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

Good afternoon, everybody. My name is John Vinh. I cover semis here at KeyBanc Capital Markets. And we're pleased to have Jeff Palmer, VP of Investor Relations from NXP. Welcome, Jeff.

Great. Thanks, John. Appreciate it. It was a pleasure to be here. Thanks.

Great. Maybe Jeff -- maybe we can start with your Auto business more broadly. I feel -- I think your message was that you're generally seeing pretty consistent strength within your Auto business. I feel like from some of your peers out there, there's been some mixed messages within the Auto business. I'm wondering if you could just help us understand what you're seeing more broadly in autos, maybe talk about it by different segments, EVs or ex EVs versus ICE and maybe talk about it by regions? Are all the regions and all the segments of your Auto business doing well? Or is it just kind of a portfolio approach where more broadly, it's doing well, but then maybe there's pockets of weakness here or there.

Yes. So I think more broadly, it is a portfolio approach. We did say that North America and Europe were solid for us. China is okay. Probably still needs a little bit more rebounding. Clearly -- and we said that some of our customers specifically Tier 1s in Europe may have had a little bit of excess inventory. We highlighted that in Q1. We think we're kind of working our way through that. That's included in our guidance and our outlook for the second half. Clearly, there's been kind of a normalization going on, right? We guided Automotive to be up mid-single digits in Q3, which is sub trend, if you would. So I think we're navigating through it. I think even inside of that, we do have some company-specific drivers, which we're really excited about. We've talked about these before, our 77 gigahertz radar business continues to do very well. We expected that to grow from about $600 million to $1.2 billion in '24. We're right on track on that. Our electrification efforts, we said we'll grow from about $200 million in '21 to $500 million in '24. We're actually quite a bit ahead of that. We hit $400 million exiting '22. I think that is more because our original model probably didn't contemplate how fast the mix of electric vehicles would accelerate. And then the really exciting areas is in our domain and zonal processors, which is really a secular move towards the software-defined vehicle, which will play out probably over the next 10 years at least. But the initial early revenue there is about $300 million of revenue in '21, growing to about $600 million in '24. We're right on track. But the real native curve of growth is in the model year '25, '26 and beyond and design win rates there doing very, very well. So all in, we feel very good about the portfolio. The thing we're really focused on right now today is really those design opportunities in the model '25, '26, '27 and beyond. From now to '24, it's kind of locked in, in terms of design wins and things like that.

Great. I know you talked about just not focusing on any one given quarter in terms of your Auto business, but obviously, you kind of acknowledge that we are kind of going through a deceleration phase as some of this starts to normalize. How do you see this playing out? Do we kind of trough in the second half of this year? And when could we get back to kind of your targeted normalized run rates in autos?

Yes. We think we've troughed in Q1 in the first half actually. Even though 5% is not bad, it's not where we'd like to be. So I'd say further into next year, we should probably reaccelerate. Global production is still expected to be up this year as well as next year. So it's not a demand problem. It's just the supply chain is so extended, and we've gone through a couple of years of supply-demand imbalances. It's just going to take a little time to normalize that, and we think we're on our way there. Lead times have normalized at this point. We think customer inventory, Tier 1 inventory is pretty much normalized.

Great. Speaking of kind of lead times normalizing, I think pretty clearly, you and your peers had benefited during the last several years of kind of the shortages really had caused a lot of your customers to kind of mix up to the high end, mix up to high-end electric cars mix up on EV. So as things start to normalize, I think there are some concerns that, that could generate a less favorable mix. One, do you see that happening with the automotive customers? And as your customers' automotive mix starts to normalize, what impact is there going to be on kind of ASPs and your margins?

Yes. So for us, I think the important thing to realize, John, is we don't have a portfolio that's one part is just for premium cars, one part is for entry level, one part is for mid-tier. We have a cohesive portfolio. We just tend to sell more of the same parts into premium vehicles. We have not yet heard or seen in our forecast a mix down. I will acknowledge on a car -- per car basis, content per vehicle, mid-tier probably has less content than a high end, but we could probably make it up in volume. So we're not that worried about it. And there's really no impact on margins whatsoever.

Great. Do you want to talk about this zonal processor opportunity, Kurt seems pretty excited about it. Can you just help us understand the opportunity on a software design vehicle -- defined vehicle. Typically, how many of these zonal processors, domain controllers are there? What are your typical ASPs on these controllers? And then you talked about an inflection in 2025. What's driving that inflection?

