NXP Semiconductors N.V. (NASDAQ:NXPI) NASDAQ Investor Conference Conference Call June 11, 2024 4:00 AM ET

Company Participants

Kurt Sievers – President & CEO
Bill Betz – CFO

Conference Call Participants

Blayne Curtis – Jefferies

Blayne Curtis

Go ahead and get started. And I’m Blayne Curtis, recently joined Jefferies, covering U.S. Semiconductor, Semi Cap Equipment. I’ve known NXP for a very long time. Very happy to have Kurt Sievers, the President and CEO; as well as Bill Betz, who’s the CFO. Welcome both of you. And I think having just initiated when I’m not talking about NVIDIA, the question is definitely analog. We’ve been very positive on that area. So I thought maybe I’d be just a way to start off, just kind of your perspective, where we are. Because I get that question asked all the time, where we are in the cycle? You definitely have had some comments. Every company is a bit different, but kind of where are you seeing it for NXP in the industry and then we’ll get into some more further questions.

Kurt Sievers

Yes, sure. Thanks, Blayne, and thanks for hosting us this morning here. I will give you a few, which is strongly biased to the automotive and industrial markets. So it’s not strictly analog. It’s more the market segment exposure to automotive and industrial, which is the lifeline for NXP.

And clearly, the end demand is not bad at all. It is just masked by inventory digestion. And the commentary I’d give you here is very NXP specific because, as we’ve discussed over and over, we have handled the whole inventory management through this cycle somewhat differently to probably most of our peers, which is we have avoided right from the start to build over-inventory in the channel. 50% of NXP’s business is going through distribution. And that is typically the trap you fall into, that you over-ship into distribution and then suffer badly from trying to build it down.

We have, coming out of the supply crisis where we had low distribution inventory, 1.6 months against our long-term target of 2.5 months. We did not follow the temptation once we were able to ship again to have enough supply to fill it up back to 2.5. We stayed down at the 1.6, which was initially a function of not being able to ship. And we’ve kept it for the last 8 or 10 quarters. And that has helped us tremendously on that half of our business to not have to digest over-inventory.

So I would call that a clean situation for us. Where we have some inventory challenge, is with direct customers. So that’s the other half of NXP’s business, where we, again, I think, did a little more differentiated than several of our peers, which is we did not enforce our NCNR orders. The whole industry had this concept in different variations of — in our case, full year NCNR orders, which were non-considerable for our customers. But in our case, as early quarter 1 of last year, customers came to us and said, “Oh, if we follow that supply commitment, then we will be over-inventoried at the end of calendar ’23. So already in the second quarter of last year, we started then to try and accommodate some of the relaxation they were asking for, such that we didn’t overfill that much. Yet, we are now left with some over-inventory in automotive. That’s the one place where we still have a bit of a challenge left. So that’s the direct customer side in automotive, which is about 60% of our automotive revenue, 40% is in the channel, 60% is direct. And we should be done with that around the end of the second quarter. This is also why in last earnings, we said second half of calendar year ’24 will grow over first half. which is not really a change of the macro. It is simply the fact that the over-inventory, which we are still digesting there should be done by that, and we are back to shipping to normal end demand, while we’ve been under shipping now for a while. And that seems to be looking reasonably well. I would say Q3 is probably the transition quarter. Q4, we see auto patterns really coming up. Q3, it will be customer by customer that we come to a clean sheet. We already have some fresh air of spring time in our noses because with 1 or 2 automotive Tier 1s, we are already clean. And we see how they then quickly came back to normal end demand, which is actually positive. You see how the mechanism works.

Inventory is digested, and we are back to normal. The whole thing is not a science plane because each of these Tier 1s and say, we have about 10 where we need to do this, they have different inventory targets. So it’s not like they all have the same target. We have a spectrum between 2 weeks.

So there was one very extreme one, which is really pushing it hard. They want to go all the way down to 2 weeks of NXP product inventory, up to something which we consider more reasonable 10 to 12 weeks, reasonable because that equals the manufacturing cycle time. So we feel that would be a good reflection of learning from the supply crisis to go to a 10 to 12 weeks’ inventory.

