NXP Semiconductors N.V. (NASDAQ:NXPI) Credit Suisse 26th Annual Technology Conference Call November 29, 2022 10:15 AM ET
Company Participants
Kurt Sievers – Chief Executive Officer
Bill Betz – Chief Financial Officer
Conference Call Participants
Christopher Caso – Credit Suisse
Christopher Caso
All right. I think we’re live. All right. Welcome, everyone. So I’m Chris Caso. I’m the new semiconductor analyst at Credit Suisse. So I’m really excited to be here at this conference. Attendance I’m told is the best that we’ve had in 10 years. So people are excited to get out and about post-COVID and find that out of things. And obviously, with a lot of turbulence in the market we have a lot to speak about. So for our first semiconductor presentation today, pleased introduce NXP Semiconductors.
With us from NXP is Kurt Sievers, CEO; and Bill Betz, the CFO; and Jeff Palmer, VP of Investor Relations is with us as well. So gentlemen, thanks for attending today.
Kurt Sievers
Yes. Thanks much, Chris and thanks for having us. Indeed, it looks like a very well attended event. I’m curious for how you run the discussion.
Christopher Caso
Well, let’s get started and I’ll — Jeff told me to be gentle. So I will do my best. But I would be remiss to not talk about some of the things that are on investors’ minds, obviously. And Kurt, one of the things that — discussions we’ve had for much of last year and probably longer than that, has been — it’s always been curious to you, I think, is that you sit in front of investors and handle questions about the potential for a slowdown and when is this going to end? And usually immediately after those discussions, you get on the phone of your customers and they’ve been yelling at you for more products.
And I guess maybe that’s a way of starting. And the macro conditions have certainly changed over the past several months. But the supply-demand dynamics in your space have not — there’s been some areas of slowdown and some not. So how do you reconcile that? And I guess the question that all investors are asking is, is it just a question of when that things are going to slow down for you?
Kurt Sievers
Yes. No, certainly, it has changed. I mean the whole macro uncertainty is not leaving us alone. I mean, we clearly see, I would actually say, a significant weakness in the consumer exports business which is our IoT business, inside 40% of the industrial IoT segment. And also the low, mid and mobile which is largely the Android business.
Now that’s not new. We’ve had this now for a while. And I think we are treating it with a lot of care which means we don’t want to jam the channel and we try to learn actually from what we have done wrong in the past in similar situations. However, at the same time, the automotive and core industrial indeed continues to be very, very resilient. So I continue to have these calls. They don’t yell at me, by the way. I mean they’ve all learned that being constructive is the better way to deal with the situation. Since there is a lot more we will jointly do in future. But the matter of the fact is that in core microcontroller product categories as well as analog mixed signal categories, we are sold out into automotive and core industrial. From an order visibility perspective into next year, we have these famous NCNR orders. That also continues to be the case, obviously, for next year.
So we have about 85% coverage of next year’s NCNR order load from those 2 markets on NXP. Now is it just a question of when it comes down? I don’t think so. Because I think the demand drivers in automotive and industrial are structurally different ones to the consumer businesses. The consumer businesses have really benefited from the surge in demand through this working from home which came with the pandemic.
While that demand in automotive and in core industrial had nothing to do with the pandemic. I mean that is a structural secular demand which has to do with electrification and all the good content increase trends which is why for sure, a weaker macro is kind of broadly bringing down the number of cars which are being sold. But we believe that the content increase is so much dumping any potential issue on the SAAR that we are quite confident on next year’s numbers to be in a good place.
Christopher Caso
Right. And one of the things that’s always sort of a hallmark of a downturn in semiconductors is inventory where customers tend to build too much inventory. And I think that has happened to some extent in the consumer space. But as we’re saying earlier, your customers have been clamoring for product for much of the last year. And is the potential that just because of the shortages that happened and the lack of industry capacity that happened during the pandemic that just in certain sectors such as industrial and automotive, those customers weren’t able to build the accesses that have been in prior cycles.
Kurt Sievers
Yes. So fully agree with you. In the consumer segment, we try to avoid the inventory build which is why indeed, we limit our channel inventory to 1.6 months which is almost a full month below the 2.5 target which we have. And mind you, now it is us limiting it. I mean in the past 2 years; we couldn’t bring it up because we didn’t have enough supply. Now we actually keep it down in order to not jam the channel.
