Great. Welcome back, everybody. I'm Joe Moore from Morgan Stanley Semiconductors. Very happy to have the executive team from onsemi, Hassane El-Khoury and Thad Trent, CEO and CFO respectively. Just quickly, I have to read this disclosure. For important disclosures, please see the Morgan Stanley Research Disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. Guys, thank you for doing this. I appreciate it. If we could just start off kind of general. If you could talk about the dynamics you're seeing in your end markets. You've talked about strong automotive, you know, some choppier trends elsewhere. Can you just kinda talk to what you're seeing in the market right now?
Yeah. Look, nothing different from our last earnings call, where we still see automotive holding up. Obviously, that's expected given a lot of the content, a lot of the megatrends that are going into automotive as far as electrification, content growth, whether it's Level 2+ autonomy or EVs, whether it's drivetrain or all the content that replaces under the hood. We see that sustaining. We've had strong the last couple of years, and we see that strength sustaining, not really tied to SAR, but tied to more on penetration of these megatrends into the general population of vehicle. We can still see the softness in the consumer and compute.
You know, we started seeing that softness in, think about it, in the second quarter of 2022, and we proactively took utilization down, so we're able to navigate and manage through that very, very well, and we'll continue to do so until the outlook changes. Then in the fourth quarter, we started looking at softness in the industrial market, but more on the consumer side of industrial. Think you know, power tools more that is indexed to consumer. Remaining strength in, for example, our core industrial like, our energy infrastructure or energy storage specifically, where that actually accelerated in the second half of the year. Where, you know, we started the year at 50% growth from 2021, and we ended at 75% growth.
That momentum will continue through 2023 as that infrastructure builds up. Overall, you know, it's at where we expected it to be and where we have been very proactive in managing through it and will continue to do so until we see a different knee in the curve.
Great. If you put this in the context of kind of the larger journey that you guys are on, you know, when you came in, you talked about a strategy of exiting businesses with less sustainable profit pools. And I think you've now exited, if I have the numbers right, $294 million in total, $17 million last quarter. You know, you remain committed to that, which I think is impressive for people to see given that you have seen utilization come down, you would see contribution margin hanging on to those businesses. You know, can you just talk to the general progress there, and then the fact that you're taking $75 million out this coming quarter or the quarter that we're now in, is a bigger increment.
Can you talk about what's driving that?
Look, I mean, obviously we thought we'd be way done with the exits that we planned, but the market environment has delayed it. That $400 million that we still have, obviously it's accelerating. We talked about the $75 million in the first quarter. That's because we've always said it's market dependent where as soon as customers are able to get that capacity elsewhere from any of our peers, they're gonna go get it elsewhere.
With the softness that we talked about in the consumer and compute, which is where a lot of these products are, we expect those customers are going to be able to get that capacity elsewhere given the overall demand. That's the reason for the acceleration. However, it remains market dependent. Look, we stayed the course, like you said. The softness in the market did not deviate. Although that business today is, you know, call it the last, the 294 is in the 20% margin. This one is, call it high 30%, low 40%. In a normalized environment when there is supply, and supply comes from other, that business is 20% margin to negative margin. The reason I don't believe, nor does the company anymore talk about contribution margin, is because there is no such thing.
Mm-hmm.
You know, when you're in the 20s, that's zero to negative. It's highly dilutive.
Yeah.
Now you're putting stuff in the fab that is, you know, what I call it's empty calories revenue. When we have constrained capacity and we have growth into those businesses, we want to make sure that we don't have anything in the fab that is dilutive for margin. Every wafer we do has to be incremental. Our new products, which are margin accretive, that's the stuff we want in the fab.
If I hold on to that, you know, dilutive revenue, well, what happens when I need that growth? You end up exactly in the same spot. We would rather take the more responsible approach, which is take utilization down. We're able to absorb it because of the strength in the rest of our company, structural strength in the rest of the company, and then navigate through it so the growth on the other side is at much better margin overall.
Yeah. I mean, we've been in this great cyclical environment for the last couple of years and now kind of plateauing, but your view would be that the businesses that you're focused on it's a sustainable source of gross profit that's in line with your overall...
