of updates, and what's going to come? And that allows us as a manufacturer to be really dynamic in being really responsive to say what features do customers want, what problems do we see in the field, what opportunities do we see? That's 1/2 of it. The other half, which is of equal importance is by controlling the stack in both hardware and software, you can really start to look at the life cycle of the vehicle very differently. And so in the consumer world, it's a bit different than the commercial vehicle world, but we'll talk about both. In the consumer world, it means you can start to think about things like residual value or risk very differently, which, in turn, positions an OEM to start to participate in the used vehicle marketplace in a profoundly different way than how historically OEMs have participated, which is just they really haven't outside of certified preowned vehicles. And what makes that interesting is the used vehicle marketplace is one of the largest profit pools that exist in the auto industry. And the reason it's so profitable in short is there's very high degrees of asymmetric knowledge. So the buyer and the seller rarely know the same thing. And because of that you, we've seen the industry sort of emerge or evolve to have lots of layers of margin stack that occur between dealers, which then sell the vehicles to auction houses, which then sell them back to dealers, which then sell them to ultimately customer again. So all of that is essentially playing on different levels of understanding of the health of the asset. The manufacture of a vehicle with a very robust diagnostics platform, where the vehicle has actually been designed to have constant assessments of its health, meaning their sensors, everything from accelerometers to microphones, to temperature across the vehicle to create really informed views of the state of health of the vehicle. It positions the manufacturer, positions us to play a much more active role through the vehicle life cycle. So used car transactions, you can remove information asymmetry. Insurance or risk assessment, you can remove information symmetry. And so I think we're at the very beginning as an industry is starting to scratch the surface of just very deeply changing how we look at some of the more painful parts of the automotive experience, used car being probably among the most painful, both as a seller and the buyer. But that requires, call it, some imagination to see what the future state will look like. More directly, we see the ability for advanced diagnostics, predictive maintenance, predictive reliability, health assessment, immediately having applications in the commercial space because you can drive uptime. And you can actually convert those skills, those algorithmic skills immediately into products. So we have, on one side of our business the products you see downstairs those are launches, those are flagship products for the consumer brand. We also have a commercial side of the business. And commercial side of the business, sort of the lazy way to describe it, it's really more accurately, I'd describe it as a fleet business or is a centrally managed deterministically operated fleet business, where, yes, we sell a vehicle, but more importantly, every vehicle is sold with a software subscription, whereby we do a whole bunch of things. We make decisions on when the vehicle should charge, how it should charge. We match routes to vehicles based upon vehicle health, we match drivers of vehicles based upon vehicle and driver skill and vehicle health, where we can make thousands and thousands of decisions through informed algorithms to holistically best utilize the fleet. And that's worth a lot of money that adds up to a healthy subscription on the asset. And so we think fleet management in the commercial space are, as I said in the essentially controlled deterministically operated space, is enormously enabled by controlling the software stack.

Yes. So you said a lot there, and I'd like to dive into some of what you did. So maybe we can start on the consumer side. And you talked about sort of incremental software offerings. One of the things that I struggle with is that if I look at the history of automotive, price increases have actually lagged producer price indices and consumer CPI. So I think consumer CPI has gone up like 2.2%. Car prices have gone up like 1.8%. And yet there's been incredible technology and innovation in cars over the last 30 years, particularly from a safety perspective.

Correct.

And invariably -- and that's partly why I started with the conversation around the industry and its competitiveness, which is it's always been a supercompetitive business, which is why returns on a cost of capital basis are relatively low versus other businesses. It's always been a supercompetitive business. And innovation gets -- there is some price taking initially. So if you think about airbags, I think Mercedes first came out with them, it's like $1,200 for your airbag, $1,200 for the passenger airbag. And then ultimately, over time, it became a standard feature within the car. Same thing with backup cameras. Now we're seeing with blind spot detection, et cetera, et cetera. So I guess the question is, why does that change? So why does incremental technology value accrue to OEMs going forward as opposed to consumers given the backdrop of a supercompetitive industry that's arguably getting more competitive? And so, a, why does that sort of conceptually change? And then b, if you were outside of the commercial fleet, which I'd like to talk about separately and outside of used cars, how do we think about -- what you think the big chunks of offerings are that people will actually pay for in a way that's meaningfully accretive to OEM's economics?