So you have to realize what the challenge for the auto OEMs is. The current state of car architecture is kind of flat point-to-point bit of a bowl of spaghetti, if you will. As you have more and more data coming into the car, more data moving around the car, that architecture is just not extensible. So the auto OEMs would like to go more of a higher structure into the car. And that's -- the most important thing for them is software life cycle management and software modularity. And to do that, you have to have a very cohesive computing fabric across the whole car. And so we -- as we talked about a few moments ago, we've rolled out our domain and zonal processors. What was missing in that initial foray was more of that gateway device, that kind of what we call vehicle computer. So a way to think about this from a hierarchical perspective, you'll have a vehicle computer. This is our 5-nanometer multicore, pretty impressive product. That will manage multiple domain controllers. Below that, you'll manage multiple zonal processors, and those zonal processors will then manage edge microcontrollers. So that's kind of the hierarchy. What we can offer to the auto OEMs is the ability to have a cohesive architectures so they can run the same software from the edge all the way to the top. And to kind of put in context what we're talking about an average edge microcontroller, call it, maybe $5 on average in the industry. One of these vehicle computers is about 10x that. We have instances where we have more than one of them in the car, and they could then manage, call it, maybe 3 other domain processors. And then those domain processors can manage another 4 to 6 zonals depending on the architecture. And then you still have tens and twenties of the edge microcontrollers. So the content in the car, it's quite a big uplift for us.

What's the range of content that you could get on this? And...

No, I think that what we've said is if you just look at the vehicle -- the vehicle computer, if that's called a 10x in microcontroller, 3 of them $150 and then you add on to that, maybe another $30 to $50 is in domain processors. So on down.

And the 2025 inflection is driven by design wins related to your gateway processor ramping at 10x ASPs.

Right. But as that starts to feather in, that is not just a vehicle computer and then nothing down to the edge. It's the vehicle computer, domain processors, multiple zones and even the edge business continues. So it's really -- it's quite, I think, an inflection in terms of the dollar per content we are going to drive.

Very nice. Any questions? Great. I'm wondering if we could just maybe talk about your China business. I think you've talked about that business being muted. What are you seeing in the China market right now? And is that more of an in-demand issue, Jeff? Or is it more of an inventory issue when we think about your IoT business?

For us, for the way we see it, we'd say more of a demand, inventory [ by then ] already [ rationalized ] out. I think the challenge was a lot of folks in our industry expected post the COVID reopening, there would be a real rebound like you saw in the West, and that is just really not materialized. In China, is a very long tail business for us, tens of thousands of smaller customers that go through distribution, both consumer IoT and core industrial. We use distribution in that segment quite heavily about 85%. [Audio Gap] I think what's important is our outlook for the second half of this year does not contemplate that this happens, it rebounds in any material way this year. And also, we don't count on the Android marketplace rebounding either.

Your assumption in the second half is sequential growth in the second half...

What we said is the second half of '23 will be larger than the first half of '23, but it will also be larger than the second half of '22. So given what we've printed, what we've guided, you can kind of get into a plus or minus [ round ] what Q4 would be.

And if the assumptions doesn't assume, a recovery in China doesn't assume or recovery any Android, what are you assuming some reasonable ramp in iOS...

Correct?

And what else...

Continued solid demand in automotive and core industry.

Got it.

Right? The base station business, which is in our consumer communication infrastructure, kind of soft and squishy, the base station piece of that, what's really helped us there is having the card-based businesses the security card, the RFID tagging, which has been pent-up demand. So that's helped. So that's kind of what -- the way we see the rest of year playing out.

Yes. Yes. Maybe we can talk about that a little bit. So obviously, the security card business offset kind of the weakness in your Comms-Infra business. So when you talk about pent-up demand, this is credit cards or government IDs or what's driving this?

Primarily government IDs like passports. They're talking by everybody traveling, nobody can get a passport. Tagging and transit, so mass transit. Some credit card, it's not an area of huge focus anymore, but it's still some credit card business and then RFID tagging? In RFID, we've been hoping for this secular trend to take off for a number of years, and it finally looks like it has. Now one thing that's interesting is the tech -- process technology that we use for these card businesses is similar to what we would use in the handset for the secure wallet. So as Android weakened a little bit, gave us a little bit more capacity, we were able to redirect it and address some of the pent-up demand that we couldn't address last year. It doesn't mean things like the cards and the government stuff is a new secular trend, it's more pent-up demand. It will normalize in some way. RFID tagging does appear to be a secular trend.

Okay. On RFID, are you gaining share in that market? And then roughly, when you think about your whole secure card portfolio, how much of that secure -- well, non-comms business is RFID versus secure card?

So we're not going to break down individual product pieces that granularly, John. What we have said in the common for business, it's really 3 portfolios. The base station piece, the RF-powered base station piece. It's our legacy network processor business, which is just kind of steady eddie, no real growth there, but very, very long tail and then all the secure card businesses. And we've said that each of those is about 1/3 of that segment.