Absolutely. Absolutely. Even, you know, if we think about, 2023 being a transition year from our margin perspective. W e've shown the sustainability of our margin net of these headwinds that we have just momentarily in this year. That's what's gonna drive not only, you know, getting to the 48%-50%, but our new products are accretive to margin. As those new products become a bigger percentaage of revenue, we're on a very, very good trajectory that we're very comfortable with.
Great. Then you mentioned fab utilization coming down a bit. You know, you've been pretty disciplined around that, around working capital. You know, what's the end goal of that? Are you trying to prepare for an uncertain environment, prepare for some of this weakness to persist? I mean, how are you thinking about? Not asking for a forecast, but...
Yeah
... how are you doing the planning?
Look, if you think about it, we look at uncertainty. You know, if there's certainty and it was soft, you can manage to it. If it's softer and there's uncertainty, then you have to take all necessary precautions so you don't get caught with it. What does that mean in our terms and what we've done from an inventory perspective? The worst thing you can do is get into a softer market that lasts longer than you think, and you piled up a lot of inventory. 'Cause then what happens is, one, you start reserving and writing stuff off, so negatively impacting your margin. You still have to put a very hard stop on your manufacturing, which hurts the margins even more because utilization drops very quickly. There's a no-win situation of trying to hold on and be in denial.
As soon as we started seeing that softness in those end markets, again, back in Q2 of 2022, we started taking inventory down. Look, if we got it wrong, we'll have to ramp back up, and it will be a few escalation calls, more than I have today. You know? That's not a problem. I got used to it. If you get it right, and you end up getting there with no, you know, low inventory, whether it's on the balance sheet or in the distribution, you're gonna come out the other side much stronger than you went in, and that's the position we're in today. We're set up beautifully, not by accident.
Mm-hmm
... for what is yet to come. As soon as the market starts becoming better than we think, we just ramp back up, and that's actually a tailwind to margin.
Great. Before I get into the end markets, the role of long-term supply agreements, you've talked about over $16 billion, like $16.5 billion secured so far. You had a $2.5 billion incremental last quarter, which is an impressive number. Can you talk about the reason that people are placing those and what the visibility that gives you in the business?
Yeah. Long-term supply agreements started off in the middle of the shortages that we've had in the last few years as a way for customers to solidify what they need and get what they need. We were supply constrained. We're now building on backlog. When customers wanted those highly constrained technologies, we had an engagement with the customer of, "Well, what do you need over a multi-year period?" Because what we needed is the visibility, do we need to invest in CapEx in order to sustain the growth? Given everything I talked about the megatrends in auto, those are sustainably sustainable markets as far as growth. In order to do that, you can't really do that with what do you need, you know, in the next six months and the next six months.
We have to have much longer view. We started establishing those long-term supply agreements. In essence, the customer says, "Here's what I need." They commit to what they need, meaning take-or-pay commitment to what they need over a long period of time. In return, we will invest in order to have the capacity to support their growth as well. It's a win-win when you do it right. That's how they started. Now fast-forward, we're, you know, we talk about phase two, and we talk about revisiting a lot of the LTSAs, you know, the $2.5 billion. You know, where does that come from?
Where it came from is existing LTSA customers that in the last few years, they've gotten what they need, and they came back and say, "You know, I know we have it on 10 products that were highly constrained. We discovered we buy 300 products from onsemi. We want all of them, and we'll give you the commitment of all of these parts in the LTSA, and we will go for five, six, seven, sometimes 10 years." That visibility is absolutely what we need for us to confidently deploy capital because we're not deploying capital based on, you know, a market share aspiration or some forecast that may or may not happen. It's a contractual obligation that we have with the customer. The other one is customers that came in and said, "We love it.
We wanna extend it, so it's a rolling kind of four-year." You know, we've had two years, they wanna extend it. Customers that in the last few years did not get what they want, and they realized it's time to make a commitment, so they came in and said, "We're ready to commit for five years. We need to get the products, especially the new products that onsemi has." Net-net, $2.5 billion, totaling $16.5 billion, and we keep adding.