Yes. So jump first on the consumer side. I do think we have to be careful. It's very easy to imagine and postulate and sort of model in a spreadsheet recurring revenues for software features or subscriptions on the vehicle. It's an easy spreadsheet to build. It's hard to get -- and to your point...

That's my profession. I do that kind of stuff.

Yes, yes. But our customers are willing to pay for it. And I think as a -- from a competitive landscape point of view, customers really love these updates. But at the moment you start to charge for them, it does change that dynamic. And you have to be very careful in our intention on what you charge for and how you charge for. So with that said, I do believe what we're going to see, and back to your question on consolidation, what I think will happen over the next 10 to 15 years is you're going to see margin unlocked not just through features. Features is sort of the obvious one to look at. But you're going to see it unlock through business models that are fundamentally different than what exists today where we structurally remove margin stacking or margin layers that exist. So direct sales is a really obvious example. So with the exception of us and Tesla, vehicle manufacturers actually don't sell their cars to customers. They sell their cars to a third party. They pay a third party effectively to sell their cars for them. And this is the world of franchise dealers. And there's lots of things you could say about that being problematic. You also can say there's lots of things that are strong about that, that dealers act in some way as a bank as a financing source for the OEMs. But at the end of the day, aside from the fact that somewhere between 5% and 10% margin is paid to someone else to sell the vehicle, it makes it very hard to participate in entirely new ways of interacting with a customer. So different forms of ownership, whether that's fractional, different forms of leasing and lower commitment consumption, which may move from a 2-year lease or a 3-year lease to a 6-month lease or a seasonal lease. Those are all doable, but you have the friction of a dealer involved. And I think what we are going to see is companies like us or Tesla with the benefit of a clean sheet, not have to start with those relationships and need to be unwound and reconfigured. And I think OEMs -- this isn't a choice. They will have to redesign those relationships and be creative on essentially designing ways to access different pools of capital, different sort of margin stacks that may exist by creating value for different ways to consume. So that's one. I think the other thing that we're going to see that creates margin opportunity in terms of restocking of the business is vertical integration in areas that have historically been handed out to the tiered supply base. So electronics and software, I've talked about already. I truly believe that this is something that the world's leading vehicle manufacturers will need to control, not just from an attribute point of view, not just from an over-the-update point of view, but also from the perspective of margin. And it's not necessarily a direct analysis to say the margin is -- you're saving 15% in buying this computer, this ECU from building yourself versus having someone else. It's the fact that you can more easily consolidate a large number of ECUs to a smaller number of ECUs. So you get both the margin removal and you get the simplification of the architecture, which, today, vehicle architectures -- every vehicle on the road has highly abstracted architectures. There's a body control module. There's a chassis control module. There's a powertrain control module. There's an HVAC control module. There's a seat control module. And it's because we've abstracted into all these different little computers effectively as a way to facilitate sourcing from a broad spectrum of different suppliers. Technically, that's not the right answer. The right answer is we want a control module that's in the middle of the vehicle. We want to control module that's in the rear of the vehicle. We want to control module that's in the front of the vehicle. And they do lots of things. They do controlling of the front trunk. They're controlling front suspension. So we think that the consolidation of ECUs into a much more simplified architectures, which results in not hundreds, but thousands of dollars in savings at a vehicle level, will unlock either margin or pricing. The OEM will have the choice. And then I'd say the last piece. So the first is different direct-to-consumer, the second is vertical integration in areas that today are abstracted across lots of tiered suppliers. The third place is actually through margin of creating products that are really interesting. And a lot of the -- a lot of what we've seen in the last 12 months in terms of growth in positive unit economics as a result of pricing that different companies have had the ability to apply to their products. And I think that really comes back to the strength of the product being created, and in some ways, links back to this idea of software because you can try to charge for software as a subscription where you can use software as a way to make the products and the experience so interesting that you can drive more margin to the asset sale. And it's not clear today how that will evolve. But I think it's going to depend somewhat on the market dynamics of how the different players are competing and different folks will apply different approaches here. Some will variabilize it. So you pay for it over time. Some we'll say you'll have a more expensive or less expensive thing upfront. But I do think the role of software is going to unlock a lot of margin potential there.

