Good morning, everybody. Thank you for attending our first session of the day. My name is Gary Mobley. I'm with Wells Fargo Securities, one of three semiconductor analysts here at, at Wells. Joining me today on stage is the management team from NXP Semiconductors. To my immediate right is Kurt Sievers, the CEO, and to my far right is Bill Betz, the CFO of the company. And in front of me here is Jeff Palmer, who heads up Investor Relations. Many of you probably know Jeff quite well. To start out our discussion today, I want to lay a foundation for the audience online and here in the room. Maybe you can just give us an overview of the business and what you're seeing from the cycle right now.
Sure, let me kick off. Good morning, Gary, and it's just wonderful to be here. It's great to get some fresh air outside as well. A quick overview of NXP. We have four end segments. Clearly, we're a leader in automotive, represents 56% of the company's revenue. The way to think about auto is, you know, really talk about three growth, accelerate growth drivers: radar, battery management systems, as well as our software-defined vehicle. The next segment that we spend a lot of attention on is industrial IoT, represents about 18% of the business. And again, think about AI edge, think about the smart home, smart devices, the smart factory. And then we play in a niche area in our mobile space, so think about your mobile wallet, and that represents about 10% of the business.
And then finally, our Comm Infra part of the business represents 16%. And think this, this is more of a catch-all. We have a legacy business that we purchased from Freescale, called our Digital Networking business. So that's just a very, very long revenue stream that we stopped investing in, but more of just continued to live it out. Followed by, think, secure payments, secure cards, passports, RFID, a part of that business is there, as well as we have our 5G base station business as well, which is somewhat lumpy and a bit weak in the current environment. From a financial model, we have a stated model out there to do about 55%-58% gross margin.
We're currently performing at the high end of that model, and we expect to continue to perform at that level through this cycle. Operating expenses, we have a good control on them. We wanna stay at 23% or below, with some leverage on the SG&A side. And operating income, we're running around on the high end of the model, Gary, of between the thirty-two and thirty-six percent. And from a share capital allocation, real strategy, if you're new to NXP, we like to return all excess free cash flow back to our owners. We are actively buying back the stock. We, we believe it's undervalued and, we'll continue to return that excess cash to you all, our owners.
Thank you for that overview, Bill. Automotive, which is more than half of your revenue, actually closer to 60% right now, you know, has been the main reason why you've actually been able to post some fairly resilient results in this in this softened environment. But the common opinion from investors is that this end market is going to roll over because all the others have, and the logic is there for automotive is, you know, more than likely gonna roll over as well. And so I wanted to get your perspective, Kurt, on the health of the automotive market and how you see that, you know, playing out next year against what is presumably a 1% growth rate light vehicle production.
Yeah. Thanks, Gary. I think it's actually more than automotive. Mind you that also in industrial and IoT, our guidance for the fourth quarter is a high single-digit growth, both sequentially as well as year-on-year, which is distinctively different to all of the peers. I mean, they are all speaking about how much the revenue and the demand and everything is dropping. So I think it is the approach which we have taken on the cycle, which differentiates us here pretty much. And that is all about the attempt to avoid overshipment into the market. That has two legs. Bill talked about the four segments we have. Another way to look at the company is we have two ways to reach the market.
One is through distribution, which is 50+% of the company, and just 50-% is direct customers. On the distribution side, six quarters ago, when supply started to normalize out of the supply crisis, we took the decision to not replenish the channel to our long-term target of 2.5 months of inventory. Back then, again, six quarters ago, we were at about 1.5-1.6 months of channel inventory, which then was a function of the supply capability.
So after that, we would have had the capability to ship more, but we said, "No, we don't wanna repeat the mistakes which we did in the past cycles, where we overshipped into the cycle, and then indeed into the channel, and then indeed had all the cycle problems." So for that reason, for the past six quarters, we took a lot of discipline. Bill and I are reviewing this every week in order to not fill the channel more than this 1.5-1.6 months of inventory. The difference, by the way, between that level and the target level is about $500 million.