The other thing I'd add there, Joe, is this has got pricing, quantity, and timing all set up over the duration of these long-term agreements. In the soft market, there's no discussion about pricing. Pricing's holding, pricing's very stable. If a customer needs help with quantity and we can allocate that somewhere else, we may do that, but we're not having a discussion on pricing. They can't come to us and say, "I can get it cheaper somewhere else." You know, it's legally binding. The pricing environment is extremely stable right now. Even as we saw consumer and compute get soft early last year, it's not a pricing discussion right now.
Okay.
I know you've said that even if there's a flexibility on timing, that you get more of an advance notice on those timing risks.
That's right. Yeah, you know, the anecdote I always use is, guess what? I'm gonna get a phone call.
Yeah.
It's legally binding and it's take or pay. As soon as the customer gets that feel of, "Hey, you know, I think in six months I don't see the demand on my customer's backlog," they pick up the phone and they give me a call and say, "I don't think we need this in six months." It gives me time to react by, like Thad said, if I have another customer that sees strength for the same part, I'll just move it.
Mm-hmm.
The customer will work with us on, okay, is there another part that we can get more share on? Net net, it's a win-win.
Great. We could pivot and talk about the end markets a little bit, starting with auto. Before we get into some of the sub-submarkets that you're addressing, you know, what are you seeing from the automotive industry in terms of supply chain management? Is there a desire for those guys to put inventory in place, either at the OEM or tier one level? Is there a desire to secure capacity from domestic vendors? Like, what are you seeing as a reaction to the shortages over the last two years?
Look, the biggest thing I would say is night and day from where. You know, I've been in automotive my whole career, but the difference that I see now is the OEMs have and want full visibility all the way to the silicon semiconductor suppliers. What they realize is there's this chasm, this big gap of what they had visibility to what. I mean, I've had conversation with customers saying, "We had no idea how many products we bought from onsemi." The focus for the OEM is really not how to solve the problem, you know, of, okay, we need to get inventory here, or it needs to be localized. What we realized is the disruption that we've had in the two years was agnostic to region.
Mm-hmm.
Right? It was, it was agnostic to inventory. You know, it sucked up inventory immediately because it was beyond that. What they realized is, how do we work together on that sustainability, you know? A lot of people focus on supply assurance. My conversation with customers is supply assurance and supply resilience. With our manufacturing footprint, that a lot of it is geographically distributed already, we are in a very good position to also discuss with customers on supplier diversification. You don't have to go to two suppliers anymore. You can go to onsemi from two different fabs that are geographically distributed. You get the same supply resilience and assurance. That's the focus of the conversation that we're having with the OEM at a much bigger and strategic level.
If you look a lot of the announcements, they've been made with OEMs, which is a very different scenario that we've historically have had.
Great. Thank you for that. Can you talk about some of the incremental auto opportunities? Maybe start with ADAS. Silicon Carbide gets a lot of the attention. ADAS is obviously a big opportunity for you as well through primarily image sensors. Can you talk to how you see that opportunity scaling over time?
Yeah. Look, I mean, if you think about ADAS, you know, if you rewind four or five years, we should have flying cars that drive themselves today. Well, that didn't happen, because there are non-technical just challenges in it. Let's take the autonomous driving as a vision, but talk about Level 2+, which now is the penetration that we talk about, where although the car is not fully autonomous, where you still need a driver, the content that's in there, you know, whether it's for 360, whether it's for occupant detection, whether it's in order to have the hardware in the vehicle that you later can upgrade into more autonomous functionality, that content is what's driving our image sensing business today. The content is being driven, that's incremental.
The penetration into the automotive market where there are more Level 2+ functionality going into production vehicles, so penetration is up. Number of cameras have doubled in the last five years and will double again in the next five. Again, content on top of SAR for ADAS. On top of all of this, there's an ASP expansion because of resolution. You know, right now the world is moving to 8 MP from, you know, one, two, and sometimes 5 MP . 8-MP is a bigger, higher ASP than the other ones. That lift is also incremental to the growth. Those are the variables that are going to sustain our growth in that market and sustain our leadership in that market. We're number one in the market today, and that growth is going to maintain that momentum.