Right. So you just -- in the latter point that you made, the superior consumer experience will manifest itself in something that people are willing to pay for.

Exactly.

Now just specifically around autonomous driving. How do you think of -- they're probably 30 people, 30 independent companies doing autonomous driving trials in California, others working on autonomy. How does autonomous driving ultimately evolve? And how does it evolve into a way that is materially accretive? And because Tesla sort of set this level of about $12,000 for autonomous driving. Chinese OEMs are already at half that level. Why do we -- a, are we convinced that can ultimately be value accretive? Or is it is autonomous driving 20 years from now like cruise control where every car has it? And how do you see that evolving?

Yes. This is an area where there is so much confusion and so much -- I'd say the media has contributed to it. I'd say the companies themselves have contributed to it. But essentially, we've bucketed all different types of autonomy from Level 1 to Level 5 and to broadly the category of self-driving. And to maybe answer the question a little bit more specifically, I think a more accurate way to think about self-driving is there's really 2 categories of topology of hardware, sensor set topology. So there's the hardware platforms for which you have very few cost constraints, which we were able to put $100,000-plus worth of sensors on the vehicle. And you're building a digital driver for which the operator of the asset is willing to spend $100,000 to $200,000. And so that's the like of a Waymo or a Cruise. But that's a very, very different topology, and therefore, a very, very different outcome in terms of the capabilities than a hardware constrained system, whereby you have thousands of dollars to work with and build of materials. So that's cameras, that's radar, maybe a low fidelity of LiDAR that you have to decide which sensors that you're going to put into the vehicle. But ultimately, if you're charging $12,000 for it or $6,000 for it, you can't spend $100,000 in sensors. And the 2 are going to result in very different outcomes. So a Waymo is looking at a Level 4, Level 5 and hardware-constrained system is typically in like more the Level 2. And I think often, we confuse the 2, and it's like comparing apples and oranges. They're very different things. And so in the world of very sort of hardware-heavy systems, the number of market applications is a lot narrower. And I think the path to market solutions requires a different consumption model. So it's some form of a transportation as a service or a true shared mobility. It's not going to make its way into personally owned assets for a while until we take the hardware cost down considerably. And in contrast, and in some ways, beautifully in contrast, a hardware constrained system, you can, in theory, put this in every single car. And so while you don't have the benefit of all the sensor set and the richness of data coming off that really expensive sensor set and associated compute platform, you have the benefit of many, many, many millions of miles. And so we, of course, as a company, had a choice to say which of those 2 systems -- 2 approaches do we want to take. And surprisingly, we took the latter. We have a hardware constrained system. We're spending thousands of dollars per vehicle on sensors and compute. And what we've launched, we've launched a Level 2 system. So in our case, you can get on the highway, you -- the vehicle can then drive itself. And we have a -- we think of the right topology of sensors. We have 11 cameras, 5 radars, of course, ultrasonic sensors and high precision GPS that allow us to have enough headroom to grow the feature set over time. But ultimately, you reach a limit with that hardware. And as hardware improves, we have a next gen set of hardware that will come out that will grow further. But we believe there's a lot of value you can create for consumers with Level 2 or Level 3 systems, meaning you get into certain operating domains, a highway or freeway or parkway and the vehicle can drive itself. You can get your time back. But the step in complexity to go from vehicle doesn't have a steering wheel from my house to someone else's house is much higher. And ultimately, the decision of how quickly we see those and everything is going to be very much a safety and a risk decision. And I think we, as a society, are still figuring out how we want to balance that. Of course, we know there are accidents that are a result of the hardware that are happening in self-driving today. There are deaths that are occurring because of this. So we -- as we look at very high levels of autonomy, it's a question of how applicable these are in broad domains outside of specific cities or specific more taxi-like operating environments.