So, think about a situation that for six quarters already, we have orders with us, where distributors would want more product, would take about $500 million of revenue, but we have resisted that temptation in order to soften the cycle. The other half of story, somewhat smaller half, is the direct customers. Here it's admittedly harder to trace, to not overship because we don't have formally speaking, any transparency or any contractual situation where we could trace exactly the inventory low level of our customers. But, we had those so-called NCNR orders, which are non-cancellable, non-returnable orders, which we entered into in October 2022 for the calendar year 2023. And that covers most of our automotive direct business and most of our industrial IoT direct business.
Early this year, those customers who had entered those commitments with us actually came to us and said, "Ah, we see we are entering into a problem. Because if we purchase everything we have committed to, we will end up this calendar year with too much inventory." And after some discussion between Bill and I and the management team, we decided to not enforce those NCNR orders. Because it was very obvious that by enforcing the orders, you create short-term revenue, but you don't create any demand. I mean, it doesn't create not a single car is being built more just because you push the product into the Tier 1 companies. So what we tried to do is we tried to find in each single case, commercial alternatives. So we asked customers for additional design wins.
We asked customers for price increases, anything which would be a gift from them, such that we could give them a relief on the NCNR orders. And with that, Gary, we believe that also on the direct customer side, we have probably somewhat less overshipped than most of our competitors. Take these two pieces together, we simply had our trough. We took the medicine already in Q1 of this year, so it's not like we don't have the cycle. We just had it earlier, and we've been not on the gas pedal. We have undershipped demand all through this year. I give you one last data point on this.
If you take our quarter four guidance for this year, with the actuals of the first three quarters, you will easily find out that revenue this year of NXP will be about flat over last year. At the same time, we also told you that we will have a price increase this year again. So the volume which we serve into the market this year is actually negative. And that's just another indication to how much we are undershipping demand, and we are kind of anticipating, the drop already. And that's why in last earnings, we could then go out and say, "Listen, from here on, we see actually all of the next quarters," so Q4, that's what we guided, but also the quarters through next year, with year-on-year growth.
We don't see this drop due to the cycle, but again, it's only because we anticipated it already in the first place.
Okay. It's very comprehensive. Thank you for that, Kurt. You mentioned something that I wanted to dig in a little bit deeper. And you, instead of just asking for the, you know, or building goodwill with customers and allowing them to push out NCNRs, you asked for some additional design wins, asked for some future pricing stability or increases. Is that any more pronounced in any one end market?
It's probably the biggest, the biggest part of that is in automotive. And that is because automotive is the largest segment on its own right. I think, Bill, you, you said 50, 50+%. Secondly, I think it's fair to say that probably, probably almost all of the direct business in automotive was and is under NCNR orders. So it's just the one which had the biggest exposure to that particular situation. Now, it's also good to do this in automotive because automotive is that market where you can gain a lot of traction with long-term design wins. I mean, it has a lot of longevity. There are many platforms out there. So it was like the ideal combination of we had something to enforce. We didn't want to enforce the, the, the volumes, so the good alternative was actually design wins.
That gets us a record design win, a year in automotive.
Okay. Well, you brought up another point that I wanted to dig more deeply into, and that is, the automotive end market. It should grow according to your overall revenue guidance, 9% this year. And I would assume within that, maybe your unit volume shipments are down, and so much of that growth is ASP driven. And that's against an industry backdrop where light vehicle unit production is up 8%. Is that in and of itself, and a softball question for you, an indication that you're undershipping into the automotive market?
Absolutely. That's exactly what I, what I said on company level before. I think in automotive, it's, it's most striking. Mind you, that the SAR this year is actually up 8%. So, I think the industry will build 88-89 million cars, which is an 8% increase year-over-year. And I fully confirm what you said, we will grow about 9% year-over-year in automotive. Actually, in the second half, only more like 5%, by the way. Price increases, we will tell you how big they were on company level, but indeed assume that there is not much left for volume growth. So, so that is exactly where we are eating away the over inventory already. That's exactly the, the reasons we have.
