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

Hi, everyone. My name is Francois-Xavier Bouvignies from the tech team in Europe, and we're happy to have NXP, Kurt, CEO; and Bill CFO of NXP. Just as a reminder, I mean, NXP is a top 3 auto semis in the world with more than 50% of revenues coming from automotive, and they have also a very strong footprint in microcontrollers and sensors where they are the leaders worldwide.

So thank you very much, Kurt and Bill, for being with us today. I will focus my talk into 2 parts, short term and then long term. So we'll start with the quarters. Don't worry, we will move to the more longer-term horizon as well. So maybe start with the short term. So maybe in the quarter, you guided maybe a bit lower than people expected, lower than seasonality. So if you look at your Q4, you guided for revenue down mid-single digit quarter-on-quarter versus seasonality around plus 1%. Can you walk us through the dynamic here? Why you are below seasonal? What is happening to justify that?

Yes. Thanks, Francois, and good morning, everybody. Indeed, like through the middle of the third calendar year quarter, we saw quite a deceleration in the macro, especially, I would say, in the automotive and industrial sectors, we try to witness this through the manufacturing PMIs, which had been growing through the first half of the year. So that's why we were also quite optimistic until then. But then like in July, August, things started to come down. Actually going back into contraction mode with the exception of China. So very clearly, the impact you are asking for, which we witnessed was on automotive and industrial, where given the weak macro, especially in Europe and the U.S., also somewhat in Japan, we see a double effect. We see that the end demand is actually coming down. You also saw that with the SAAR, which was now down corrected to minus 2% for the year. When we did the guidance for Q3, it was still seen flat. So everything deteriorating. And the second impact is that the inventory digestion, which we've been dealing with is actually getting a little tougher because those customers, and that's especially automotive Tier 1s, which had seen a reduction in the call-offs from their customers, from the OEMs. They, of course, then further down adjust the inventory levels because they measure the inventory as a function of their revenue. And if the revenue comes down, then that percentage further drops in terms of absolute numbers. So what we are seeing is an end demand deceleration and additional inventory reduction into Q4. That was the reason why we had to guide like minus 5% down quarter-on-quarter into the fourth quarter. Clearly, we are undershipping end demand. So when we look at these numbers and when we look at the car production on the automotive side, which is relatively easy to trace, we know that we keep undershipping, which is -- which means that we keep building down inventory at our customers.

And maybe a follow-up on that would be like where are you in this inventory correction. Can you share maybe some visibility of what you see the level is? Is it the right level? And importantly, the big question is when we get out of it?

Yes. So first of all, about 50% of NXP is shipping through the channel through distribution, where we have a very precise and very disciplined handle on the size of inventory. That's the famous month of inventory at our distribution partners. By the way, we are moving now to measuring it in weeks. That makes it easier for you guys because we found all of our peers are measuring in weeks. There we are at a low level of 8 weeks. That is well below our long-term target, which is more like 11 weeks. So the inventory size, if you will, on the distribution side, which is 50% of NXP is very right. I would actually say it is too low, but we are very cautious at this point in time to increase it further because we want to wait for the recovery and then actually ship more into the channel. So that side is clean. The challenge which we have is on the direct customer side, so that's the other 50% of NXP's revenue, where I would call out mainly automotive. So the largest part here is actually the automotive business. And in automotive, we have not a strict clear visibility into the inventory of our customers. They don't have to tell most do. So we have a pretty good relationship with most of them in terms of being very transparent on the inventory size. The challenge is that they have not stable and also very different targets. Numbers which I would give you is that they want to achieve something between 2 and 12 weeks. That's the inventory targets of our Tier 1 customers. You asked me if that is right and when it will be right. My view is that 2 weeks is absolutely not right. It is way too low because the reaction time to any change in demand when you think about our manufacturing cycle times being more like 3 to 4 months, -- with 2 weeks, you cannot react. So I think this is actually irresponsible and will create issues again once the demand comes back. 12 weeks is probably more where I would say the learning after the supply crisis landed all of us. There was quite some consensus in the industry that 12 weeks is probably a reasonable size of inventory. Now it is also not even, Francois, between all the different products because some products are really single sourced, have longer manufacturing cycle times. So we are more granular with customers to make sure that we have the right inventory size for the right product, which is -- but on average, then something around 10 to 12 weeks, I would say, is normal. Our feel is that currently, the average is moving lower. So we are moving in a territory given the poor macro, given the working capital challenges of these companies, we are moving in a world where things are moving lower than those desirable 12 weeks. which is part of the reason why we keep undershipping end demand.