On the auto and industrial side, you never have ultimate visibility of customers’ inventory. I mean I would be lying if I said, we know exactly how much is there. But it is very obvious that it is still very, very thin. I would go that far that the supply chain is still dysfunctional which means any little change in say, we ship a few days later. We have a test on maintenance in our — in one of our factories or there is a FedEx issue in transportation. We immediately have the calls that a car factory is down which tells me that there is nothing in between. I mean it’s just empty. So I think there are 2 stages, Chris. The one stage is to make that supply chain functional again. And that means all the different stages between us and the car company are filled to a level which is enough to buffer any variations in supply.
Second stage will be and I think that’s what you were referring to, is the target or the learning from this period of supply shortages to build some buffer stocks. More than actually they had in the past. I am very, very sure. The second one is nowhere near. Nobody has been able to do this because there hasn’t been enough supply. Everybody is still fighting to get to the first stage which make — which is about making the supply chain per se functional, bringing it back to fill levels like we had before this whole turmoil.
Christopher Caso
Right. So right now, your customers still — the hand to mouth in that industrial and auto space, that’s still where you think we are.
Kurt Sievers
Yes.
Christopher Caso
Interesting. What about…
Kurt Sievers
Which is, by the way, a very — it’s a very interesting situation for us because we — the past 2 years, everything was short. I mean that was pretty simple. It was just everything short and we tried to produce and ship as much as we could, any place. Now it is, of course, much more differentiated because indeed, the — those consumer markets don’t have that anymore. While for us, the majority of our business being auto and core industrial, they absolutely still have it. And the trouble is, it is not fungible with each other because you might be tempted to say, yes, what the heck, I mean, if you have enough consumer IoT, why can’t you shift this and help the automotive side of the house and the core industrial side. The trouble is, it is largely not 100% but it’s largely different process technologies which we can swap.
Christopher Caso
Right. And that was going to lead to my next question which is the difference in the supply demand dynamic for the different processes. And whereas some of the leading-edge stuff or maybe kind of the trailing edge or the leading edge is freeing up because of consumer. But we’ve heard from your peers is sort of 55, 65 nanometers even some of the older geometries are still pretty tight.
Kurt Sievers
Yes. It’s really, I’d say, between 28 and all the way up to 180 nanometers. There is still a lot of stuff going in 180 which is not old products, by the way, it is specialty processes with very high voltage capabilities, et cetera. And that whole bracket probably really with a specifically concerning field 55, 65 and 90. That is indeed what keeps us busy. And you know that a lot of this we are sourcing from foundry partners which I think gives us very transparent forecast for their supply capability through all of next year. And that’s why it is what I said in the beginning, from anything we can see today, that demand which we have on hand will continue to surpass supply capability also through next year in those particular nodes. There is other things which are fine. But those are the majority of the auto and core industrial supply.
Christopher Caso
And what’s being done to add capacity in those nodes? Because, obviously, it’s been short for some while. The availability of process tools has also been constrained. So again, another potential difference in this cycle is that some of the excess capacity that the industry would have or had adding prior cycles has been difficult to do so. So maybe you could speak to that.
Kurt Sievers
Yes. So tactically, indeed, unfortunately, I have to totally confirm what you say. Bill will tell you that we thought we would move to 10% CapEx in Q4 which we don’t because we don’t get the tools. I mean, we wanted to spend it but there is a — also that part of the supply chain is actually so much in delay that we couldn’t even make the CapEx operation which we wanted to. How is our capacity being added? For us, it is a 2-leg approach. We have our internal manufacturing facilities which cover around 40% of our wafer supplies and 60% are with third-party foundries. We maximize for anything which is NXP proprietary technologies, our internal footprint. I mean that’s what we spent the CapEx for. So we don’t do the overflow model anymore but we only run internally what only we can run to maximize the output for that. That’s in full swing and it will deliver gradually every quarter more supply.
On the other hand, we, of course, work with the foundry partners to get more and more supply from them. They are adding in their world. And all of that has helped us through the past — yes, it’s 10 quarters now, every quarter more and we see this also continuing through next year. So when I say next year, our supply capability is just maybe 85% of the demand. That is on a higher level than what we had this year. So I mean, it’s not like we are stuck on the same level. It keeps going up. But it’s just that the demand is also higher. What doesn’t work and that has led to some confusion is there are those Tier 2 foundries. And I think we’ve all heard that Tier 2 foundries, especially in China, are actually underloaded.
Bill Betz
Correct.