So we're in, not. I would say in not very-
And by the way, let me just add to this. This is markedly different to our peers. While we had outgrown everybody in 2021, the whole peer group was about the same in 2022. In 2023, we undergrow our peers. And, and again, this is very intentional, and that's exactly the, the story we had. So that's why I do think indeed, we all face the same cycle, but we are treating it differently.
Okay. I think it's fair to say that the light vehicle market from a sales perspective is less than ideal. Borrowing costs are high. You know, China's doing well despite the state of their economy on the automotive side. But my question to you is, you know, do you think we're gonna plateau at this 89 million annualized rate for light vehicle production? Or given the deferment of upgrades by the consumer over the last three years, is it possible we could see 95 million plus?
Well, I don't wanna be in the position that I could judge the future car production. It's really, it's really hard. What I would say indeed is that the 89 million this year in a very weak macro, I mean, all of the inflation, et cetera, is probably not stimulating people to buy a lot of new cars. So I would say it's fair to believe what S&P is saying, that next year SAR is flat plus. I think S&P says it's gonna be 1% up or something. So say about flat next year. What matters, however, much more to us is obviously what is underneath the further penetration of electric and hybrid vehicles. Because from a semi-content perspective, that is actually much more important than the absolute number of cars.
And the what we call xEVs, so the combination of hybrid and battery electric vehicles, that reached this year 33%. So 33% of the 89 million cars are either hybrid or fully electric, which, by the way, is a stunning number when you know that that same percentage was 6% back in 2019. So from 2019 to 2023, the share of the total car production globally went from 6% to 33% electric, which is massive. And that also explains some of the very nice growth which we could do the past years. Now, I think given the bigger numbers, that growth is gonna slow from a percentage perspective, that's normal, but we absolutely believe it will continue to penetrate.
I have to say, here in the U.S., there is a much more negative sentiment currently on EVs because you hear about the non-profitability of these cars, and you hear about big OEMs kind of delaying some of their platforms under the pressure. I believe that is all true, but from a global perspective, it doesn't matter, because most of the EVs which are being built and which are being sold are in China. So we talk about Tesla, and we talk about the Chinese OEMs.
Yeah.
So therefore, I am confident that EV penetration will continue. Maybe the pace is slowing, and it's a little, it's a little confusing to take the U.S. sentiment on this. I think from a global level, it actually does continue.
As a car guy and owner of a Tesla, I can talk all day about this. But, let's move on. You know, you, and Infineon are the top two players in automotive semiconductors, and close number one and number two. And what has become clear post-pandemic is that the automotive OEMs and their Tier 1s need to work more closely with the chip companies, and you've forged a couple of direct collaborative relationships with some of the automotive OEMs. My question to you on this topic is, do the big get bigger in automotive semiconductors, given, I guess, the dependence on the automotive OEMs? And related to that, do you have to feel a need to fill in product holes to satisfy all of their needs?
That's a great question because indeed, I think that is the case. And it is the case because through both the supply crisis, but also in consequence, the need to redo the electronic architecture of the cars, and by the way, Tesla has been a big leader in doing that. The collaboration and the proximity between semiconductor companies and OEMs has massively grown. I mean, that's a. It's a totally different world to five years ago. Five years ago, most of our business was both created and fulfilled through Tier 1 companies. In the meantime, there is a very significant share and innovation work, but also logistics, direct logistics contracts with OEMs. Now, if you put yourself in the shoes of an OEM, you cannot deal with 10 different semiconductor companies.
So indeed, I think there is clearly a trend for the bigger getting bigger, because they are just the stronger partners for the OEMs. So I mean, that's simple. The background is because OEMs need system partners. They don't need a niche player for one application. They need somebody who is in a position to discuss with them and shape with them the future full architecture, which includes digital, it includes analog, mixed signal, et cetera. Now, by the way, you compared us to Infineon. I think we are distinctively different companies. Infineon has a very large power business, which, so discrete, which is a great business. You all know the discussion about silicon carbide and stuff, which we don't do at all.