And what is your lead time today? I mean, the visibility you have right now? What -- how long is that?

Well, visibility is really low because we have basically every product available. So when customers want the product, we have it, which means we have a lot of turns business. And that is exactly what I've seen and witnessed in earlier cycles in my career. That's always the start of the problem. I mean it start of the positive side, if you will, when you think about demand coming back, but it's also start of the problem because if that becomes too much, there will be a moment we cannot serve it anymore for each and every product. So visibility is extremely limited. We are getting a lot of orders within lead time. Lead time is the classic 10 to 12 weeks.

Makes sense. And maybe moving on to gross margin, Bill. I mean if I look at NXP, I look at a lot of the analog space, you have been standing out in terms of like -- I call you like almost a sniper in terms of gross margin. You're like very focused, very good in terms of guiding it. And if you compare with the others, you are not very far from the peak of the gross margin. It's like less than 100 bps. The others are like down 800, 900 in some cases. So clearly, you are standing out there. So it's a positive thing. Now the guidance is a bit slower or lower than we expected. So maybe can you walk us through maybe how do you manage this gross margin so well in the past 1.5 years? And what is driving maybe the lower-than-expected gross margin for the December quarter?

Sure, Francois. Good morning, everyone. When it comes to gross margin, there was a lot of lessons learned in the past. And what we've learned, it takes years to improve your gross margin over longer periods of time, driven by the mix, the NPI, the fixed cost structure. So in the short term, what can really drive gross margin is any large swings in revenue or mix challenges in any given quarter. And as we can see in the current quarter, we are guiding slightly down, driven purely by the mix that we talked about. And you could see it actually in our end segments where the industrial, specifically in Europe and Americas remained weak. As we go forward, again, I think the 2 levers -- well, there are several levers of what gets gross margin higher and what gives us that confidence of our new gross margin of 57% to 63%, bringing the midpoint at 60% will be the utilization of our front-end assets. The operational efficiencies back to normal pricing. We'll see that in the low single digits starting next year. And clearly, more longer term, we're super excited about the tailwinds with our investments in VSMC, our consolidation of our 200-millimeter factories and really to get NXP in the 60-plus range in the years to come.

Okay. Great. And what is your utilization rate? I mean you mentioned utilization rate.

We're still in the low 70s.

So if I may add, Francois, it's really the result of a long-term strategy, which was fixed cost reduction in our hybrid manufacturing model, but it also comes back to the discipline with the channel. A lot of the gross margin variation comes, obviously, if your revenue has a lot of peak-to-trough variation. By not overshipping in the channel in the first place, we avoided this, which means we didn't grow that much like 2 years back, but now we are declining less. I think with the guide for Q4 this year, we will be down 5% or so, which is really different to the peer group, which is just a result of not overshipping and not having to correct in the channel, which helps then again the gross margin. Gross margin is a function of variation in revenue.

Very clear. Moving on to the '25 and maybe beyond outlook because you had the Capital Markets Day not a long time ago. When you look at '25, it's very uncertain, let's say that, especially for the automotive industry. You have China threats, EVs, the car OEMs are having issues like every quarter, high inventories in some cases, like you highlighted, Kurt. So what makes you confident that '25 will be the year of recovery? I mean what -- with all this uncertainty, what can you say about your confidence around that?

