Okay, let's go ahead and get started. Good morning. Welcome to JP Morgan's 51st Annual Technology Media and Communications conference. My name is Harlan Sur, Semiconductor and Semiconductor Cap Equipment analyst for the firm. Very pleased to have Bill Betz, Chief Financial Officer, and Jeff Palmer, Vice President of Investor Relations, at NXP Semiconductors here with us today. NXP is a top three global microcontroller supplier and the world's second-largest automotive semiconductor supplier. Bill and Jeff, thank you for joining us today.
Thanks, Harlan.
You know, semiconductor industry revenue is forecasted to be down about 10%-12% this year. We're in the midst of a cyclical down cycle for the group. Consensus has your revenues down 2%-3%, significant outperformance by the NXP team. Relative to your 2021 analyst day financial targets, even with downturn, you're driving a two to three-year revenue CAGR that is still within your financial target of 8%-12% revenue growth on an annualized basis and margin profile that is above your target range. The outperformance is clearly driven by exposure to auto, core industrial, which has held up well, disciplined inventory management by the team and the six accelerated growth drivers, which we'll talk about, and that you also laid out at your Analyst Day about 18 months ago.
I guess starting off from an end market perspective, can you guys just take us through what the team has seen in the first half of this year and anticipated profile of the business through the second half of the year?
Yeah. I'll take that one, Harlan.
Mm-hmm.
What we saw in the first half of the year, we did see some better supply that allowed us to grow our auto and kind of core industrial businesses better. We probably were overly conservative in our outlook into the first quarter, given a kind of unknown of what was happening with the China reopening. We did do better than we originally had anticipated versus our Q1 original guidance. We did make this statement on our earnings call a couple weeks ago that we thought the second half would be larger than the first half on an absolute terms. How to kind of build that up is it assumes that auto and core industrial continue on the same trajectory they're on, so no deceleration there.
We do have a premium handset customer that has pretty normal product launches in the second half of every calendar year, and they appear to be on track for this year, so that'll provide some tailwind. On top of that, we've been able to provide some supply to our secure cards business, which has had some pent-up demand for the last couple years. Also the RFID tagging market is finally becoming material after a number of years of waiting for it. You put all those together, and mathematically you do see the second half being larger than the first half. What it doesn't contemplate is we don't anticipate a significant rebound in China.
Mm-hmm.
If that was to occur, great, but it doesn't contemplate that. Secondarily, it does not contemplate us refilling our distribution channel above the 1.6 months level that we've been running it at for quite some time. Pretty natural growth, if you would.
Perfect. On the supply side, the team mentioned back at earnings, you know, one-third of your products still carry 52-week lead times. This mix should continue to drop as you move through the second half of the year. If we look back, however, your historical average lead times, I think is around 15, 16 weeks.
Historical.
You know, my guess is that although a majority of products have fallen below 52-week lead times, they're still well above historical pre-COVID-19 kind of mid-teens levels. Is that the case, number one, and does that suggest that mature node wafer availability is still somewhat tight?
Yeah, let me take that question, Harlan. The way to think about the 1/3 comment that we made, as you know, we entered into non-cancellable, non-returnable orders for about one year. Think about that supply being used up related to those type of orders.
Uh-huh.
The rest of the portfolio is what we try to indicate is back to normal.
Yeah.
Related to what's still tight on the technology nodes, we still remain tight on 55, 90, and 180, and we expect those to hopefully by the year-end work each other out and be back to somewhat normal as we look ahead into 2024.
Perfect. Your channel inventories, you know, customer distribution inventories are 1.6 months and still sitting 35% below pre-COVID-19 levels. The team has been pretty smart about keeping channel inventories low, which drives maximum customer responsiveness and a more real-time gauge on demand inflection, and also provides NXP with some flexibility. I know you said that the second half outlook doesn't contemplate any channel replenishment, but given the high responsiveness that you get from your customers, especially on inflections that they see on your business, has the team thought about keeping the channel inventory targets sustainably below that sort of 2.4, 2.5 month sort of historical range?
I'd say no. I'd say one of the things that the 2.4 months is kind of based on how you could provide optimal service, right? You do wanna have material on the shelves for your distribution salespeople to sell readily. The whole idea currently about keeping it at 1.6 has been to keep the inventory on our balance sheet because it gives us maximum flexibility.