You're seeing that growth, like, this year, or is that more down the road where you see it?
No, no, it's, it happened last year-
Yeah
It will continue to happen.
Yeah.
Yeah.
Okay. Great. You know, silicon carbide, obviously a big story for you guys. When you first gave the $1 billion guidance for this year, it was a pretty surprising number that's taken shape now as we've seen a bunch of the customer announcements. You know, can you talk to us about why you're doing well in that market? Is it the capabilities that you had already in silicon and IGBTs and modules that are bringing their way into this world? Just, you know, how are you securing all these wins relative to people who were bigger than you last year?
Look, in general, we've been very consistent in our approach in this market. What you need to win in the market is you need compelling and competitive technology. I've always said, I define technology as the device and the packaging. We're talking about power semiconductors. If you have the best device, when you pump a lot of power in it, if you can't cool it effectively in a small form factor and so on, you end up putting a lot of cost at a system level. You have to have the two, and you have to have the two tenets very competitively in order to win in the market, and we have that. The testaments from all the announcement that we've made with the customer are about us winning against everybody out there. That's for a few reasons.
One, we decided to heavily invest in silicon carbide and the packaging capability, but it's not a new investment. It's an acceleration of a decades of IP and decades of know-how into that market that we are just reapplying from silicon power to silicon carbide. We have that core competency that we have redirected and invested and accelerated in silicon carbide. On top of that, we are able to go to the customer and say, "Okay, now that you have evaluated our products, evaluated in your system, and you have concluded that it is superior than everybody else's, therefore we get awarded the business."
We get awarded a bigger share of the business because they come and review our supply chain all the way from our substrate capability all the way to our fab, and they say, "Yep, you got it. You can scale with us, and you got the award." 'Cause the worst thing you can have is you do all the technical work and the customer says, "Great, we are successful in the market. It's actually surprisingly better than what we thought. You need to scale." We go, "Sorry, we can't-
Mm-hmm.
-because we don't control our own destiny." They're not gonna double down with you. Our ability to have that vertical integration allows those customers to double down with us. Then as we get more visibility, we will implement that capacity and scale it accordingly, just slightly ahead of them. That's why we've always said we deploy CapEx for LTSAs. We're not putting CapEx for some, you know, market share dream that we have.
Mm-hmm.
We have customers, we will put LTSAs, and we will ramp those customers ahead of when they need it.
You mentioned the vertical integration. You did the GTAT acquisition allowing you to build your own substrates. We've all toured the facility, which was really impressive. You know, does that make it more difficult to ramp? Are the OEM customers aware of what's an internal wafer versus a Wolfspeed wafer? You know, does that? Or at this point is it sort of they're just looking to qualify the device?
It's the device. Qualification happens at a device and package. You change the device or you change the package, there's a qual. Anything before that, is not qualled. It's like... which is the same as silicon...
Mm-hmm.
...by the way. That's why I put the, our Hudson facility or the GTAT facility as a supply assurance and vertical integration, not why we're winning. Why we're winning all of the volume and all of that is because of the supply assurance, where we're able to provide that. The first anchor point, you have to win on your device and packaging capability because, look, if you have all the substrates in the world and you can't get the efficiency at a system level, the customer's not gonna care because they're not gonna be competitive in the market if they don't have the range that our efficiency brings them. We win because of technology. We build, scale, and win the platforms that you've seen us announce in the last few months because of the vertical integration.
Great. Then, you know, I think when we talked probably in December, we did one of these and we said, you know, "How come you haven't announced any customers?" Since then, we have Tesla, JLR, BMW last night, I guess, VW, a lot of customers now.
Mercedes-Benz.
Mercedes-Benz, yeah. You know, it still seems like $1 billion is a lot. You know, you're talking about a close to a couple million vehicles. You know, do you take a lot of time thinking about this bottom up relative to the size of the market? I mean, you know where your customer demand is, but-
100% bottoms up.
Yeah.
I mean, what we have is what we're building to.