On the comment for business, what are you seeing out of India these days? And when are you expecting that comms business, which is obviously being driven by 5G. When does that come back for you?

I don't think we would hold our breath through this year. I think it's kind of soft and squishy. We had been hoping that the India build-outs would take place, doesn't seem to really have to materialize. And I think you listened to the Nokia and the Ericsson calls, it's not great. I think we're probably, as an industry, I could be wrong, but I think as an industry, we're probably past peak 5G build-outs. You'll continue to see some occur and there will always be maintenance, but I think the steep curve up isn't behind us in the industry per se.

Maybe we can talk about pricing. It sounds like you're really confident about being able to pass through these incremental price increases from your input cost to your customers. Is this mainly an Auto and Industrial comment or are you able to pass through higher input costs in some of your other consumer-facing segments as well? And then conversely, as supply and lead times start to normalize or have been normalizing, your input costs are being -- are probably coming down as well. On the flip side, are you expecting to kind of pass through those cost savings to your customers?

So the reality is our input costs, specifically around foundry costs aren't coming down. So pricing will be up again this year. They were up -- kind of up low single digits in '21. They were up about 14% in '22. We know they'll be up this year, not as much as they were last year, and we're getting indications from some of our foundry partners that prices will probably go up again this year. And we've been very transparent to our customers. The only thing we are doing is passing on inflation and input costs to maintain our gross margin percent. So your question is, if our prices were to go neutral or even down, are we mandated to give them to our customers? No. But if let's say you went to a neutral inflation -- input inflation type level, we could decide at our discretion if we wanted to offer some small ASP reductions if we wanted to.

And then...

It is primarily in terms of segment. It is primarily a Automotive, core industrial type of situation. A lot of the consumer business is one in the core -- consumer IoT business is weak. So there's really not a lot of demand there right at the moment. And in Mobile, it's a different beast altogether, as you know.

Any questions there? Okay. Industrial IoT for you guys declined, I think, 19% year-over-year but did grow 15% on a sequential basis. Can you just talk more specifically about what's been weak here? Is it mostly China across what applications? And then regarding the sequential growth should we view that as a sign that you are seeing some modest improvement in the China market?

Yes. So let's set the table for what the -- this is a very long tail business. About 60% of it is Core Industrial, 40% is consumer IoT. I would say the sequential trends are positive. The year-on-year trends are still down, but they're getting less down, if you will. And what we're looking for before we refill the channel is for those year-on-year trends to turn positive in a more consistent way. So what's been weak? It's really been around the consumer IoT stuff. And this ranges from wearables all the way up to home automation type products. That's what our consumer IoT business is. Our Core Industrial is primarily factory automation, building automation type products, so a little bit of health care. And the Core Industrial piece has been relatively solid in North America and Europe.

Maybe switching gears to your Mobile business. It does look like you're seeing a seasonal ramp in third quarter for flagship phones. Can you give us what's driving your Mobile business these days? Should we think about your Mobile business going forward is tracking the handset market? Or is there opportunities still for you to grow incremental content there?

There's opportunities to grow. But let's talk about what -- the majority of the business is secure wallets. So if you've ever used your phone to pay for something, you're more than likely using our technology, whether you use an Android or iOS. In that area, that is primarily an attach rate type of a driver. Believe it or not, still only about 50% of all smartphones have a mobile wallet embedded in them. The Android players tend to be much more adaptive or quickly moving if the market doesn't look like they can generate a return on it. They may design it in, but then they'll defeature it. Whereas on the other handsets is a little bit kind of single SKU type of model, right? We do have an opportunity to continue to do some custom analog products that we've done in the past. And we -- I'd say the big secular opportunity in handsets for us is really adding the ultra-wideband radio to the handset. Now in that part, in ultra-wideband, it's really an Android play. And the customers have been very interested in it, but the Android market demand has just not been great. So you think about what you'll do with the ultra-wideband as you basically virtualize your car key into your phone. And so there's 2 halves to it just like on a credit card, you'll have the key in the phone, and then you'll have this opposite system in the car. So all the car OEMs that are designing ultra-wideband secure access solutions are using NXP. We feel very good about that. That's actually coming along okay. But we need the Android piece to come along a little bit more.

Great. That sounds like an interesting opportunity. Can you talk about what sort of attach rate or penetration rate we're looking at in cars right now? And what's the adoption on the Android side?

I'd say on the cars, it's very -- still very early days. It's not even a number that I would say it's X percent. But you can start to see in BMWs and Hyundais, a few others that are hitting the market, it's getting out there. We did not put out an attach rate target. We put out what we thought the revenue would be in '24. Initially, we said it would be about $80 million in '21, growing to about $400 million in '24, and that's both the handset side and the car side. Clearly, the handset side is behind a little bit. So we'll have to see how next year plays out.