Mm-hmm.
Once you sell a product into distribution, it basically title transfers, and we can't redirect it.
Yeah.
In this current environment where things are tight, it's an uncertain demand environment, it's best to keep maximum flexibility on our balance sheet. Over time, the goal is to bleed off the balance sheet inventory into the channel, but that'll take multiple quarters. That won't happen all at once.
The size of that magnitude is roughly about $500 million, when we believe the market and our customers are needing it. Like Jeff said, it's sitting in die bank. It takes us about three to four weeks to finish it out, and we'll be able to serve it. We feel very good about that.
Yeah. So it's in die bank, so as you mentioned, right, very quickly for the team to sort of bring it to market, so you're primed for that. My question is, you're seeing a better second half outlook. I think a lot of your consumer businesses you feel like are sort of bottoming. What are the metrics, triggers, signs for the market that you're looking for to start to take those channel inventories back up?
Sells out of the channel. If you think about a channel as sell in, where we recognize revenue and sell out, we monitor both of those as well as the inventory in the channel on a daily basis, and we review it as a management team weekly. If and when we see sales out of the channel begin to incrementally, steadily improve and re-heal as we kind of go through this trough of the environment we're in right now, we will take the action to slowly rebuild the channel.
I guess to follow up on that, your China business is mainly through a sort of distribution. Post zero COVID policy elimination end of last year, post Chinese New Year this year, you know, consumer and corporate demand in the region has been more muted, right? Although it did perform better than you anticipated in Q1, China exposed businesses which are more IoT and mobile appear to have stabilized, at least. Has the team started to see signs of a pickup in distribution sell-through in the region? I'm curious.
Well, it was better than we had anticipated.
Mm-hmm.
Let's kinda If you think about the conversations we had internally in the fourth quarter getting ready for the earnings call, we were uncertain about how the China market was.
Mm-hmm.
gonna evolve. We took a very conservative view in how we projected our guidance for Q1. It was better. It is incrementally better, but we're not. We don't think it's a huge rebound at this point.
Is it just because that it's not broad-based? Are you just seeing it primarily in, let's say, your auto end markets or?
No, it's across the different end markets that are serviced by distribution, but it's not back to where it should be.
Mm.
Right? Where we'd like to see it.
Yes. Okay.
I think we've maybe gone past the trough.
Mm-hmm.
Are we on the accelerating other side of the trough?
Yeah.
Harlan, it's not certain yet.
Okay. Perfect. Any questions from the audience? If you have any questions, just raise your hand and we'll get a mic to you.
Yeah.
You have a question up here.
Hi. Thanks. You gave us some brief comments about auto in your opening remarks. Could you maybe just put some more numbers around the inventory levels and how you see that playing out for the second half of the year?
When you say inventory, you mean inventory at customers?
Yeah.
Yeah. As, as probably most of you know, it is near impossible for us as a semi vendor to give you an exact sense of how much inventory is at the tier ones, who is our primary customers we sell to. Through conversations, we do hear anecdotal, there might be some small pockets. We spoke about this on our earnings call, two particular tier ones in Europe. It really wasn't an excess across the board. It was really as a result of kind of a golden screw problem, where OEM couldn't get a component from another vendor. They had had enough of our product as other and other products. We don't really have a sense of how much inventory is there. That is a fundamental challenge of the way the supply chain works.
Thank you. I was wondering if you could make some commentary about what some people see as a lot of incremental capacity coming online globally, especially from Chinese domestic players on the microcontroller market?
Yeah. For us, our primary partners for foundry is... About 60% of our revenue we source from the foundry marketplace, just to level set you. Our primary suppliers are TSMC, GlobalFoundries, a little bit of Samsung, primarily those vendors. Our customers have become very risk-averse, and they are not open to taking product built in kind of what I would call second-tier foundries in mainland China. While there is probably more capacity coming online, our customers are not qualifying that product.
I think the second part of his question was just, competition from some of the up-and-coming China microcontroller, players.
We do see some low-end microcontrollers in China, more for the industrial IoT side of the market.
Mm-hmm.