Kurt Sievers
And they are operating in these mature nodes. But first of all, we have no real appetite to source from China, given the geopolitical situations. But secondly, it is also that our — especially automotive and core industrial customers do not accept those foundries anyway. So we are, from a safety, reliability and quality perspective, we have to work with the Tier 1 foundries. So it is traumatic [ph]. But while there seems to be overcapacity in those foundries in China, we still continue to be short because we can tap into that capacity.
Christopher Caso
You can tap it. Yes. I mean there’s some bifurcation between the capacity. Maybe shift a bit and speak to some of the product cycles and some of your product-specific things. But — and I think what was — one of the things I noticed from the last Analyst Day was that your growth drivers, we speak about auto, it’s not all of auto that NXP seems to be very targeted in certain areas. Things like Radar, some of the domain controllers and things like that.
Before I get into some of those things, maybe you could speak to how you choose those markets, kind of the project selection process which gets you into some of these markets where what you used to say in the past 1.5x market share of your competitors?
Kurt Sievers
That’s exactly the key. So we have — and when I say we, it really starts with Bill and I. We have a very, very stringent focus in the company to only operate in segments where we can achieve a relative market share of at least 1.5. The whole idea behind that is that we believe by doing that, we can out-innovate competitors through scale. I mean, if you spend 16% R&D but have 1.5x more revenue in absolute terms than your competitor, you will spend more on R&D and you can actually come with a better roadmap and a better product proposition to your customers.
Now not the whole company is at that stage. I mean we constantly have to renew and we add new segments and we grow into them. I think at the moment, it’s like 60-plus percent of all of the company’s business is actually in this relative market share leadership position already. And that leads then indeed into 2 things. It is an enormous focus to be the best in certain areas. But it also means that Bill and I say no to certain things. I mean if the teams cannot come with a plan where they credibly achieve that position over a period of time, we will either immediately stop or we do it for a year and see if we get closer and stop them. So it’s a tough process. But I think for this ever need of scale in our industry, that’s really what makes us successful.
Now bringing that back to product examples, I mean, this is how we got to the leadership position in mobile wallet, for example, in mobile, where we — I think we have a relative market share of 7 or 8 or something. So really, really strong leadership. It’s similar in Radar, in Automotive, where the 3-year growth is guided to $1.1 billion in 2024 which is 22 — I think, 25% CAGR over the period in a large and growing business and we made it now to number one. And I think with the design wins which we have in-house already which we just have to execute on now. We will even expand that number one position going forward.
But you also have positions where we don’t have it yet. That is in battery management for the electrification for the car is a good example. I mean, we are actually behind ADI, they — after they acquired Maxim, they actually got into a strong number one position. But I believe with our value proposition in battery management which is full system solutions, including automotive ASIL-D microcontrollers, we are design winning faster and work towards that relative market share leadership position.
So you see we have that in all stages. Mobile wallet, we are far ahead already. In Radar which is a fast emerging segment, we are right in the middle of just number one now. And now expanding to the number two. In Battery Management, we come from behind but I think the dynamic, the momentum we have is also going to boost us ahead. And those are what we call growth businesses which were like $3 billion revenue for the company in 2021. And we’re going to double them to $6 billion in 2024.
So when you think about the growth of NXP, as you said, 8% to 12% which, by the way, is twice the speed we had before. I mean we came from more like 5% or so in the 3 years before. Now we are doubling this to 8% to 12%. It’s really made up of 2 elements. It is those growth businesses which are more like 20% to 25% CAGR, $3 billion to $6 billion and a core business which we think is going to grow with market, say 5%.
Why is that core business only growing with market? Because this is where we have these high RMS positions and we can’t outgrow the market. We are literally the market in those positions. But — so it’s really those 2 elements, those 2 legs which are driving NXP’s growth.
Christopher Caso
Right. And how do you segment out? I mean, in between the 2, the sort of higher RMS sort of baseline business and new businesses? In the past, you used to say something like 75% of your business would grow at SAAR plus a few points. And then there was a 25% of the business which is higher than that. Does that kind of framework still apply right now?
Kurt Sievers
Well, it has tilted because the relative share of those pieces, especially in Automotive which are growing much faster, has actually become bigger. I think the 75%, 25% doesn’t fit anymore. It’s — that portion has got bigger which is also why we have — we guided automotive 9% to 14% which is also faster than we had guided in the past. And admittedly and now I know I opened a big box of discussion but all of those 8% to 12% for the company or 9% to 14% for automotive were not comprehending price. Now we all know that we are in a, I’d say, at least somewhat unusual period for the semiconductor industry where price increases. And I actually think consistent price increases are the name of the game.