Well, you see, Francois, we have -- as I said, we have little visibility into '25. If you would really look into the order book and stuff. I mean that is really limited. However, we clearly see that the algorithm of content growth into cars is perfectly in place. So the transformation to EVs, the transformation to SDVs, software-defined vehicles, which is a lot about processing and networking structures in the car is ongoing actually at an accelerated pace. And yes, quite a bit of that is driven by China. The acceleration is driven by China. So the development times for new car platforms in China are about half of what we used to know them in the Western world, which is a very positive thing for us because it gives us access faster to bring higher content and higher value product into cars. All of this is currently masked clearly by the inventory digestion, which we discussed a minute ago. Now admittedly, the SAAR is in not good shape, Francois. I mean, this year, the SAAR is going to be down 2%. Next year, there is different forecasts, but we assume it's going to be flat, say about flat. So we are not counting on the SAAR to deliver growth. We count uniquely on the content increase of semiconductors into the car, which again is then a function of EV and SDV transformations. Now China, we currently certainly see that China from a global perspective, is winning share in automotive over the Western world. That's a very clear, very strong trend. We have a good exposure to China. So we are actually benefiting from this because China is faster, and we are hedged between China, Europe and the U.S., which means if China is now continuously doing better, then that helps us in our revenue mix from a growth perspective.

So your market share, let's say, in China versus Europe and U.S., you don't have a big difference. You have a very...

It's rather a little higher in China than in Europe and the U.S. And that is increasing at a pace because they just renew faster, which gives us that growth opportunity quicker.

Yes. This -- like this renewal like upgrade faster in China has been a debate for the industry. Some could say that it is an opportunity, but it also could be a threat, right, especially given the China localization trend. How do you stand on that?

Well, so first of all, in principle, theoretically, you are right, but I -- it's a matter of attitude. If something goes faster from a competitive perspective, we see this as an opportunity. Now we are not naive. Of course, you can also lose faster. That's the same thing. But we are clearly in a position that we know what we do, and that's why we see it as an opportunity to grow faster. Now the localization, et cetera, question is a different question. That is also ongoing, which is why we have put in place a very decisive China for China strategy. Meaning 2 things. We are increasing our local manufacturing to serve local customers from local factories. That has become a distinct -- a really distinct competitive advantage of NXP over our typical peers, which is thanks to our hybrid manufacturing strategy. So we can do this because we work a lot with foundry partners, which means we don't have to fill factories in the West, but we can lean on local manufacturing partners. We -- in the Analyst Day in Boston a few weeks back, Andy, our operations leader, clearly showed you what it is. We work with TSMC on 16 and 28 nanometers in Nanjing in China. We work with SMIC, and we work with [ Grace ]. Those give us local support from a manufacturing perspective, which for our Chinese customers ticks the box of local content. So that's a very strong element. But there is another one, and that is about the competitiveness of the product, not from a cost perspective, but from a performance perspective. And here, the world has rapidly changed, and we do everything to stay ahead of this. The change is that in the past, we designed our products in automotive with and for Western customers. And then we would sell those same products into China. In the meantime, this is upside down because the leadership in EVs and SUVs is from Chinese companies. So we use them as lead customers. We design for them and then we eventually sell the same product into the West. So it's a complete change of where are we building our road maps and our performance focus on because their requirements are different. So when we speak to BYD, for example, I just had a call with the Chairman 2 weeks ago. They are now moving from leadership in EV, which they have achieved to leadership in SDV or they call it intelligent vehicles, which is exactly in the middle of our strategy. But that takes, of course, that we do silicon solutions for them and not treat them as a secondhand customer, which has to take what we've done with Western customers. So China for China strategy has 2 legs. One is local manufacturing, where we -- 30% or so is already in place. Secondly, is about treating a number of Chinese customers as lead customers and do what they need in order to differentiate against local and other competitors.

Makes sense. And if we -- you mentioned BYD, obviously, #1 now in the world in terms of EVs. We know that they have also a strategy to do things internally. Is it something happening as well for your products specifically that they want to do themselves what you provide today? Or how do you manage that risk? -- big customer, obviously?

Clearly, they have that internal silicon capability. I think they have a strong focus on discrete semiconductors around the battery also. We have not had this, and we have very open and transparent dialogue with them because that's where they want to work with us on where we can offer something they don't have. I mean that's the whole idea. So there is always a risk that something gets copied over time, but we haven't had that in our dealings with them so far.

You mentioned the pricing for next year. Maybe you can reiterate the pricing outlook. But the bigger picture is if you look at the industry, we are in an oversupply right now. And you have on top of that China localization with more aggressive pricing behavior, let's put it this way. How do you intend to secure good pricing? What's your strategy? And how do you see pricing in '25, but also beyond '25?