Great. Maybe we can shift gears a little bit. You've talked about a market share goal of 10%. When you say 10%, what is that 10% of -- I did some like really basic kind of market data segment. And there are 1 million SUVs sold globally with greater than $65,000 price tag. And there are 200,000 pickup trucks globally with an ASP greater than $40,000 as starting prices. And so when you say 10%, are we talking about those addressable markets? Or are you thinking about your gen 2 offerings and thinking more broadly about categories? How do we square that?

Yes. Well, it comes back, Toni, to your first question, which is the scale of what we're going to see over the next 10, 15 years is extraordinary. We're going to see a world -- my oldest is 6 years old. By the time he is my age, it will be a very exciting event to see a gasoline-powered car on the road. So it's like -- it'll sort of be like if you were to see a horse on I-87. So it's just going to be a very different world. And to get to that, there needs to be a lot of volume credit, and there's going to be a restack of who are the strongest and largest players in the space. And so when we talk about 10%, it's more emblematic of our broader goals as a company to become a large company, to be a significant player when it comes to market share. So we're talking about 10% of the overall electric vehicle market, which eventually will be 10% of the overall market. And the reason it's important to recognize that for us as a goal is it results in a different set of decisions around how we approach certain parts of our technology stack. So if the goal was to build 100,000 vehicles or 200,000 vehicles, it would be really hard to rationalize the level of technology development and the vertical integration that we're driving. So we're developing inverters, motors, power electronics, all in-house. All of our computers in the vehicle are developed in-house. The full software stack from the low-level OS, bare-metal all the way up to the application layer is in-house. We're building a charging network, a large charging network of 300-kilowatt DC chargers. So we're making a lot of very large forward investments to make sure that as we go beyond our halo products, our flagship products, the R1T and R1S, which as you said, are, by definition, positioned at a higher price point, and therefore, by definition, have more narrow addressable markets, but the purpose is to build brand. Following those as we move into what we call R2, which is a larger volume -- much larger volume vehicle priced in that $40,000 to $60,000 range, that sort of step comes with the tailwinds of this technology set and with the tailwinds of some of the infrastructure building, whether it's charging, whether it's service infrastructure. And of course, that then creates the pathways for future products as well. But we do really believe that some of this sort of core focus on technology and creating these structural cost advantages will create opportunities to grab significant market share.

And I want to address cost economics in a sec, but even just like practically in the near term, like your Normal facility, I think your capacity is 150,000. Is there enough addressable market at $70,000 plus vehicles given some of the numbers that I said that you probably actually have to have more than 10% of the market initially to build that facility? Is that realistic given you're still at the early stages of your brand building?

Yes. It's a -- I'm glad you asked it. It's a fun question. So it's really interesting for us. So when we started developing the products and the brand and thinking about it, you can imagine like on day 1, there's no product, no brand, no strategy, no technology, no money.

It's like you, a car guy thinking about what would be cool.