But again, we have a range of 2 to all the way up to 10 to 12. And they are also not consistent. I mean some of them agree with us in the beginning of the year, it should be 6 weeks, and then they give me a call last week and say, “Oh, but we really have a working capital issue for Q3. So can we not have 4?” Or we have a ramp in the second half; 6 was a bit tight, so can we have 8? So it’s kind of moving around, but if you put it all together, it feels like end of this quarter, entering into next quarter, we should be [indiscernible].

Question-and-Answer Session

Q – Blayne Curtis

Perfect. I do want to ask you about the actual auto markets. So you’re now shipping below demand. I think I get this a lot from investors. During the pandemic, I think people were cash flushed. They bought loaded cars because that’s all that was on the lots. So it’s kind of the question is just are people’s wallet smaller, are you worried about just sales of autos? And then in terms of the mix of cars, semi content is rising, but I think maybe there was a mix of certain features that maybe was overexaggerated during the pandemic? Kind of just your perspective on both of those?

Kurt Sievers

Yes. So clearly, the macro this year in auto is weaker than last year if you look at the total SAAR. Last year SAAR growth, I think, was 9% year-on-year. This year, it’s going to be flat. So you have already from that a 9% less strong market.

Secondly, pricing plays a role in all of this. We were quite transparent. We had last year as a company, 8% price increase. This year, we’re going to be flat. So this year — and that’s not only auto, it’s across the company. We’re going to be flat. So also there, you have a softer market, if you will. The content thing on premium versus volume cars that has already moderated before. I think the peak of this was in ’21, which was a maximization of profits for the OEMs.

So as you say, in times of shortage, they, of course, build those cars, which delivered the biggest margin, which were the premium cars. But from ’21 to today, that has already normalized. So we don’t see a big impact anymore this year. The other one is EVs, where I offer a somewhat more differentiated perspective than what media would consistently tell you. I mean there is a bit of a sentiment in the Western world that the EV party is over. We totally disagree with it.

EVs and hybrids together reached a 33% share of the global car production last year, which is material. I mean, 33% is really big. And there is a total consensus that this year, from that 33% penetration, they will grow another 20% in unit volume. So in a flat SAAR, that 1/3, which is electric, they grow 20% in units this year, which is massive growth. I mean, that’s very, very good because they carry a lot of semiconductor. The only little thing is most of that is in China. So that perspective of a moderation of EVs and kind of things are not good is, I believe, really more a U.S. perspective, where it’s true, but the U.S. don’t matter. They don’t move the needle for this because the volume is in China.

And Europe is somewhere in between. Bottom line, we continue to be quite positive about the EV penetration trend, which for us, the difference between hybrid and EV doesn’t change that much. So we are — we don’t suffer from maybe a bit of a snapback in hybrids versus EVs. So above all of this, I would say, the long-term growth trends for NXP, which are driven by ADAS, software-defined vehicle and EVs are fully intact. So the macro is okay from our perspective. It is more an inventory question, which is a short-term rationalization.

Blayne Curtis

I’m curious, being an analyst in the U.S., Chinese auto OEMs are not in the U.S., but marketing around now that have initiated. And you’re even seeing some in Canada, in the Europe, meeting investors, that’s actually a big thing where they see these cars. They’re actually quite feature-heavy and good price point. I’m kind of just curious in terms of the success of Chinese EVs globally. What’s your perspective? And then for NXP, your content in those cars versus global?

Kurt Sievers

Yes. So on the latter, we are very hedged. So we have an equal share in China as we would have in Europe and the U.S., which means if — and that’s actually the case, if China EVs are winning over European or U.S. EVs, it’s good for us. So we don’t lose from this on the contrary. I would judge them actually as very capable, Blayne. And that’s not only cost leadership.

I mean, often, the messaging is reduced to, it is subsidized and there is a cost leadership. I do believe they have really mastered 2 things very well: one is speed of development. They develop a new car platform in 2.5 years, where the Western companies traditionally have taken 4 to 5 years. And in a world where it’s all about electronics innovation, which, as we know, is very fast. If you do it in 2.5 years instead of 4 to 5 years, you are just ahead of curve in actually bringing new electronics experience to us as consumers.

And I think they’ve done this masterfully so far. And secondly, many of them didn’t have the legacy of combustion engines, which is just a way to right away design a car from scratch as a realtime computer on wheels. That’s how I would call it, which makes you just much more agile than when you try to convert your history into something which is a bit of a hybrid mix into the future.