So the reality is when we — when Bill and I, when we gave that guidance in November last year, we didn’t comprehend price increases because we didn’t know how much we could actually do. And we also didn’t really comprehend much movements of the SAAR. So those elements are kind of modulating now positively, I have to say, where we are currently going.
Christopher Caso
Right. And as you speak of price increases, this is about the time of the year where at least your U.S. auto customers will typically negotiate for next year. In addition, we’ve heard from some of the foundries is that, particularly, on some of the lagging edge processes that the price increases are starting to be implemented for next year. Given all that we’ve seen and the changes in the macro, is that still likely to continue? And maybe you can speak to us about the conversation you’re having with your customers?
Kurt Sievers
Yes. So both is true. The one is — it all starts with the imbalance between demand and supply which I spoke to earlier, in parts of our business. And that has to do with this trailing edge problem. And by the way, I would say it’s just that this industry, we all collectively have underinvested in the years maybe between 2016 and 2020 or so which is why we have that whole situation now. And somebody has to pay for it because there’s a lot of depreciation of all that capacity which is being added. So yes, the annual price negotiations with the car customers are happening as we speak. I mean this is always this time of the year.
In Japan, it’s actually shifted by a quarter. They do this in April. But fundamentally, there is this annual once-in-a-year price negotiation. Secondly, yes, the — those nodes I spoke about before which continue to be chronically short like 55, 65, 90. They continue to see from our vantage point, cost increases. So our input cost with our foundry partners are going up next year for those nodes. So we have no other choice. It’s not that I particularly like this but we have no other choice than passing on that increased input cost to our customers in such a way that we actually protect our gross profit percentage.
We’ve done this, I’d say, successfully over the past period. And I see no escape from continuing to do this also into next year because, again, the cost continues to go up. We will not let loose on our gross profit performance. So we are increasing prices as we speak.
Christopher Caso
Right. And although the gross margin percentage hasn’t been changing as a result of that, certainly, gross profit dollars have benefited as you raised price on customers, you keep the same percentage, the gross profit dollars go up. Bill, maybe I can ask you, does that lead to some degree of resiliency in the margins as you go into next year, both because of, one, the pricing environment? And then secondly, just kind of these tight supply conditions, what sort of levers do you have to pull if we use these some more incremental macro weakness as you go into next year?
Bill Betz
Sure. On the pricing, it really doesn’t impact our gross margins because, as Kurt mentioned, we’re just offsetting the higher input costs. Now what’s a bit different is internally, what we manufacture is greater than 90 nanometers very specific IP proprietary mixed-signal type of technologies. And we see these utilizations staying in the high 90s. They’re not coming back down.
If you recall, when COVID hit, we brought utilization for these factories well below 50% as well as the company’s revenue — overall revenue structure of the revenue fall through, the company has been really built when we merge with Freescale. From a scale standpoint, of the fixed cost is really above $10 billion. So we feel very comfortable that we will stay in our gross margin long-term Analyst Day model that we provide between 55% and 58% regardless of what the revenue is going to do next year.
Kurt Sievers
And if I may add, Bill?
Bill Betz
Yes.
Kurt Sievers
This is a big difference to the past. And it comes back to how we are retooling our internal factories. Our internal factories used to be exposed to kind of all the markets which we serve which is why we had a pretty significant swing in gross profit when demand came down. Now because of the supply situation, we have actually maximized these factories for these proprietary technologies which are almost exclusively for automotive and industrial which means our internal loading or utilization is, to a very large extent, a function of the automotive and core industrial market. And since that continues to be very resilient, Bill says what he says which is that we feel quite protected next year from a loading perspective which is very different to how it went in the past.
Bill Betz
Yes. Over 80% of our internal load is linked to the industrial and auto market.
Christopher Caso
Right. We’ve got about 5 minutes left. Is there any questions from the audience? I’ll let you guys think about it if you do. But just bringing — following up on that one as well. One of the things is also different in the industry now versus the past is that, for some of those foundry agreements, there are long-term agreements that are in place. And we’ve seen with GlobalFoundries, for example and some of the handset that they’ve been in force. That was one of the questions when things slow down, where they actually be enforceable it looks like they are. How has that applied to NXP. And in some of the areas that have slowed down, there are long-term agreements in place, I’m sure for you. How have you been able to handle that?
