NXP Semiconductors N.V. (NXPI) Earnings Call Transcript & Summary

Good morning, everyone. My name is Bob McCooey, I'm the Vice Chairman at NASDAQ. It's great to see all of you. It's great to be back here for NASDAQ'S 47th annual -- well, I guess we do it twice a year, so 47th Investor Conference held in association with Morgan Stanley right here at the Mayfair hotel. It's nice to see all of you back in person after a little respite and so many familiar faces as well as so many new ones. The NASDAQ Investor Program is built to show our dedication to the growth and advancement of our companies and the investors in this community. As we say at NASDAQ, our work with our clients doesn't end on their listing date, it extends well beyond Day 1. And we hope to celebrate many more milestones with them and opportunities to engage with them, and this one investor conference is exactly one of those. European investors play a vital role in development of the global financial markets, and there are more than 4,500 asset managers operating in Europe with the total assets of just under [ $31.5 trillion ], which is actually 169% of the European GDP. I'm pleased to say that there are actually 630 investors who have signed up for this conference today from over 300 institutions in the U.K., Switzerland, Scandinavia, Germany and Spain. A key contributing factor that we have been continuing to experience with this strong growth and momentum is in ESG investing. In fact, European companies are a major driving force behind the rise and interest in sustainability. According to NASDAQ'S own eVestment, ESG-focused products account for USD 1.2 trillion of AUM, of which over 50%, $665 billion, was European. There's certainly a lot to talk about in the world today, and you're going to hear from some amazing companies over the next 2 days. And it's our role to make sure that we're able to facilitate this meaningful dialogue between NASDAQ-listed companies and you as the investment community. This year, we will welcome over 50 companies to the conference, 14 of which are part of the exclusive NASDAQ 100. We're joined by the likes of companies at the forefront of each of their sectors from health care into technology, companies like Apple; Palo Alto Networks; Bookings Holding; Etsy; and for the 16th year, Akamai, just to name a few of our companies. On behalf of NASDAQ, I'd like to thank you all for being here with us today and the extraordinary work that you all do on behalf of investors in certainly some challenging years that we've had. Please don't hesitate to come to me or any of my colleagues over the next couple of days to offer any insights that you have or feedback on the conference as we continue to want to make it better and better for each of you every year. Before we kick off this year's program and our first group of presenters, I'd like to invite Frank Ducomble from Morgan Stanley, who's the Head of Global and U.S. Equities here in Europe, to say a few words. And we want to again say how much we appreciate the support of Morgan Stanley because without them, we wouldn't be able to put on this conference. Thank you all for joining us.

Thanks a lot, Bob. It was great seeing you back in person for having not met you for the last 2 years. Yes, it has been a great partnership with NASDAQ. I think this is the biggest conference of U.S. companies in the U.K. and Europe. And as you said, our corporate access team received more than 2,700 demands for meetings, of which they allocated around 60%. So I hope you are happy with your schedules. I'm joined by more than 10 of our U.S. tech and health care analysts. So please join them for their group meetings. We will also host a panel today at 12:00, where we discuss the key themes as well as the key picks going into next year. Obviously, please go out there, harvest the alpha, and I obviously invite you to partner with Morgan Stanley to implement all the good ideas you find here. I think it's now time to turn it over to Kurt Sievers; as well as Jeff Palmer from NXP; as well as Joe Moore, our semiconductor analysts in the U.S. Thank you.

All right. Good morning, everybody. As Frank said, I'm Joe Moore from Morgan Stanley. Very happy to have with us today the CEO of NXP, Kurt Sievers; as well as Head of Investor Relations, Jeff Palmer. You might see in your schedule. Bill Betz was originally here. Unfortunately, he came down with the flu last week. So thanks a lot for joining us, guys. I appreciate it. I wonder if we could start off with just a general discussion of what you're seeing with semiconductor industry conditions. When you reported, you talked about 70% of your catalog having greater than 52-week lead times, which is a lot, but it's down from last quarter. So maybe starting to see some normalization, and you said you're seeing some weakness in some segments continue to strengthen automotive. Can you just talk generally to those conditions?