No one yet looking like they're delivering auto qualified type products. It's not a matter of never happening. Probably will at some point, can't deny that, but we're not seeing a design win loss or anything like that at this stage. Still very low level type of, competitive products.
Any other questions? I've got one right here.
Yeah, thank you for taking my question. On the topic of microcontrollers, when do you foresee we're gonna reach a balance between supply and demand? At the same time, what would that imply for ASPs for 2024?
We think that by the end of this year, end of calendar 2023, we should be in a more equilibrium of supply and demand. Not just on microcontrollers, but our total product offering.
Thank you. Just as a follow-up on the China market, do you think we could see a bifurcation in terms of pricing very aggressive, for the lower quality products and less so for the kind of products that you're selling?
It's a case by case. We try to not sell our products on a price basis, but on value to the customers. If the customer does not value the technology we're bringing to the table, then it may not be the perfect opportunity for us to engage on.
The highest volumes for NXP are centered around, I believe, 28, 45, 55, 65 nm manufacturing nodes, which I assume are technologies that you outsource to your foundry partners. You mentioned 55 90 in particular as products that are still in tight supply. Not that long ago, I remember we had a discussion, you know, discussing the longer-term industry capacity expansion profile on these mature and specialty nodes, you know, that would probably continue to grow slower versus the actual demand curve, keeping the supply-demand relatively tight. I'm talking about over the next few years. Even as we move through this period of slow demand, longer term, in discussions with your foundry partners, your internal capacity plans, do you still see industry capacity and mature and specialty still trending below the demand profile?
I think in the short term, the answer is probably yes, Harlan.
Mm-hmm.
You know, our foundry partners are deploying capital and CapEx for new capacity. I would say it's not a, just a floodgate. For ourselves internally on that 40% of our capacity that we build internally, we have reoriented our factories to be primarily mixed-signal proprietary processes. We've pushed a lot of all bulk CMOS out of our own factories.
Yep.
Our foundry network internally is all mixed signal. We aren't building any new greenfield capacity there. We have fixed four walls in our different facilities, and we're just reconfiguring it right now. That will take probably a couple more years till we're completely done with that project. We've been probably in process for a couple of years already.
You know, in the U.S., we had the National Defense Authorization that was signed off by Congress back in December. It is basically, you know, sign off on the defense spending budget for the year. There was some additional verbiage that was put in there around U.S. government applications and restrictions around China-focused semiconductor content or, you know, China foundry-based content going into government applications and having restrictions put around them. This has catalyzed a lot of the OEMs out there, auto, industrial OEMs. They see the writing on the wall, and they're already starting to make moves around over the next few years, trying to secure their suppliers that are more U.S. or European-based, right, with manufacturing in the U.S. or European.
Has the NXP team seen that level of engagement step up because of some of these dynamics, just even over, you know, past six to nine months?
Yes. We've had a number of large OEMs who've come to us and said, "If you wanna bid on this program, you have to show us the supply is sourced either in domestic, in the United States or in Europe." You know, of our 40% that we build internally, we have one factory in the Netherlands, and we have three factories here in the United States, two in Texas, one in Arizona. We have a joint venture in Singapore with TSMC. I think from a already owned factory footprint, we're very well aligned to that kind of mindset of domestic or geopolitical de-risking, if you will.
Yeah. Let's drill down into some of your end market segments. Automotive, you know, based on first half 2023 annualized revenue estimates, you guys are driving a 16% CAGR on auto over the past two years. It's tracking well above your 9%-14% revenue CAGR target. Strong outperformance here. You've obviously got a solid anchor of core portfolio of sensing, processing, control applications. In electrification and EV, let's start off there. Good portfolio of products. Business has basically, you know, doubled in electrification, you know, in calendar 2022 to $400 million. You've essentially hit the targets that you laid out at your analyst day about two years ahead of plan. You know, one of the drivers, obviously, is EV production, right? That continues to be strong.
What else in electrification is driving the strong growth dynamics for the NXP team?