Just like me, right. And so we went through a lot of iterations on what we should do, why we should do it. And what we ultimately sort of arrived at in terms of the R1T and the R1S is this sort of sibling set of products. So we wanted to develop products that were going into a space that really redefined trade-offs that we had long become accustomed to. So the vehicles you'll see if you've seen outside, these are among the most capable vehicles in the world off-road. They're also incredibly refined on road. They're also incredibly quick. There's only -- there's very few Ferraris that could out accelerate our pickup or our SUV. And they're also very efficient. So it's just like this really unique combination of features and attributes. And one of the reasons we wanted to create such a unique set of features and attributes around segments that we felt were sort of embodying this idea of adventure and this idea of going places with your kids, with your friends, with your gear, with your pets, your stuff this platform for memories is we recognize that it would enable us to go into what we would think of as a white space. Meaning if you were to look at the cross-section of customers that we have and what they're coming out of, no single vehicle represents more than 2.5%. So whether it's F-150 or Tesla, these are -- there's a broad spectrum of different vehicles customers are coming out of, which is really good. It would be unfortunate if all of our customers are simply moving from one thing into Rivian. But because of that, the dynamics that you would think about normally in terms of market size and percent of market share are sort of hard -- it's hard math to do. So you're equally having people trade in a Porsche or Tesla or Ford or GM. So it's sort of everything. And we went into that knowing we have this -- or hoping I should say we'd have this broad-based demand. We now have -- we actually know because we've generated the orders, and as of our last earnings call, we've talked about this, our ASP is around -- just over $93,000, close to 100,000 preorders, and that's just for the U.S. And so the rate of acceleration of demand, coupled with the growth of our ASP really demonstrates that there's a market that's quite large for vehicles that have a lot of function but are also efficient, that are also exciting and new and we believe will take significantly more market share vehicles at this price point than 10%. Now I think with all that said, the challenge that we have as an industry going forward is a lot of the segment dynamics that we've historically looked at are just going to start to change. You're going to have customers that switch between different vehicles. And I think with that, it's often hard to predict what demand will look like. So what we've seen in the last 6 months has shown us that we're not concerned about demand. We're very much focused on ramp and supply chain and production. But as we go into our next set of products, as it moves down on price point, we do think it will be -- it's just a different segment. You're going to have more cross shopping. You're going to have more closely comparable products. So it's in those lower priced segments that the differentiation continues and we have to be really strong. And so we're spending a lot of time on R2 making sure that we create the same set of dynamics where we're pulling customers from lots of different brands, not one single brand, creating new electric vehicle customers, but really at its core, preserving as much price as we can.

And if we think about just on the Normal, Illinois facility, is that envisioned to be an R1 facility exclusively and then R2 will go in the new facility, is that fair?

Yes. So Normal was designed to be R1, our commercial bands. And today, it has an installed capacity of 150,000 units. We'll eventually grow it, we can make an expansion to our paint shop, and we'll grow that to 200,000 units between our flagship consumer products and our flagship commercial products. The commercial products, what we started with -- it's -- we wanted -- you think of what's a flagship in commercial. It really is defined by the customer. So we pursued Amazon as a partner. They're now a very large shareholder, but more important than their shareholder -- shareholding is the fact that we have a big offtake with them for 100,000 vehicles.

So if the -- I'm just -- I'm sort of doing this math real time, so maybe we've talked about it publicly, but I think your orders over the last 2-plus months have been about 10,000, it's probably a gross order number. I don't know if it's a net order number. So call it, 8,000 to 10,000. If I annualize that 50,000, and that's just U.S. But U.S. should be the majority of your market. And then maybe your commercial, I think you said would be 1/3, right? So how do I -- what happens between a 50,000 or 60,000 run rate plus some commercial business to get to 200,000 over time just sort of conceptually from a market development perspective?

We think R1 -- between R1T, R1S, on a global basis -- or by the way, the R1S does fit global markets. But we see that as a 110,000 unit year. We do anticipate that we'll continue to be constrained by production, meaning demand will exceed our ability to produce, which is good. And then the remainder of the capacity would be dedicated for a range of different commercial bands, the first set of which we're now ramping and delivering to Amazon.

Okay. Now maybe we could talk a little bit about cost because we've alluded to the R2. And I was talking with Jim Farley from Ford yesterday, and he was saying, look, like, the cost right now of our -- the Ford Mustang powertrain is like $25,000. And it's tough to make the math work with, a, be profitable; and b, be competitive with ICE offerings with that kind of powertrain costs. So maybe you could provide your perspective on what you think the relative cost of your powertrain is on the R1 Series relative to a nice powertrain like I sort of go close to $20,000 for the battery pack, maybe another $20,000 for the powertrain, commercial F-150 truck, total powertrains maybe $10,000, like is that sort of the way to think about it? And part of the reason I ask is, how do you then make the move to an R2 platform? And what changes to get you more competitive and get you at a $40,000 starting price -- $40,000 to $60,000 price for R2?