Okay. That's fair.

It's an incremental few dollars on the handset. In the car, it's actually -- because the way the system works, it's actually quite a bit larger. So if you have maybe $2.50, $3 in the handset, it can be $20 in a car.

That's helpful. Maybe we can switch over and talk about your gross margins. You've done an excellent job there in terms of managing gross margins through this kind of soft spell, if you will. Despite lower fab utilization, you're running at peak margins. Can you walk us through what's allowed you to kind of keep gross margins so still strong.

I think the thing to remember is over the course of the last several years, our fixed and variable costs have flattened. So our fixed costs are only 30% of our total cost. So 70% of our costs are fully variable. And that goes along with the fact that we build 60% of our wafers externally. That's one. Two, the company was structured to get to about at least $10 billion to be in the sweet spot. If you remember back in '19, we struggled to hit the 55% number because we just couldn't get the revenue there. So now here we are solidly above that $10 billion. Utilization on a standalone basis, yes, is dilutive right now, but we have some interest segment mix things going on this counterbalance here. But positive trends would be as utilizations go back up into the mid-80s, that should be a positive trend to gross margin. As we refill the distribution channel, which is normally about 55% of our revenue that would be accretive to margins. So I think that's the way to think about it.

I think you might have said that you guys are not resetting your long-term gross margin target...

At this point.

Given that you're at the high end, and it sounds like for the foreseeable future, given the potential tailwinds that you're still potentially going to see long channel, favorable mix, utilizations kind of trough, like why wouldn't you...

We usually don't change our model except for our Analyst Day. So our Analyst Day is usually every 3 years. Current model is 55% to 58%. I think the way to think about it is 58% is not the destination. It's a place to get to, hold on to and then put a model together further out. I think that's something to look forward to the Analyst Day next year.

Okay. I need to ask the obligatory AI question. It seems like investors are just hyper-focused on it. I mean indirectly or directly, are there opportunities for you to benefit from AI?

I think if I'd be objective -- I think indirect is the way to think about our opportunity. So as more and more data centers start building the systems, the capability to do these large language model development, you need to be able to run those models at the edge. And that's really where we play at the edge, right? We're not a data center player at all. [Audio Gap] Recognition in the radar processor. And a lot of that learning was done in the cloud and the data center and we download the models to our processor. That's one example. In industrial, we definitely do a lot of object identification as well as human interaction like if you go to an industrial equipment and you want to be able to log in and reach your eyes or reach your fingerprints, that's kind of AI object type of classification type of things we do. We're not a competitor to something like an NVIDIA or even Marvell in the data centers. That's just not what we do. Think about it is as these models are developed and you need to put them into real-world situations at the edge, that's where we'll play. That's our strength and our focus.

Yes. Is there -- you have a sense of roughly what percentage of your revenues are kind of edge AI type of...

Well, if I wanted to AI wash with the revenue, I could come up with all kinds of numbers, John. But the reality is, if you think about our radar business, which we talked about, right, almost over $1 billion next year. That has basically AI applications. Talk about our Core Industrial business, which is 60% of a 20% end market, right? So it's about 12%, right? That's got AI opportunities. We don't want to be seen as just jumping on a bandwagon because everybody is excited about it.

Got it. On channel inventories, you're running at like 1.6 months in the third quarter. I think you had talked about if you were to replenish the channel, it'd take potentially a couple of quarters, potentially that would be a gross margin tailwind for you. What specifically are you looking for from a demand signal perspective to give yourself confidence to start replenishing the channel? Is it backlog? Is it book-to-bill? Or is it something else?

It's year-on-year growth. So right now, while we believe we've started to resume sequential growth from the trough of Q1, what we really care about is more year-on-year. And right now, in some of our business, just obviously look at the numbers, we're still down year-on-year. So as we start to get more positive trends on those year-on-year trajectories, we have the ability to refill the channel. And the delta between 1.6 months and 2.4 months is about $500 million. And right now, we're keeping a lot of the inventory on our balance sheet in die form, so that we can quickly turn it to our back end in 2, 3 weeks into quarter. But ultimately, once we feel like we're in a more normalized environment, we would move about 25 days of that balance sheet inventory into the channel over a multi-quarter period, get us back to the 2.4 months, right? And as you do that, as you start to bring down your balance sheet inventory your working capital metrics get better, but that at a certain point, you're going to trigger the ERP systems to do starts in the factory again, right? [ It's little pre-built altogether. ]

Great. I think we're out of time. Thank you, Jeff.

Thank you, John.