Yes. So we see for the next couple of years, and that includes then explicitly '25 we see that we are going back to what we had pre-COVID with low single-digit ASP erosion per year, but from the level where we are. So when I say pre-COVID, that doesn't mean we go back to the pricing levels in absolute terms, which we had pre-COVID. Since then, we have increased with, I think, 2%, 14%, 8% over 3 years. But we are now more on a curve which has then this low single-digit ASP erosion across NXP, which we can easily compensate from a competitiveness perspective, from a gross margin perspective with cost reduction and other efficiency measures, which we have. And that's explicitly, and I think now it's a good time to say this since we are early December already, that also holds for next year. So a low single-digit ASP erosion, which is nothing unusual. I mean that's what we had before. Now you mentioned China, you mentioned different markets. it's always a mix. It's not like every customer, every product is low single-digit ASP. I mean this is a mix across the whole portfolio geographically and product-wise of NXP. But we are very, very confident that across that mix, we keep the low single-digit ASP. Now how do we do this? Clearly, that includes walking away from parts of the portfolio which might commoditize. So we are not shy to not bid where we would only win or would only be able to win on price. That's clearly not our strategy. We've done this with certain low-end microcontrollers in the past. We've done it with analog ASICs at some point. So there's a constant portfolio process, which Bill and I do, where every new product which we design has a margin hurdle, which has to be higher than where we currently operate. So that makes sure that over time, we are grinding higher. But it also means that in the running business, -- if things fall too low, we will just get out of it. So the name of the game is very, very strong product differentiation, which allows us to command a fair price, a fair value for our product.

Makes sense. Now if you look at your Capital Markets Day, you guided for revenue growth of 6% to 10% in the next 3 years and with a gross margin of 57% to 63%. So maybe, Kurt, first, could you run through like what is the main driver of the 6% to 10%? If you can give me 3 products or 3 trends that will fuel that growth that you see would be very interesting. And maybe Bill, on the gross margin side from 57%, 63% is still a big range because like I said in the beginning, you have a low variation of your gross margin. So what brings you up to 63% and what brings you down to 57%? And how do you see that shaping?

Yes. So briefly on the revenue, we clearly differentiate between the segments. So the growth is led by automotive and industrial, which are both forecasted to grow with 8% to 12%. Mobile and communication infrastructure is much lower. So from a segment perspective, the focus, the strategic focus from a growth perspective is on automotive and industrial, which is 75% or so of the company anyway. So that's a strong part. So 8% to 12% for those 2. Within them, we are differentiating between what we call core business and growth business. The core business, and I very, very explicitly say core, not legacy because legacy always sounds like that stuff which goes out. Core is a business where we have very high relative market share, where growth above market is hard because we are the market. So we have a couple of those positions like in-vehicle networking transceivers in automotive or the keyless entry solutions in automotive. They will continue to be there, but they cannot grow much above market. So all of that core business growth with about 3%. And then we have the growth businesses, which grow on average with around 20%. And they represent currently just short of 30% of NXP. And through that next 3-year period, they grow to just short of 40% of NXP. So 40% of NXP by '27 is what will have grown with around 20%. So there was a strong differentiation between these high growth and the core business. Now within the high growth, I would clearly mention the software-defined vehicle, SDV, S32 in automotive, which is a plate of microprocessors and microcontrollers, which, by the way, has grown with 35% over the past 3 years. So a very strong track record, but also electrification and radar. These are all businesses which are around $1 billion or north of $1 billion. So solid big businesses growing really, really strong. But certainly, also our intelligent edge solutions in industrial clearly see an acceleration in growth going forward, where our main differentiator is in the focus on processing. Our main competitors and peers in analog -- in industrial are focusing on the analog side of the world. But with much more intelligence coming into building automation and factory automation, our portfolio, which is more biased to the processing side of the world, including microprocessors and microcontrollers, actually operates at a great advantage over those competitors. So that's another one which is then in the 20% kind of growth bracket. And Bill, maybe you'll speak about the margin.