Sure. And thanks, Joe. And good morning, everyone. Indeed, the way we talked about it in earnings, and it hasn't changed that much since earnings a couple of weeks ago, was we called it a dichotomy, which is really like 2 very orthogonal trends we are currently seeing. We do see, as you noted, a weakness and softness in the consumer mobile, in our case, actually consumer IoT and mobile and compute markets. While at the very same time, we do see very, very strong resilience in demand in automotive and core industrial. And that goes to the extent that in automotive and core industrial, we continue to be sold out. So the very long lead times of 70% of the portfolio is actually coming largely from core industrial and automotive. When we look at our NCNR orders, which are being placed in those 2 sectors for next year, for the next full calendar year, it also shows that we will continue to be sold out. We see about a 85% coverage level of demand with our supply capability for next year. And mind you, the supply capability is improving gradually quarter-to-quarter. So there is a growth in demand, a growth in supply. But still in that part of our business, which is the much larger part for NXP, we continue to be sold out, where, indeed, at the same time, we do have that softness in the consumer mobile and compute markets. Now in my view, this is a very understandable situation because the demand trends in those 2 different legs are having very different origins. I believe that the surge in demand for the compute, mobile and consumer markets was really a function of the working from home through the pandemic. While the demand in automotive and core industrial had actually not really anything to do with the pandemic, it is a secular trend. It is about content increases in electric cars with ADAS in factory automation, which was not accelerated by the pandemic at all. So we would have seen that anyway. And that's also why, in my view, the resilience of that demand is here to stay. Because it is not something, which now as a function of the pandemic being behind us, would come down like it does in consumer mobile and compute.

That's helpful. And I guess maybe if we could put some of the supply issue into perspective. I've seen a few of these sort of situations in the last 25 years. I've never seen anything as severe as what we've seen in the last couple. Can you talk to that? Is it unusual in your experience? And how do you think it changes customer behavior in the long term going forward?

Well, it's -- certainly, in my 30 years in the semiconductor industry, I've never seen anything near this. I mean very, very extreme gap between supply and demand. The majority of it, at least from our perspective, is in these trailing edge nodes, so I'd say between 16 and 19 nanometers. And again, I don't think that, that shortage was a function of the pandemic, but actually a function of collective underinvestment from us and all the industry participants in, I'd say, 2018, 2019, 2020. So we were -- in a way, we were hit COVID by the demand, which we haven't prepared for, for these trailing edge or mature nodes. Now it has done a lot, I believe, in terms of customer relations, how we experience the way we can do business. What started in the second half of 2020, really as a supply panning crisis, actually, customers once in a while yelling at you and a lot of hectic has morphed into what I would call 2 things. One is a much deeper understanding of long-term demand of customers and our customers' customers. So for example, the car OEs, which we speak to directly now. When I say much deeper understanding, then this is not about the next few quarters, but it's actually about the next couple of years, really understanding which platforms would be ramping, which technologies are needed and what does it mean from a demand perspective on us. While we have gained that much deeper understanding, we built relationships and now is the second leg, which really help us to build a much better innovation partnership, because most of these customers have figured out that they not only depend from a supply perspective critically on semiconductors, but also their future differentiation with their end product in their markets largely depends on semiconductors and software. While we haven't worked directly with them in the past, we do now. And I think there's a couple of giving long-term demand assurance and being awarded long-term design wins, massive platform wins is, I would say, a completely new game. So we have massively stepped up how we can work with these customers. I'd say in hindsight now, thanks to this situation. Now it was very painful by the way. So not an easy reward, but it got us into a very good place.

And in terms of resolving that now, how fungible is the capacity between the markets that are soft? Can that capacity be deviated to sort of resolve the issues in automotive side?