So I think very capable. It is more a political question, Blayne, where they are allowed to sell it, they are not? Under which taxes? You know that also Europe has a discussion now about this. I fundamentally believe they will be successful because in the end, I believe that economic forces will prevail because they have a very competitive offering.

So the consequence and the implication for NXP is that we are very focused on them. So, so far, we have joined that success, which is good. But we also want to stay ahead of this because we think they will push the envelope globally. So we think if we are able to be successful in China with EVs and SDVs, then we will be able to do that globally because that’s the leading force now.

Blayne Curtis

Right. And I guess the natural follow-on, and I get this is probably the #1 pushback to the whole analog group is the concept of Chinese localization. So it’s an auto question, but actually broader as well. It does focus, you’re not in the silicon carbide market, which is one hot area of it. But there are areas, microcontrollers and such, we you might see — what are you seeing in terms of the Chinese companies?

Kurt Sievers

Yes. So I offer you a few perspectives here. First is currently, and that’s not a naive statement. It’s just the current status quo. The future is very different than I speak about the future.

Currently, indeed with our portfolio, we have virtually no local competition. Because we see them focusing on power discretes, which include silicon carbide, as you say; low-end microcontrollers and catalog analog. None of these 3 are really in our world, but that’s a snapshot of today.

So today, we compete mainly with Western companies in China. Going forward, I’m sure that’s going to change. I mean they work hard to change that. Our way to deal with it is twofold. Number one is we have a firm customer requirement across the board of industrial and automotive customers in China to localize manufacturing. So they clearly want us to both back-end and front-end manufacture the chips, which we sell to them, which is easy for us on the back end because our largest back-end test and assembly site is anyway in China. It has always been there. It’s in Tianjin, which is an hour from Beijing. 4,000 people factory, which used to serve the world. Now we are reconfiguring that factory to become a China-for-China factory.

For the front end, it’s a little more complicated, obviously. But given that we are largely a foundry-based company, we are quite flexible. So I think we are currently building a big competitive advantage over IDM competitors, who have in their own countries a lot of assets. So they cannot just swap to China, we can. What we do is we use the TSMC Nanjing, 16, 28-nanometer factory in China, which is a copy exact of their other 16-nanometer factories. And it’s good to know that our workhorse microprocessors for automotive are all in 16-FinFET. So that’s a very heavy move which directly gets us into China manufacturing.

Secondly, we’ve worked for many years with SMIC outside of automotive, that’s more in the industrial and mobile space. And we will continue to leverage that relationship. And we are in the process of qualifying a analog mixed signal, a local China factory. So the typical mature node analog-mixed signal, which is more complicated.

But we also need that to have the full portfolio covered from a local — for local manufacturing perspective. I do believe that for the next couple of years that makes us sufficiently Chinese and gives us also access to cost. By using local foundries, we tap into the same cost competitive environment, which the local companies would do. However, mid to long term, we clearly believe that the most sustainable and structural differentiator which we have to build is product performance.

Only if we have an offering to, say, a BYD or Xpeng or the likes, offering which they absolutely need and cannot get locally, only then they will use us. I mean, let’s not be naive. They — of course, they will prefer a local company if that local company offers the same we have.

So we must avoid the commodity trap. And what that takes from a cultural perspective in NXP is, and I work hard on our 12,000 engineers to get there, to change our mindset. We have always used Western companies as lead customers. So our products were defined along the needs of Western car companies or Tier 1 suppliers. And we have — we are now taking a portion of our R&D, which has to be used for Chinese lead customers because only then I feel we can guarantee that we also make a product they absolutely need and not just a copy of what we had in the West since they are so much leading.

We are in the middle of that process. And with that, I’m actually quite confident that we can successfully compete in China. All of this is, of course, excluding any government regulations, including export control from the U.S. government, which is beyond our control. I mean if there is anything, of course, we have to comply, and that could be a stop very quickly.

Maybe last point to size this. NXP has 30% of its revenue into China, and about half of that is China for China. So anything I said now in the last few minutes relates to about 15% of NXP’s total revenue. So it is relevant.