Sure. In the short term, turning to gross margin, the 57% to 63% is a large range because a lot depends on the timing of revenue. Clearly, probably the next couple of quarters will probably be toward below the 60%, right, if you think about it because of the revenue levels. And what can change in any given quarter, as I mentioned, is the dynamics of mix in any given quarter. But more fundamentally, and Kurt alluded to this, is over the long term, it's about your fixed cost structure. And today, we're about 30%. And so you do the simple math, you can see and understand how revenue has an impact over your fixed cost structure and the fall-through. We think in 2027, we get to a fixed cost structure of about 25%. And then more longer term, we see an uplift with the 200 basis points we talked about VSMC. We have not sized the 200-millimeter factory consolidation yet, but that's a plus to get you above and comfortable above the 60% and plus-plus. And then lastly, I would say it would be, again, the continuation of that new product introduction, that mix that Kurt talked about as we continue to invest above the corporate gross margins today. And revenue takes about 3 or 4 years to come out for NXP and all our investments. And I'd say our fixed cost structure probably around 2030 will be around 20% is where we are aiming. So that should help the variability and reduce that variation in gross margins as we go ahead.

You mentioned EV penetration. Kurt, I mean what -- for '27, what's the percentage you have in your model in terms of penetration?

Yes. So I think this year, the latest I've seen from S&P was like 37%, 38% xEVs of global SAAR, which, by the way, is already big. And in China, by the way, it's 50%. So September, October, I saw that China was on 50% already. I think for 2030, we have something like 75%. So it just interpolate between the 2. I mean now we are not we are not heavily sensitive to how much hybrid within versus full EVs because we don't have that high voltage exposure that much. So for us, it is about processing and battery management. And there is a slight delta between the value of a battery management system for an 800-volt car versus something 400 or 240 volts, but it isn't all that great. So we have a little bit less sensitivity, which is why we are also not obsessed about what is the exact mix within that. But we see it marching towards this, say, 75% or so by 2030.

Bill, maybe on the free cash flow. I mean, you have targets at the Capital Markets Day, above 25% free cash flow to revenues and 100% excess return. So can you elaborate your model here? And how do you plan to do on the phasing side?

Yes. So if we just look back when we put our capital allocation strategy in place back in 2018, we have returned 109% of our free cash flow or around $19 billion. We're pretty disciplined here. And the way to think about it is when our net leverage ratio is below 2x, we may buy back the stock. But when we're above 2x, we definitely don't buy back the stock until we get back into model. So we use that as a perimeter of making sure we continue to return all excess free cash flow back to you all. From a cash standpoint on our balance sheet, we like to hold for the size we are in any given quarter, it's about $2.5 billion worth of cash. Free cash flow obviously has been punished a bit because of our higher inventory levels and some of our payment terms, which we are correcting and hope to get back in order in '25 and '26 to help us leverage to drive our free cash flow above that 25% over the next couple of years and continue to return that excess back to you all.

And this last minute, I wanted to ask on this trend around zonal domain architecture and the content opportunity for NXP because I think it's very interesting what you are doing here. And I'm going to be honest with you, I thought it was not real 5 years ago. I thought it was just kind of a marketing thing, the zonal architecture. But then I have a lot of trips with OEMs and they actually talk about this to investors this trend happening. So -- it's real, much more than I thought. What is your ISP opportunity here? And how is it driving NXP revenues as well?

Well, it's totally having, Francois. I mean we -- I've said we've grown 35% with our processing solutions over the past 3 years, which is now yielding a lot of domain and zonal processors, which are shipping. And I have to point out that most of them are actually microprocessors, not microcontrollers. It includes some microcontrollers, but the majority is microprocessors. And I say that because we have peers who talk about their success in microcontrollers, which comes at much lower price points. So our 35% growth is also a function of the higher price points. Price increases, but just higher-value products, which are selling at a higher rate. And maybe the best example for this for the next generation, which is then coming out in the next few years, is our 5-nanometer-based vehicle processor, as we call it, S32N. Jens talked at the Analyst Day about some price points here. I mean we have a $50 kind of price range for that to start with, but there is higher performance versions coming out with more software, which will sit north of $100 per piece. When you know that classic automotive microcontrollers are selling below $10, you see why we are very excited about the growth opportunity. So we don't need that many because the individual value is huge. So that definitely continues. I would just call out, as I said before, also here, China really is driving the pace, which is a great opportunity for us.

Great. On that note, thank you all for listening in, and have a nice day. Thank you.