Unfortunately, very, very little. That's a common question also from customers, by the way. Customers read newspapers, and they see the demand drops in the computer and mobile markets. And of course, they -- at first stance, they don't understand why that wouldn't help us immediately to give them more product. It doesn't because most -- it's not black and white, but the largest part of what we need into automotive and core industrial are dedicated process technologies like with embedded flash, with extended temperature profiles, which is not what becomes available from the mobile and compute and consumer markets. So it is the mission profile of the technologies, and it is also the node size. I mean a lot of what is becoming available now is actually sub 16 nanometers, which is not what we need in automotive and core industrial. So that's why the fungibility is very limited. I mean we and our foundry partners, we do what we can in trying to retool to the extent possible, but it's not like that this solves now the critical shortage.

Okay. Great. And on that note, foundry pricing for those wafers has been going up. You guys have successfully passed that on to your customers over the last couple of years. Can you talk about where we stand now for calendar '23 in terms of those prices?

Yes, we -- I mean if we say this year is almost behind us. We indeed, we have now 2 years of price increases behind us, which, by the way, Joe, is about as unusual as the supply situation from a shortage perspective. I mean the one, of course, is the reason for the other. And indeed, the input cost, which we face and the largest chunk of that is indeed wafers from foundries, is set to continue to go up also into next year for these mature edge nodes. With that, we are being forced to continue to raise our prices to our customers. We have had and indeed successfully, if you will, a very transparent philosophy in how we are raising prices. We pass on the input cost to the extent that we protect our financial model. Or to be more specific, the gross profit percentage is what we are protecting. So not padding the margins, but also not losing out in the middle. Given that input costs will continue to go up, my answer is yes, we also will continue to raise prices into next year.

And can you talk to the cyclicality versus sustainability of that? Is that -- the foundry arguments, I think, are that they need to make new investments to support these nodes versus using depreciated capacity the way they did before. So they're arguing that's permanent. Do you think that's the case? Or do you think there's a cyclical decline in pricing at some point in a way?

So first of all, I think the argument is really a true one. I said earlier, there was a critical underinvestment for a couple of years for mature nodes, and that's exactly what has to be caught up now. So there is a lot of investment. It is new factories, new capacity and not depreciated assets as it used to be in the past for the mature nodes. And because of that, I do believe there is indeed a continued cost burden going forward, which is coming from these assets. So in my view, for the critical mature nodes, and again, let's talk about 16 nanometers to 90, in some cases, 180 nanometers, I do not think that prices will flip back down. So I see about 3 years of price increase, calendar '21, calendar '22 and calendar '23. Then hopefully, maybe towards the end of next year, we get more in the balance of demand and supply, and then we will operate from the new level of pricing. Maybe then we will see years again with very small annual price erosion like it used to be, but not slipping back to the level where we came from initially.

Great. So I want to dig into the markets a little bit, but maybe just one last question.

And maybe, Joe, one other thing I have to mention here. What is burdening cost further and with cost also pricing is a strong trend to regionalization in wafer manufacturing. So the geopolitical situation, as we all unfortunately see and feel every day, is leading many of our customers in the Western world, in the U.S. and Europe to make regional wafer manufacturing a condition for business awards. So we have customers which clearly say we only award you that chunk of business if you can give evidence that you will be in a position to manufacture it a 100% on the front-end side in either Europe or the U.S. Now we all know that manufacturing in the U.S. and Europe is more costly than doing it in Taiwan. And for that reason, I think you have here another lever, which is actually kicking the cost rather up than down and with that also pricing going forward.

Yes, that's a great point. So referring into end markets, maybe just one last kind of question on this unusual supply cycle. You mentioned that the automotive companies are taking measures to avoid this in the future, one of which is for the Tier 1s to be directed to build inventory. Where are we in that process? How high does that inventory go? And does that create headwinds for you at some point when it stop building?