Yes. For sure. So I'd separate -- when you think about propulsion and separate the drive side of it, so the motor, the inverter, the gearbox from the battery system, and from the point of view of the drive side, the motor gearbox inverter, that should cost less than an engine. And I think we can also imagine all the components and pieces that go into an engine. It's a very complex to...

Only a 4-drive.

Well, we have 4. So we have an expensive -- so on R1, we have a very expensive -- yes. We have -- the R1 product has 2 motors per axle. It's really incredible performance, but over 800 horsepower. But that, by definition, is a flagship configuration. It's an overkill configuration. It's really fun. But the R2 will have a combination of either 2 or 3, depending on the package motors. And when you think about 2 motors delivering all-wheel drive performance, 0 to 60 in 3 seconds, that's a really cost-effective way to get to that level of performance relative to an ICE setup. Now in contrast, the fuel tank on a combustion-powered vehicle is very cheap. It's -- we're talking hundreds of dollars, including the fuel pump versus a battery, which depending on the size, is -- let's make a size independent, call it, $100 a kilowatt hour or something like that. So 100-kilowatt-hour pack is pretty expensive. We're talking about $10,000. So these are -- that's the biggest difference. And the challenge here is that the supply chain needs to really be thought about. So you need to understand where you're getting your raw materials from. The raw materials then, of course, have to be processed. The conversion of hard rock mined lithium into lithium carbonate or lithium hydroxide, that going into cathode-active material, that going into the cell, and of course, making modules and packs. So it's in this world, but we also believe much like on the driving net side where vertical integration, we think, creates cost advantages. By the way, everything I said about cost -- if you have to source a gearbox, an inverter, a motor, someone to assemble the 3 of those things together, you create a lot of cost layering that's painful. So for us, really foundationals owning every part of the drive unit down to the rotor and the stator in the motor. On the battery, the CapEx associated with vertical integrating everything to start is too high. So you have to pick your battles. So we've picked everything north of the cell to starts. We buy cylindrical cells, in our case, from Samsung for our launch products, and we do everything else. Over time, I do believe manufacturers will have to play a more active role in the upstream supply. Whether that's owning or not, will depend on the commodity, but definitely participating in very structured agreements that have floors and ceilings to pricing around offtakes. In some cases, we'll have co-investment. In some cases, we'll completely own outright to try to lower the cost intensity of storing energy, which in an ICE vehicle again is close to nothing.

Yes. And I mean just maybe I misinformed on that. But -- so let's say the nonbattery part of the powertrain is a push on ICE versus EV.

It's close. Engine is 1000 -- engine driving $1,000. For equivalent power levels driving it $1,000.

Right. So -- but on the battery side, like I think about battery costs today, maybe entering this year were like $125 per kilowatt hour, maybe they are closer to $150 now. So if you're talking about 100-kilowatt hour, you're still at $15,000, maybe we get to 100-kilowatt hours by $100 per kilowatt hour by R2. So you still have a $10,000 gap, it would seem. Am I thinking about that incorrectly? Or...

I think you're right. And I think that is -- the challenge is why this segment selection is so important to make sure the segment has the ability to support that. If you look at electric vehicles today relative to the cheapest electric vehicle we can buy is still $15,000 more than cheapest ICE vehicle you can buy in the U.S. market, different in other markets. So I think we're going to continue to see that. But over time, costs will come down. It is worth noting -- and you said something here, which is I do want to comment on. There are different chemistry topologies that create different economics. And to oversimplify it, you have high nickel-based chemistries, which are great from an energy density point of view, really good for like a premium product, but don't necessarily fit commercial vehicles or base models that, well, the iron-based chemistry. So lithium iron phosphate we are quite bullish on. We're launching our LFP pack later this year. It'll first go into our commercial vehicles, and our entire commercial business will leverage LFP. And we'll also have LFP as our base pack for entry-level models. And I do think that the advantages that come from LFP are pretty extreme. I mean I don't think -- this is just a statement of fact. The challenges, the supply chain around those is complex. 100% of the world's LFP cathode active material comes out of China, and the tariffs that exist, and of course, the geopolitical challenges that we know so well have to be considered in terms of long-term reliance on LFP for non-Chinese markets. And so it's something, I would say, if you're an OEM that's planning to do a lot of volume, you need a robust strategy in LFP that extends production outside of China.