So our whole focus is on what I would say, true semiconductors, so logic, which is analog, mixed signal, and processes, in which we are by far the number one, because, again, we have zero discrete business. And in Infineon, that's a large part. So it's kind of, if you will, complementary, and that's the kind of partners the OEMs need going forward. Now, are we missing something in our portfolio? I'd say we are consistently working on tuck-in acquisitions, where I would specifically call out the need for software. I think we just published, Bill, I guess, 2-3 weeks ago, for example, an investment into a startup company in a software company for imaging radar. Radar is one of our big growth drivers in automotive.
Next year, like $1.2 billion revenue in automotive radar, number one worldwide, and the trend goes to higher and higher performance radar systems, which is more and more software, more and more AI. And that's, for example, an area where we just invested in a company. So it's not like the big areas that we would have a miss in the portfolio, but in those areas where we are, we wanna be absolutely at the forefront of innovation, and often that is in the fields of software.
Okay. Let's expand beyond just the scope of automotive for a minute. On your most recent earnings call, you gave some rough parameters around fiscal year 2024, and mentioned that each of the foue quarters should grow year over year, and thus, overall, maybe some modest growth for fiscal year 2024. What are the puts and takes to this? What can go right? What can go wrong?
Yeah, first of all, I have to frame this. We do not build this on a recovery in China. I make that comment because I remember that early this year, several companies, and we tried to be modest on this, were kind of building or hoping on a big recovery in China, which would be, of course, a big tailwind. So with everything we said there on the earnings call, which you just quoted, that doesn't need a recovery in China. Secondly, we said we will start to replenish the channel, but we don't need to fill the channel to achieve the numbers which you spoke about.
I just framed this because otherwise it sounds like it's a bit tricky, and you just push product in the channel, and that's kind of fake growth. That's not what is behind these numbers. The way we think about it is the following: We do believe that the content increases, which are growth drivers for us in automotive and industrial, in principle, are in place. That's the electric vehicles, that's the higher levels of autonomy going forward, that's the industrial automation, factory automation, et cetera. However, it is, of course, masked at the moment by this inventory digestion, which we discussed in the beginning of this chat.
Now, we believe that with our low channel inventory and with maybe into the middle of next year, some more inventory digestion in the direct customers, we don't have too much of that problem, which is that we are pretty nicely exposed to the ongoing content increases, which will drive growth in industrial and IoT and automotive. In mobile, which, as Bill said, is a much smaller segment for us, we have also no more inventory issues. I can clearly say that in the Android space, we have all the digestion behind us, and it was big and horrible. I mean, if you look at our Q1 from this year, it was really bad, and it was bad because inventory digestion, among others, in Android. That's behind us.
So with early positive signs on the Android market, we should participate in that growth. And then we have the U.S. premium handset maker, where our fulfillment model is such that there can be, by definition, no inventory. There is no inventory build. So as long as that goes in normal ways, I would say also mobile should see a path to growth. The fourth segment, communications infrastructure and other, which Bill talked about, that is the one which will not grow next year. Because we see the mobile base station market, the 5G build outs, very lumpy, very weak. So we don't think there is gonna be growth to be expected.
Secondly, we have the secure card business in that segment, where we had a good run this year, but a lot of that was pent-up demand from back in the supply crisis, where we had allocated capacity away from that segment. This year we shipped it, but it's done. So from that perspective, also, no, no growth to be expected next year. Finally, there is a piece in there, and Bill, you also mentioned it, the former Digital Networking business from Freescale, which for us is more of a cash cow business, and we will start to end of life a few products in there. So big picture, there should be solid growth in automotive and industrial because of our inventory policy, and the end market from a content perspective is in a reasonably good place.
Mobile, which is smaller, should also see some growth, but we will see a decline in communication infrastructure and other.
Just to be clear, you're not expecting a full replenishment up to 2.5 months of inventory?