Yes. So they take a lot of measures, by the way, or they try to take, which in itself to start with is a very good thing. So I feel the industry for the first time, though, is not going through such a crisis and immediately starts at square 1 again once the crisis is behind us, but they actually try to take structural learnings. And the bigger ones, by the way, are the much more direct work with companies like NXP in order to be at the source of where the product and the innovation is coming from. Now building inventories is another one. There is not a uniform trend. What we are seeing and hearing is that there is, with many of them a target to have safety stocks between 1 or 2 quarters. The logic for that is simply the manufacturing cycle time for a typical product from us is 3, 4, 5 months. So 1 to 2 quarter safety stock would kind of protect you from cycle issues. Now from anything -- first of all, I should also say we will not hold that at all. So I mean we will make sure that it is being held elsewhere. With all of that critical product, which is in shortage, I do not see at all, Joe, that any of this is being built as we speak. I mean it remains a target. It remains a target for next year. I don't know when the capability will be there. But at this stage, that inventory hasn't been built at all.

Great. So maybe if we could shift to some of the growth drivers, and I know you guys do kind of an every 3-year Analyst Day, which is really good, really rich kind of view into where you guys perceive the growth to be coming from. Maybe we could just talk to some of those. Starting with battery management. You've talked about revenues growing 2.5x in 3 years. But you've also talked about sort of being #2 in that market, and I know you always want to kind of get to a leadership position. Can you talk to your progress in battery management for cars?

Well, that's the one which sees enormous traction. I mean it's sailing with the enormous tailwind of electrification ever accelerating. I think it's like 7 -- to 27% or 28% of all cars this year being produced will be electric, hybrid or fully battery electric. And I think it's going to be 34% next year. So it's really like 1/3 of the car production globally needing a big battery. So -- and that's when you refer back to the Analyst Day, we did like November last year, we did not have that estimate show. So we are running ahead, simply because the industry is converting much faster. Now our value proposition is very specific, and I think it's actually unrivaled in the industry. What we are offering is analog precision front-end chips, which are sitting on each of the battery cells in these large battery packs and a microcontroller. And the microcontroller is this element in the middle, which is steering the cell balancing between the individual cells. And the whole system has to be ASLD fully-functional safety compliance in order to make sure the car wouldn't catch fire, et cetera. That value proposition show is actually what differentiates us from also the #1 player in that market, which is why I do see us catching share. We are still behind because, as we all know, that #1 player actually acquired another one, which made them even bigger. So we have to run harder to catch up. But I think with that specific value proposition, which I just described, we are in extremely good shape. And when you refer back to the Analyst Day, I think it's fair to say that we are quite a bit ahead of the target, which we had set out with the $500 million for 2024. Now I should also mention, it's not only about battery management systems, we also work on inverter control. So think about anything which is logic and analog mixed signal needed for electrification is what NXP is doing for the electric car.

And how do you think about markets like silicon carbide, which you guys haven't, I think, had a big presence in? Is that going to be a target market for you? And is it -- that's where a lot of the increase in content comes with the fully battery-powered vehicle but there's still a lot of opportunity for you guys, I assume, around that.

Well, the way it works is if you think about the semi content, which is specific to the electric drive train, about 50% of that opportunity is in power discretes, which includes then silicon carbide IGBTs, et cetera. The other 50% is logic and microcontrollers. Our absolute focus is on the logic and microcontroller side because that's the technology and the -- let's say, the IP, which we are mastering. We leave the discrete power side aside. Reason, to be honest, being that we think the capital intensity and also the typical margin profile in power discretes has not proven to be really compatible with NXP's model. So we focus on the other half, which is as big as the first one. And battery management, inverter control, hybrid microprocessors are actually examples for that.

Great. And then on the topic of microcontrollers, where you guys have a very strong position, you've talked a lot about the S32 family. What's happening generally to microcontroller content in the car, whether it's unit content or pricing content?

Well, pricing is going up like everything. I mean they are just in the middle of the shortages.

But more than just like-for-like there's an increase in -- to the volume as well?