Kurt Sievers
Yes, I can. We have like $3.8 billion or so of long-term agreements sitting there but that’s over multiple years. I think it’s 5 years or something. So it’s a really long period of time. So when you calculate that back to revenue for the year, it’s actually very little. So I can very straight and very clearly say, we absolutely face no issue with this at this particular moment. And that is largely because most of those agreements were actually for automotive. And there is no issue on demand. Actually, I wish we had more of these agreements to get more supply. So it’s — we have the opposite problem. I mean, I wish we had a bit more because then we could actually supply more into automotive.
Christopher Caso
Right. I’ll check again if there are any questions, otherwise — make sure it’s hard to see. Just pivot on to maybe cash flow as well. And what you’ve talked about is about a 25% free cash flow margin. With the plans that you have going forward, how do you manage that, especially given the uncertain environment? And obviously, your CapEx, for example, is not as high as you would have liked which is benefiting cash flow. As we go into 2023, is there any impact on cash flow as you sort of catch up on some of these capital investments that you wanted to make in 2022?
Bill Betz
Yes, Chris, let me take that one. This year, we’ll spend about 8%. As you mentioned, we wanted to spend $10 million. Next year, we’re projected to spend between 6% and 8%. And we’ll be able to continue to drive to our long-term target of 25% of excess free cash flow that we want to go drive for the company related to it. So it’s balancing it, all the needs. And then our return policy, obviously, our number one priority is to continue to invest in the business. That’s the best use of our cash to grow the top line in a profitable manner. Then we’re going to continue to do these small tuck-ins and which we’ll demonstrate them. We disclose depending on the size but we continue to go after talent, small teams of talent around the world. Specifically, in AI software capabilities, low power and so forth.
So that’s the next step of our capital allocation policy. As you could see, our dividends, our dividends are running on a trailing 12 month around 21% of cash flow from operations. So I think over time, we want to get that to about 25%. So we have some room there. And clearly, the buyback in Q2, we were able to put some extra cash on the balance sheet just to prudent thing to go do. And then you saw we continued to buy back in Q3 and Q4 recently. So no change in our capital allocation policy and we feel very comfortable with our balance sheet and cash flow generation of the company.
Christopher Caso
Randy, sure. We can’t — the mic wasn’t on.
Unidentified Analyst
Yes. Question on inventory. I think you build a little bit, you’re at about 100 days. Where do you target? And do you think you’re going to have success building inventory? And then if you could talk about customers, where do you see inventory levels? It looks like balance sheets are quite high. But are you getting feedback of any inventory overhang customers wanting to work down?
Bill Betz
Yes. For us, our inventory is still quite low if you think about it. If we’re controlling the channel at 1.6 to make sure that we don’t pull it all to 2.4 overnight, that’s something that we looked at very carefully. If the demand is not selling through, we’re not going to ship it in and we watch this on a weekly basis and it gives us opportunities to potentially redirect this material or allows us to actually us, NXP to rebuild our value bank, to be able to give us more options for that future growth, specifically industrial and auto spaces. So we’re going to be very careful. We’re going to be comfortable of running a little bit more inventory on the balance sheet. And to be honest with you, compared to our peers, we’re probably one of the lowest.
Kurt Sievers
Yes. And let me add on the customer side. There are pockets of certain products which tend to be high for a certain period of time which is this golden screw problem where one part was missing from somewhere in the industry. Everything else is piling up and then it’s flushed through at the moment. But when you overall look at, say, car company or Tier 1 inventory which appears high, then you really have to discount this for the higher prices. I mean inflation sits on this quite massively. Secondly, the largest part of the value of the inventory of our customers is not semiconductors because they do high-value goods with steel and plastics and what have you which is all much more valuable than semiconductor. So it doesn’t say much — if you look at their DIO, doesn’t really say much about the semiconductor position. And I can only say in the auto space and also in the core industrial space, so with customers like Schneider, ABB, Honeywell, et cetera, we are not worried at all about inventory at this stage. I mean, I don’t have the crystal ball. I don’t know what this is going to look like in 4 quarters. But at this point, absolutely not.
Christopher Caso
Great. We’re going to have to leave it there. But gentlemen, thanks for your comments.
Kurt Sievers
Thanks, Chris.
Bill Betz
Thank you.
Christopher Caso
Thanks, everyone.
Question-and-Answer Session
End of Q&A
Be the first to comment