I think the biggest one, Harlan, you kind of hit on it there, is our view. When we originally put out that view of our electrification business growing from $200 million to $400 million, we had a certain view of what the penetration of electric vehicles would be in global production. Clearly, it has gone much faster than we had anticipated, and so that has benefited us. We've done very well with a lot of the local Chinese kind of startup companies, very well. We do have a number of European OEMs who have adopted our electrification business. We feel very good about that trajectory continuing as it's progressed so far.
One of the other dynamics, and one of the accelerated growth drivers for the team is your S32 domain and zonal processor solution platforms. That was about a $300 million business in 2021. It's, I believe, tracking to your 25% CAGR target through 2024. The trend in domain zonal computing is leading to significantly higher requirements for compute power in the car. You guys have highlighted a number of wins with the major OEMs. You've also said the ramp, you know, you're hitting your targets, but the ramp, the strong ramp really doesn't begin until 2025. We understand the growth dynamics between now and 2024. How does that revenue profile unfold beyond 2025?
Yeah. Great question, Harlan. It's probably one of the most exciting areas of the business, and the whole kind of driver fundamentally pushing that is this idea of the software-defined vehicle. The challenge is the OEMs are re-architecting their vehicles completely ground up. That does take time. From what we can tell from the design wins we've been awarded, the knee of curve, the growth begins in 2025 and 2026. We've not put a growth rate on it, but it is significantly higher than the roughly 25% that we're gonna experience between 2021 and 2024.
Let's move to your core industrial. You know, 60% of your industrial business is core, what you guys consider core industrial. It's levered to strong multi-year demand trends, factory automation, energy infrastructure, building automation. You know, the team here continues to leverage your strong processor and market leadership, drive increasing attach rates to your connectivity, analog, power management, security products. I think a good reflection of this is that your Crossover and i.MX application processor families grew 50% year-on-year in 2022. Help us understand what's driving the design win traction with the i.MX portfolio. You guys probably do track this. What's the rough attach rate of your analog power connectivity products when you do win an i.MX socket?
Yeah. Yeah. For the i.MX family in particular, Harlan, I would say the attach rate of PMICs is probably 100% because for most advanced processors, the PMIC is kind of designed in cooperation with that design. Some of the crossovers may not need a specific or custom PMIC, so it may be less than 100. I'd say we are still learning how to sell the complete kit, not just the processor and the PMIC, but also the Wi-Fi attach and the other type of connectivity attach. I would say we're still in the early days there. Probably clearly not, you know, even 50%, but doing well.
Any questions from the audience? Let's talk about your comms infrastructure and other semi, 'cause you did mention it in your prepared remarks. In addition to auto, core industrial, your comms infrastructure and other has actually been a very resilient business, right? It turns out you have some nice drivers of the business in RFID, secure card access. What are the applications driving these two product segments, and what's your view on the service provider spending trends in the comm infrastructure side of that segment?
Like all companies, Harlan, we have, our comm infrastructure segment is a catchall of multiple-
Mm-hmm.
different product portfolios and really consists of three unique non-interrelated portfolios. One, our RF power products for the base station market. That market continues to be very, very lumpy. I think the one opportunity we see in the, in the intermediate term is kind of the build-out, carrier build-outs in India, but it will be lumpy. So far, it's not been substantial so far this year. The second one is our cards and tagging business. There's two parts of that business to think of. There's been a pent-up demand in our e-government and some of our banking card businesses and transit cards. This is a business that during the pandemic, we just were short capacity, and it was a bit starved. Now we have been able to apply some capacity there, and that's been good.
The part that's more secular in nature has been the RFID business, and this is something we've been investing in for a number of years.
Mm-hmm.
It's been one of those kind of ideas long in the tooth waiting to really take off.
Yep.
The idea is to be able to do individual tagging of components and track components through a supply chain, like consumer products is what I mean. That business has really started to take off quite substantially and will probably play out very well over the next several years. The third portfolio inside of that comm infra is our multi-core digital networking business. It's a kind of a sunset business that came to us from Freescale. Not overly focused on it, but it's continued to hang in there pretty well. At some point it will sunset in the, in the distant future.
I'm gonna rewind back to auto because I forgot to touch upon one product, which is one of your other accelerated growth drivers, which is radar, right?
Yeah.
Strong leadership position, 77 GHz.
Yeah.