What I would say the content is growing massively, and I believe it's fair to say this is not just about microcontrollers. It's about microcontrollers, DSPs and microprocessors and hybrid innovations in between. What really is happening is that you get this trend to domain computing and sonar computing, so a more orderly electronic car architecture, which, in the end, leads to significantly higher requirements to compute power in the car with very, very dedicated silicon to the specific real-time compute requirements in a car. Think about over-the-air updates, which have to be managed. Think about the whole idea of the software-defined vehicle, where the biggest pain point for the car industry is actually not hardware, but software. It's the life cycle management of software through the life of a car. When you get a new car, it feels kind of new. But what you want is that after a year, it still feels new. So it gets new functions coming in through software. So the hardware has to provide enough headroom, and the architecture has to be coping with the possibility for that change. So software compatibility, scalability is the name of the game. And this is the S32 family, which you just mentioned. I think it's absolutely the industry-leading concept to master that requirement. So enormous amounts of compute power, scalability with a full portfolio from relatively low-end microcontrollers all the way up to 5-nanometer-based microprocessors, which we have designed in now.

I mean one of the really appealing things about this market has been the durability of design wins that the repeatability from year-to-year. But you're also describing quite a bit of change, whether it's the sort of computation of over-the-air updates and those kinds of things, migration to batteries like. How does that change the content that isn't directly impacted? I mean people rethinking the entire car. Are you seeing design wins come up for renewal that are different than we've seen in the past?

Changing the architecture show means the whole car is changing. It's a very, very deep transformation. What, however, is important to understand is that, for example, when you think about electrification, the increase in semi content is not only what comes with the electric drivetrain. But typically, electric cars are much more advanced from a technology perspective overall. Because you as a consumer, you expect a more techy car when it's electric, more displays, more safety functionality. So for us, we actually found out over the last 2 years that we are over-indexed to electric cars, which means the more electric cars we see, the more acceleration we have, relatively speaking, in our semi revenues into automotive because it also pulls all of this other content. They are just much more loaded with semiconductors.

Great. And then the third growth driver that you talked about in autos was radar. And I know you guys -- that's kind of your ADAS positioning. You're less focused on vision and things like that, you're focused on radar. Can you talk about that and where you're seeing the increased adoption of radar in these vehicles?

Yes. Radar is already quite advanced. So we are coming from a 600 million run rate last year, shooting for almost doubling it in '24, so to 1.1 billion with a 20% to 25% CAGR, I think, which is really standing on 3 legs. Radar is penetrating into cars. So more cars which will have radar. Then you will have more radar nodes per car because inside the car, there is more applications. So what used to be automated cruise control and emergency break only has now also blind spot monitoring in the side mirrors, et cetera. So more radar nodes per car. And then we see a trend which we love to push, which is actually more dollar content per radar node. Because the performance requirements for systems like imaging radar, where radar builds like a cocoon around the car to get a perfect digital image around the car with many, many radar transceivers sprinkled around the car, is actually picking up speed. So 3 legs of acceleration: more cars with radar, more radars per car and higher silicon content per radar node. And that, in our #1 position, is actually a beautiful trend. Now a lot of technology push, by the way, so we really have to innovate fast. Industry is moving from 24 to 77 gigahertz, so higher frequencies, and industry is clearly moving from silicon germanium to CMOS. Because in CMOS with the 77 gigahertz integration, you can get the whole system into one chip products, which from a performance, cost and space performance is actually the best we can get. Now that's exactly where NXP has been leading for the past couple of years.

Great. That's helpful. Next, maybe we'll shift to industrial. And then after this, we'll open it up to the audience for any questions. But industrial and IoT, you've talked about kind of a mixed picture there. Can you talk to that a little bit? And can you talk to the growth drivers that you're most excited about?