Everything I just said is not built on that. However, our long-term distribution inventory model is 2.4-2.5 months. So this discipline to stick to 1.6 is because of the current negative environment. Once the environment gets into better waters again, we have to make sure that we have enough competitiveness on the shelf space of our distribution partners, and then we will replenish. But I can't say when and how much next year. In any case, it's not gonna be like one quarter, a big push. It will anyway be thoughtful and gradual because we don't wanna spoil the big advantage which we've built over the last six quarters.
Okay.
We will treat that with a lot of care.
Okay. Let's get Bill back into the conversation. So, so most of your, your peers are reporting or guiding for gross margins that are well below what they were probably over-earning during the 2022 timeframe, but you're still running at the high end or even above your guidance range of 55%-58%. What's, what's allowing you to, to keep this high level of, of gross margin?
... Yeah, no, thank you, Gary. I think as Kurt mentioned, is how we're treating this cycle is very different than our peers. We started much, much earlier, saw signs, and we're very cautious by keeping the inventory in the channel at 1.5, 1.6. Internal factory represents 40% of our foundry, our wafers that we do in-house. We start tapping the brakes on utilization over a year ago. As you recall, we were in the high 90s, kind of throwing money to solve supply issues. The sweet spot's around the mid-80s. We brought below the mid-80s for the last couple of quarters, I think, low 70s, mid-70s. And so we're having a nice offset because that's a headwind right now for the last couple of quarters, with our mix.
You could see that the mix of distribution, which drives higher gross margin because it's a long tail of customers, thousands of customers. So as you can see throughout the year, that channel of our revenue grew from 48% to 52% in Q2, in Q3 was 57%, and we see the same trend continuing at the moment. Those two are nice offsetting each other, but more structural, more foundational. What's changed the gross margin of the company happens over years. 10 years ago, this company was 45% gross margin. It had 70% fixed costs. Today, we're 30% fixed costs. So you reduce the variability when you have these revenue swings as you go forward through these different cycles. Another one is our investment strategy.
The way we make investments is we have hurdle rates after the attractive of the market and so forth. So 10 years ago, we accepted business that was greater than 45%, and that business came to be, and, you know, five years later, we improved to 50%. So all the revenue you see today is based on a hurdle rate, probably three or four years ago, around the mid-50s. And so that product is just coming to the market today, and all our new investments we make today of that 16% R&D, will then have a hurdle rate greater than that 55 in three or four years from now as well. So very disciplined. So think about new product introductions coming out over time, layering on top of our existing business, representing the quality of the portfolio.
Then lastly, obviously, scale helps. I mean, we struggled to be 55% gross margin until we got to that $10 billion mark. So as we could grow based on our CAGR figures of 9%-14%, that should also help us continue to improve the gross margin. Again, we're at the high end of our model. We have our Analyst Day coming up in a year from now. We expect to update our models as well at that time.
It is really the gross margin—the consistency of the gross margin performance, Gary, which led us to do what we do about inventory management. It is exactly that, because had we brutally overshipped and really just drove revenue hard, we would have a similar problem like others now, because then we could not live with 70%+ loading in the factories. We would run them lower, and you would see the impact, of course, in the gross margin. So the gross margin, the consistency of the gross margin over the quarters through the cycle, was actually the reason why Bill and I decided to be so cautious on the inventory into the channel. And again, I mean, we are now—I think we are largely through this.
So with that, Bill, you made the comment also in earnings that we don't see a reason that the gross margin has to drop next year. So we don't think there is a peak, like with many peers, which is now coming down, but we should just move forward. And if you then think longer out, and as Bill says, we're gonna give a new model in November next year. I don't think we have to stop at the 58.5. I mean, mind you, we do now 58.5 with low 70s loading.