That was a $600 million business in 2021. Is expected to drive to $1 billion in 2024. Is the team meeting its growth targets there? Obviously you've got an expanding pipeline, more radar sensors per car. Yeah, I just wanted to know, is the team on track to hit its or exceed its targets for radar?
Clearly on track. Where we're at, where we're focused right now is actually winning the new opportunities that will go into production post 2024. If you think about it, the design wins that will drive that $600 million to over a $1 billion-dollar business have already been won, and we're monetizing them. The way to think about that business, what drives that business, it's kind of a three-way multiplicative business. Every year there are more cars with radar. In each car there are more nodes...
Mm-hmm.
of radar. Each year there's more demand for functionality in each of those nodes. Each of those kind of like opportunities drives revenue for us.
Got it. Any questions from the audience? Okay. On the financial front, Gross margins are already above the high end of your target model even as the team is driving mid-70% utilization levels here in the June quarter. Additionally, I believe the team's new product pipeline is at gross margin levels that are accretive to your current gross margin. How should we think about the trajectory of gross margins from here? Because it looks like the bias appears to be to sustain or expand above your target range as revenues expand from trough levels here in the first half of the year.
Yeah, Harlan, for the rest of the year, we expect our gross margins, what we can see, to be at the high end of the model, plus or minus the 50 basis points in any given quarter, related to it. Yes, you're right, we're running utilizations in the mid-seventies because Jeff mentioned earlier, we are retooling, reconfiguring our factories. 8% of the CapEx we're spending on, 75% of that CapEx is going into the front end internally as we focus on the IP proprietary information. What's offsetting that, because again, 85% is roughly the sweet spot where you recover enough fixed costs. What's offsetting that is the mix, the way we service our revenue. It's not by segment, obviously by product. Inside the segments, gross margins are different.
It's really, if you think about it, between direct and distribution. Last year, we ran distribution in the mid-50s. The way to think about distribution is low volume, high margin. In Q1, if you can tell, we serviced distribution only at 48%, and we expect distribution, how we service our customers going forward for the rest of the year, be higher than that. Therefore, distribution having the lower volume, higher margin versus the direct business, which typically has higher volume, lower margin. That's a nice tailwind to offset the headwind of the underutilization on the front end.
I'm gonna go back to your industrial and IoT segment. Here, you know, 40% of the business is leveraged to more consumer IoT applications like smart home, connected gateway, and so on. This business has a higher sort of tier 2 business with a long tail of customers and particularly strong exposure to China. I know this goes back to our prior question, but this is one of the first segments, I think, where you guys started to see weakness.
Yeah.
What are you seeing there in terms of demand consumption? I know we talked about maybe seeing some better trends in POS in China. Is this one of the segments which is driving that sort of goodness that you're seeing out of China distribution sell-through?
Harlan, I'd say the consumer IoT business is better than our conservative outlook in Q1, but it's not back to where it was or where we would expect it to be.
Any questions from the audience? Capital return program. The team has had a very, very strong program targeting 100% free cash flow return. Obviously, you know, continues to be, you know, underpromise and overdeliver on that front. How do we think about how the team thinks about allocation between dividend and share repurchase?
Yeah, there's no change to our capital allocation policy. Obviously, like I mentioned on the call, we are buyers. We like to cost average. We are buying back the stock. The way we get comfortable around that is we take our most pessimistic view internally, and we do the intrinsic value on that, and clearly, the stock price is very attractive. Dividend, as you all know, we raised it last year 50%. This year, another 20%. I think it's running around a bit higher than the 25% of cash flow from operations. I think it's around 30%. We feel very comfortable with that. That's more fixed in nature. We are holding a little bit extra cash on the balance sheet. That'll give us the opportunity if we do choose so. A lot more options at play here.
Buy back the stock, increase the dividend, but as well as maybe pay down our debt that's due in March 2024. Again, the best use of the cash is to reinvest it in the business, and we believe we have a consistent policy here and expect to return 100% of the excess cash flow back to the owners.
Perfect. Well, we are just about out of time. Bill and Jeff, I wanna thank you today for your insights and look forward to continuing to see the team execute as we move through the second half of this year.
Great.
Thank you so much.
Thank you, Harlan Sur.
Thank you, everybody.
Thank you.