And RJ, just thinking about sort of near-term economics, sort of getting to contribution, positive contribution, is it -- do you feel now that you've raised price on your models that there's a point at which you can see positive gross margins? And is that just volume at this point? Or is it principally mix? So even at volume, do you feel on an entry-based level model, you'll have positive gross margin or contribution dollars at volume? I realize volume is a huge issue when you're only doing a few thousand units -- less than a few thousand units a month. So how do we think about even with the raised prices, whether base entry prices, whether you can have positive contribution margins or you really need that kind of average mix up to kind of $93,000 to have realistic contribution margin at scale?

So for us, for our production facility in Normal, there's 3 big levers that will ultimately drive positive economics. The most powerful and most important is volume. So we have a plant that's designed for 150,000 units. Because of some of the supply constraints we've been dealing with, we're operating at a lower output rate than that. So if you think of it like this in terms of hours in a given week, you have 168 hours in the week. So if you work 24/7, that's how many hours you should be able to run the plant, of course, it's practical or possible. But at a minimum, you'd want to be running your plant 80 hours a week, maybe the maximum 130. Today, we're running our plants at about 30 hours a week, meaning we're not able to support a full shift because of some of the supply constraints we've had. Now we do believe the second half of the year looks a lot better. So we've talked about this in the last earnings call. But the supply chain improvement, particularly on semi and our ramp with our suppliers, is really strong. And because the plant is now running so well, we were burning through supply components really quickly in a good way. But the most powerful is to start to be able to run not only a full solid shift, but to go to 2 shifts and really utilize the CapEx, utilize the fixed costs associated with that. The second element, of course, is building materials. And we talked about this in the context of vertical integration. We also talked about it in the context of purchasing power. A lot of the challenges of a highly vertically integrated business when you start it is at the beginning, you have none of the benefits of vertical integration. You have all the headwinds of vertical integration at the start because you have all the fixed cost carry through. So that starts to flip pretty profoundly for us as we get into the second half of next year as we start to really see ramp. And then the last, of course, as you said, is price. And for us, fortunately, we've seen the dynamics around pricing have been more positive than we had modeled or anticipated when we designed the plant and design the product and building materials. We certainly love it, but we don't know how long that will persist. Today, we're in a world where it almost seems like unlimitless in terms of willingness to pay. Used drives are selling for twice the price of new Rivian. So this -- we don't believe this will forever be the case, but there is a world today where pricing is very, very strong.

Maybe you should drive them around the block and then put them up for sale and [indiscernible] that could really help ASPs. But -- so if you were -- you have a $93,000 ASP on recent orders. Do the marginal economics work at a starting price of $67,000 or $72,000 on your offerings at scale when you plow through those? Or are you -- is there still some expectation post the price relief that, look, we're going to naturally mix up, and that's sort of how we price to get to a 25% gross margin -- target gross margin?

Yes, a big enabler for us. I talked about LFP. Here's our LFP pack on driving positive unit economics to the base model. We've seen others do this. Of course, Tesla has done this well. It is a material reduction in cost, which if you're willing to do it in a base model where you don't need the biometric energy density because you have the space that was set up for a larger pack, you can create some pretty advantaged costs. And that's what we're launching, again, as I said, it goes into the consumer vehicles early next year, but the commercial vehicles 100% of those will use our LFP packs later this year.

Okay. Yes. I mean I sort of think of that as maybe being 20% or 25% on -- if it's a smaller battery it's going to be $2,500 or something.

Yes.

Okay. So just given the market's recent focus on kind of free cash flow and so the recalibration evaluations and -- how has that changed either your strategic priorities or how you think about ramping over the next 6 to 18 months?