Blayne Curtis

And everyone ask you, we’re going to run out of time. So in terms of growth, there’s a lot to talk about. But the growth drivers within auto. I mean you’ve done a good job highlighting these at the Analyst Day, RADAR, domain zonal controllers. So people might not fully understand what that is, but it was a big $1 billion-plus opportunity for yourself. And maybe you could just update us in terms of the progress there on the multiple growth drivers in auto?

Kurt Sievers

Yes, for all — the third one is battery management solutions and inverter control. All 3 are on track or above targets. The most sizable one currently is RADAR. So we think about $1 billion revenue this year, and we hit the growth target of 20% to 25% over the last 3 years.

Battery management is ahead of target, which is simply a function of the faster EV penetration than we had anticipated. So here, we were too conservative in our assumptions, so that’s good. And we have a large battery management exposure to China.

You know our biggest competitor, which is actually larger than us, is I’d say we have an advantage over them in terms of system solutions because we offer microcontrollers and high-performance analog front ends for the battery cells, and China won system solutions. So that gave us an advantage. In China, we won share. And since China is the fastest-growing in electric vehicles, that is a great — that was a great run over the past couple of years.

Now the one which you said people understand least is going to be the most impactful for the future, which is the automotive compute architecture. It’s our S32 CoreRide Platform. Maybe the easiest way to explain this, because we are all consumers, is the following: it is about enabling the software-defined vehicle.

And the software-defined vehicle, what it will do to consumers is that when you buy today a car, the tragic thing is that the car, at the moment you buy it, you drive it off the dealer lot is actually the best moment in terms of performance. It will only degrade from there. It gets worse because it’s not updatable. And the whole disruptive change in how to think about cars with software-defined vehicles is that the moment you buy the car is the worst, and you will be able to upgrade the performance of the cars through software over the next 5 to 7 years.

I mean we are all used to this on smartphones and other devices, but it hasn’t happened in the car. And NXP is the leading company to work now very closely with OEMs. It’s not a Tier 1 thing, it’s an OEM discussion to offer a compute architecture, which allows to do this. And that is a portfolio of processes ranging from a 5-nanometer vehicle computer, which is a $4 billion transistor machine, which I dare to say is the only real-time capable 5-nanometer processor in the world. So we’ve really gone here ahead of anybody.

But it goes then through, as you said, domain computers, sonar computers, all the way to edge nodes, which might be in 40 nanometers. All of them software compatible. I mean the whole trick is to have this whole portfolio, which is under 1 software regime, and we help OEMs to co-architect their new vehicles.

That one is when we have Investor Day in November 7, probably going to take center stage in terms of what’s the revenue impact of this go forward. Maybe one last data point on this. Think about this vehicle computer, this 5-nanometer device, which we are now sampling with customers. It will sell at around $50, which is a 10x factor over the classic microcontrollers in automotive, which are around $5 — $4, — $3, $4, $5. So we have a significant value increase here just by going to much higher performance levels in compute.

Blayne Curtis

This goes for either of you or Bill, but just in terms of going back to the cycle, you navigated the cycle , I would say, better than anybody in terms of you’re not really even down, most are down 20%, 30%. So you mentioned how you kept the supply chain lean. I’m kind of curious in terms of your view coming out of it, you’ve talked about adding back to the disti channel. What are you looking for? I mean, it sounds like you’re through it. You didn’t even see a decline, and you’re talking about hopefully auto is cleaned up by Q2. So what else are you looking for, for the rest of the year to kind of add back that inventory industries?

Kurt Sievers

Well, to be specific, first of all, we indeed said that there is one we need to mitigate with this inventory strategy. When the market comes back, we have to be well represented on the shelves of our distribution partners. If then we have less product than they are sitting than our competitors, there might be a tendency that we don’t sell as much, which is why we have agreed to carefully stage back.

We want to go to 1.7 months, from 1.6 last quarter to 1.7 this quarter. And then in the second half, we might move up a few more digits, not all the way to 2.5. We just don’t see an environment which would allow this, but a little bit higher than 1.7.

The signals we are waiting for is just a consistently growing sell-through. It’s very simple. And a large part of that has to come from China because our distribution exposure is strongly biased to China. And it includes very much the industrial and IoT market. So we have industrial and IoT market for us is 80% served through distribution. So I’d say that is probably the biggest metric which we look at is the sell-through consistently coming up and not just incidentally, and then we would refill.