Okay, well, I think that picture is clear. So we don't have to dig any more deeply into that topic. I wanted to shift to the topic of China. So as you probably hear firsthand, a lot of investors are concerned about China trying to foster a domestic chip industry, become more vertically stacked and especially in the automotive market, since, you know, 35% of current light vehicle production, you know, is in China right now. So they're particularly motivated to, you know, have an internal automotive chip supply chain. So my question to you to start out the discussion, you know, what do you think your China indigenous mix is? The mix of revenue assigned to China consumption.
Yeah. So, we have 30%-35%. It moves between the quarters, but say 30%-35% of the total NXP revenue goes into China, and about half of that, we believe, is China for China. The other half is sub-assembled and moves out of China, again, into the Western world. So round about 15%, if you will, of total NXP is then China for China. And I completely agree with you, the relevance of the China automotive market is, in my view, big and is growing. So I believe they are the winners. So it is, for us, indeed important, to keep that leadership position, which we have in China automotive especially, but also Industrial IoT is, for us, a big, a big deal in China.
It is also correct that in the past many years, to my surprise, I have to say, we have not had any significant local competition. So most of the competition we faced in China was actually Western companies, plus Renesas, which is amazing if you think about the potential in the country. So, in our view, it's absolutely to be expected that there is more local competition coming up. We saw this already years ago in low-end microcontrollers, outside of automotive, I have to say, which wasn't a big deal for us because that's kind of a fringe segment for us, where we never wanted to play on price anyway. We see now the first moves also into the more catalog type of product in analog mixed signal, where local startups are starting to try and be competitive.
And we do see massive investment in manufacturing in China. Manufacturing both in logic, so like the famous mature node kind of technology, say 16 nm and bigger, but also specifically in power discretes like silicon carbide and IGBTs, which doesn't matter for us. Actually, we are happy to see this because that keeps them busy on something which doesn't really hurt us. Now, how do we think about this, Gary? First step is, and I say this from a very fresh experience of being almost two weeks in China in late October. What all the customers want from us is localization of manufacturing. That's the first step to be a bit more Chinese there. They want that the front-end wafer manufacturing of NXP for them is in China.
Now, our largely foundry-based model, Bill just talked about the numbers earlier, allows us to do that. So we are—I mean, it's, it is effort, but it is not impossible. It would be much harder if everything we do was internal factory, because we couldn't build a factory in China. I mean, that would be really difficult. Since we are largely foundry-based for these products, we can actually work with local foundry partners, including TSMC, I have to say, because also TSMC has a big place in Nanjing, which does 16 nm and 28 nm, which are workhorse technologies for us. So the first big requirement, which is important to compete against that local competition, is actually to have wafer manufacturing capabilities through foundries in China.
Secondly, given the size and the criticality of the Chinese market, we started to change the way how we think about China from, yeah, we sell products, which we did in the Western world, into China, to making dedicated products for China. Because if you think about BYD and the likes, we see now companies which have the strength and also the differentiation to justify dedicated investments, which we do for them. And then, you know, as long as it is not regulated by governments, we should also not be overly paranoid on this. We have learned to be competitive. All of our business is about keeping competitive. Years ago, we had the same situation in Korea. So, I think there is a bit too much angst about local competition.
Mind you, that the bigger Chinese companies, they also need us and want to work with us because they want to be successful without- with us outside of China. They want to export their cars into Southeast Asia, they want to export their cars into Europe, and they want us to understand what the requirements are. So again, we are certainly, absolutely watching this very carefully. We do whatever it takes to stay ahead, but I also don't see a reason to be panicking on the local competition. Again, unless there is government regulation. If, of course, U.S. export control rules would hit more the type of products which we do, then we have a different discussion. Or if the Chinese government would have rules which say, you have to use 60%, local product, yeah, then also we have a different discussion.
But if it is about fair competition, I think we are in a great position to counter that. But I still agree with you, there is more competition coming up. We see that.
Okay. Well, we have 30 seconds left. This is a good stop point. Kurt, Bill, Jeff, I really appreciate you joining us here today. I appreciate all the people in the audience and online as well. Thank you.
Gary, thank you. Thank you all. Thank you.