So we've talked about this a bit in the past, our last earnings call, we focused on -- I think we all can recognize the world is a very different place today than it was 6 months ago, heck, 3 months ago, and it's incumbent upon us and the leadership team to recognize the world is a different place. So the capital markets, appetite for risk, appetite for lots of parallel growth is just different. So we've adjusted our plan to ensure that the balance sheet that we have today, which, as of the last earnings call, around $17 billion, supports us launching and ramping our R2 program, meaning we don't need to go back to the capital markets to raise more capital. And in some ways, we can have a level of -- we can focus on the business and not be overly concerned with some of the short-term perturbations to share price. And that long-term orientation is enabled by the focus on simplification of what we're doing and how we're doing it. So we had some additional R1 variants that we're planning in our production facility in Normal. We've stopped some of those. We had some additional commercial variance that we've stopped -- I should say, paused development of those. And I think one of the things that's hard to fully appreciate is the OpEx intensity of launching a product is high. And this year, we're launching 4. So they have R1T, our R1S, our EDV 700. That's a 700 cubic foot version of van, and then the EDV 500, which is both shorter and narrower and to stack 4 launches on top of another, it's a very, very heavy year of spending. Each one of those has a ramp curve. Each one of those are [indiscernible]. Each one of those are parts supplier bring up to validation. The beauty of 2023 is we're not launching any news. So we get to sort of harvest a lot of what we've invested. Now we're putting new technologies in, whether it's battery or drivetrain, but the heavy lifts are being done this year. So 2023 becomes a harvest year. 2024, again, is a harvest year, and then we have R2 being developed in parallel that we then launch in '25. So it's -- we've really simplified how we're focusing efforts to make sure that, of course, ramping Normal fully utilizing the 450,000 units of capacity and then adding in our Georgia facility with R2 and focusing on that. And of course, R2 recognizing that's a global platform that'll eventually be built in Europe and China as well, meaning that's a very high-volume platform. Getting that to market as quickly as possible is really key for stepping up considerably in terms of the overall volume of the company.

Final question. This has been a long personal journey for you. Reflecting on it, what are lessons learned? Like I think if I were to ask Elon Musk, he'd probably say everything takes longer than you think, manufacturing is hard as heck, and s*** happens. I don't know if that captures any of your sentiment, but what are your reflections kind of 12 years in, 13 years in? And what are the most important lessons learned for you as the CEO for the next 5 or 10 years?

Yes. There's many. I'd put it into 2 categories. I think in terms of some of the strategic decisions we've taken, recognizing the necessity of really building out a robust supply chain in the environment that we're in, and the areas recognizing the importance of controlling and owning certain areas or having very deep relationships in certain areas in the context of semiconductors. That's something that if we can approach things with that lens 2 years ago, we would have looked at semi sourcing differently. And I think that is -- I'm saying this for Rivian. I think every manufacturer, not just in automotive, broadly in the industrial space is thinking about their sourcing strategies and relationships in a different way. And as you said, it is -- these things are absolutely harder than they look. A vehicle has -- an R1T has 2,000 parts coming from 400 different suppliers at the Tier 1 level. If you look at Tier 1, Tier 2, Tier 3, there's over 25,000 components on the vehicle, and you can't make a vehicle with 24,999, you need all of them. So it's a very complex orchestra. And doing that across for different vehicles does amplify the complexity. But the other learning, I would say, which is very actionable for us as well is making sure we have the right teams in place. And we've spent an enormous amount of time really focusing on getting the right mindset, the right teams, getting the teams to work well together, taking the hard decisions. We have the wrong leaders, meaning removing them from the business. And as we look going forward, that's ultimately the degree to which we're effective in achieving our targets is the degree to which we can be effective at navigating very complex landscapes, whether it's the headwinds that we see today from a supply chain point of view, headwinds of the pandemic, headwinds of global conflict. But there's -- we're talking about an organization that will need to take many thousands of decisions every week. And the strength of the team doing that is really key. So it's something we focus on heavily, and it's something we continue to focus on going forward.

Great. Well, thank you very much for your time and for your candor. Appreciate you being here.