Now there might be a moment it becomes actually hard to refill, I’ve seen this before because when the sell-through is very strong, you refill and it sells — it just sells through more. So that’s why I cannot say how fast, when, but that is the direction plan, absolutely. And of course, it’s going to help Bill and I to build on DIO because we have a somewhat elevated internal inventory that will be built down and then we can also slowly start to ramp up our factory again from the 70% loading level.

Blayne Curtis

Actually, a perfect leading because I was going to ask Bill on the gross margin. So I mean, you are trying to bring down your own inventories, but you’re actually starting to see your volumes take off. In terms of the gross margin for the rest of the year, are you going to be able to show some improvement? I mean, obviously, overall volumes help, but kind of the counterbalancing is to bringing the inventories.

Bill Betz

Yes, Blayne, I think the way to think about our gross margin over the last couple of years, we’re able to stay at the high end of the model based on how we’re navigating through the cycle. As Kurt alluded to, clearly, our internal utilizations are in the low 70s. So there will be a point in time when we’ll start to bring that back up to the sweet spot toward the 85% and get some recoveries there for gross margin.

Another aspect is, as Kurt alluded to again, is we’re at 1.7. As we refill that channel, this revenue of about $500 million over the next several quarters has accretive margin to the corporate average because we’re selling to many low or — low volume to but many high customers, thousands and thousands of customers related to it.

So you have that lever with the distribution. You have the lever with the utilization. Then you have also, I would say, just higher revenues, higher revenues over a fixed cost structure of roughly, call it, 30% falls through related to it.

And then last week, where it’s probably more strategic gross margin levers that we talked about. Last week, as you saw in the press, we continued to build upon our hybrid manufacturing strategy. Nanofab Lite, non-IDM model, but a hybrid manufacturing strategy, where basically, we focus on the supply we need. We share the risk and instead of doing a complete factory by ourselves that is worth about $7.8 billion, we decided to work closely with TSMC and partner with their sister factory called Vanguard to create the SMC.

And our portion of that investment is roughly $2.8 billion of the $7.8 billion. Again, $1.6 billion is related to, call it, the equity portion, and that equity portion gives us basically 2 aspects: one, it gives us access to flexible supply. It generates for us a capability of once filled an incremental $4 billion of revenue on an annual basis to NXP. On top of that, from a margin perspective, we now move from 200-millimeter to 300 millimeter, so we gain the geometry scale, as well as getting competitive costs as being both an owner as well as preventing the margin stacking when we buy from this joint venture. And that impact, once we’re full, call it, in 2028 or 2029, will add another 200 basis points on top of total NXP today.

So just assume if this factory was running tomorrow instead of our gross margins being 58%, they would be 60%. So very strategic for us as we continue to think more longer term. Now what we haven’t figured out yet is, as you all know, we have factories in the U.S. and Europe that eventually will become uncompetitive. And so in our press release, we talked about a Phase II. And over time, if you had to think about 10 to 15 years out in the horizon, we will eventually consolidate our legacy factories into this joint venture, which will give us additional margin accretion in our portfolio and make sure we’re competitive in the landscape that we play in.

So we’re super excited about more really transforming the company more longer term. And as I indicated many times before, 58% is not a final destination. Clearly, we have an Analyst Day on November 7 in Boston, and we’ll update the financial model there.

And going back to the joint venture, there are more benefits as well, again, our portion is $2.8 billion. Vanguard, including their customer access side of the house. Basically, it’s $3.1 billion. You add that together, you get $5.9 billion. And then the rest is obviously coming from Singapore as they have their own somewhat Chips Act and favorable government lending into — inside the factory, of course.

But overall, I think that takes care of this and focuses on where we were constrained. We have a lot of design wins ramping in this time field. And the most strategic reason why we did this is to make sure we can lock in that supply and deliver to our customers of what we want. And so that really starts to peak in ’28, ’29 related to for that $4 billion incremental revenue for us.

Blayne Curtis

With that, we’re out of time. Went by quick. Thank you.

Kurt Sievers

Blayne, thank you very much. Thank you.

Bill Betz

Thank you.