Yes. Our industrial and IoT segment is actually, you could say, split into a business-to-business and business-to-consumer part. The business-to-business is what we call the core industrial. This includes factory automation and systems like that, which is 60% of the revenue segment. The other 40% is more like a consumer business, which is what we call the industrial IoT, consumer IoT business. And that's the one which is seeing the softness currently. So 40% of our industrial IoT revenue segment is seeing the softness. The other 60% continues to be resilient, as I said in the beginning. Now what makes us a leader in this segment is our system approach. So I would really say over 6 years now, we have carefully architected a system portfolio, which gives us everything from sensing over the compute element, connectivity, security and actuation. The connectivity part was the Marvell connectivity acquisition, which we retreated a couple of years back. And actually, the processing versus analog side came from the Freescale-NXP combination 5 or 6 years back. That system capability is a big plus in that market, where we work with a long tail of small customers. And think about automated vacuum cleaners. Think about cloud connected coffee machines. The manufacturers or designers of these systems do not have a lot of experience in electronic design for cloud-connected robotic solutions. They -- all of them have robotic elements. They are cloud connected. So cybersecurity, functional safety plays a big role. So what we have learned to do is give them reference designs. There, more or less everything is done, but the user interface. So they basically only have to program the user interface. The rest is being provided by NXP, and that's a huge plus in our value proposition. So I'd say that's the result of a long architected portfolio, which is now coming to fruition.

Great. So let me see if we have any questions from the audience. If not, I can keep going. Wait for that mic.

So getting back to the consumer markets, which you mentioned, they're still soft, especially around mobile and compute and because there was a push forward during the pandemic, I'm wondering when do you see it bottoming out because we're getting almost to the 3-year mark in the pandemic, and it has to cycle eventually.

I really can't say. I have to -- I also have to admit, we are not a prime mobile or prime consumer player. I mean that is, for us, the smaller part of our portfolio, and I don't have the crystal ball when that would be bottoming out. I can only confirm that it keeps being soft. I mean we -- in our guidance for Q4, we also clearly reflected that softness. In our case, that hits us in the Android space, on the mobile side, so the lower-end mobile, as well as this consumer IoT business. But I really can't call when that would be bottoming out for now.

Question from the audience? Charles?

How important are analog semiconductor design engineers for you now? And is it easy? How difficult is it to recruit them?

Is the question how important are analog semiconductors?

Yes, and especially the engineers who understand how to write the designs.

Very important. There is no electronic system without analog semiconductors because as long as we human beings are using electronic systems, there will always be this match point between the digital world and the analog world because we are living in an analog world. And these are analog chips. So there is -- it's just -- there will never be a moment where we don't need analog chips, so we will continue to need understanding them. We will continue to need engineers to support them. And I think the sophistication level for analog chips, which includes, for example, very sophisticated power management units for the big processors. So there, you'll see that the one -- it's like yin and yang. I mean the one needs the other one. It's ever increasing. And by the way, those circuits are needing those famous trailing edge or mature nodes. I mean so it's not only the engineering competence, which continues to be needed, but also the manufacturing capacity continues to be needed. Absolutely, it goes hand-in-hand with the digital systems.

We have time for maybe one more question.

Maybe quickly because we're running out of time. What about -- you mentioned the higher cost of production in Europe specifically. What about energy supply? Has that been a concern, an issue? How do you navigate through that, either you or anyone else in your value chain? Has that been like -- what's your...

So I got to say that in our case, we are lucky, if you will. We only have one wafer facility in Europe, and it's not in Germany. That helps a little bit. Structurally, I would say, certainly, we have to look at competitiveness of energy supply in Europe. However, for semiconductor manufacturing, while that is important, it is not the biggest deal. I mean there is really more -- there is more important other elements. I'm more worried about the overall cost level for semiconductor manufacturing, both in Europe as well as in the U.S. Now in the U.S., the energy problem might be a little bit less, but still the cost is relatively elevated. And as I said before, in my view, this is an unfortunately irreversible trend from a geopolitical resilience desire perspective, and that's why we will all have to carry that. In the end, it's going to land with consumers. And I just mentioned it to close the circle here in the context of pricing because it's yet another element even beyond the shortages, which is going to keep pricing rather up than bringing it down again.

Okay. We'll wrap it up there. Thanks very much.

Thank you, Joe.