Yeah.
We already are qualled in these accounts, so it's not. We already started ramping in these accounts.
All those customers are bringing up production...
Mm-hmm.
...vehicles now.
Yep.
Tesla made some comments last week that people found unsettling, that they're reducing content 75% of silicon carbide. It seemed like they were speaking about a future vehicle rather than across the board. Can you know, I don't wanna put you in a perspective of saying about a customer you don't wanna say, but just in general, you know, when you look at the content, do you see that content coming down in certain parts of the market over time?
No, it's not coming down. Silicon carbide is a net positive growth story, and therefore, you know, even with the, with the announcement, what I will tell you is there is absolutely no change to our $1 billion or the $4.5 billion in the LTSA, committed revenue LTSA that we've been talking about.
Mm-hmm.
Why? If you go back to a lot of commentary I've said publicly or in forums like this, we've always said when we engage with a customer, we engage at a system level. You know, what is the customer trying to accomplish from range, performance, and so on? We will drive down of what technology we offer, whether it's IGBT or silicon carbide. The, the benefit of where we are with our portfolio is we're able to support both at the most optimal price point for the required performance the customer needs. You know, we don't walk into an account saying, "We got silicon carbide. You gotta design it in." Because it is, like, more expensive. If the customer doesn't need the performance, maybe IGBT is more suitable.
I've also said that there are some customers that are opting or will even opt in the future of the front axle, like the non-drive axle being IGBT and the drive axle being silicon carbide, 'cause that's where you need the performance. Take all of those in context to the commentary made last week, and they're the same, right? When you talk about a $25,000, you know, $25,000 vehicle, you're not talking about a 500 range vehicle, right?
Mm-hmm.
Having that vehicle actually will increase adoption of electrification, which means it's incremental for us. Even the amount of silicon carbide in that, whatever % it is of the current generation, is incremental to what the industry has. Forget about where we play here. The benefit for us is obviously it's incremental for silicon carbide and it's incremental for IGBT. 'Cause it opens up yet more IGBT lifeline, and our IGBT business, we're still investing in it. We just transferred IGBT to our 12-in fab in East Fishkill in North America, so you have localized supply, you know, back to your localized supply here on 12-inch, and our roadmap is not stopping because we have silicon carbide. We still have a roadmap on IGBT.
I know how, you know, the angst that was around the announcement. When I saw that one, I looked at it going, "Oh, it's great." It's incremental and it's favorable for us, especially because we have both.
Great. Let me pause and see if we have questions from the audience. Like, we have three more minutes. One right here.
I got 'em right here.
The mic's coming over.
Hey, Hassane. Just a quick question on that. You know, if Tesla is able to reduce the amount of silicon carbide that they need, do you think that you could see ASP compression at either the silicon carbide traction level or at the onboard charger level? Like, could that $500 ASP come down to $300 or anything like that if they do require less silicon carbide?
Look, because that's all incremental. You need less silicon carbide, but again, I've always. If you remember when people ask me about ASPs and content, I always say, "Well, what kilowatt are you talking about?" The content doesn't scale with the price point of the vehicle. We have to be very, very clear. If you have a 100 KW vehicle and you have a 500 KW vehicle, you're using way less silicon carbide in that one versus the top one. That's how it's gonna scale. It's kilowatts. It's proportional to kilowatt, not to anything else. Look, our efficiency, by the way, when we engage with a customer, they come in and they have their model about their packaging and so on, and let's say we have 10 die of silicon carbide in there.
We go in to the customer and we model our packaging, our capability, and sometimes we end up with nine or eight. You can argue there's a 20% reduction right there, but we do that already. Why? Because what is the kilowatt that you wanna get out? That's what the ASP is compared to. When I look at ASP. I normalize it to a kilowatt, a dollar per kilowatt. That will scale ASP, so it goes from, you know, $500 to $300, but not at equal kilowatts.
Got it.
That's why the vehicles I always talk about platform with a lot of our customers. Every platform, and even in the Volkswagen conversation, every platform is a different range platform. You have a high range platform, high performance platform. Every single one of them is a different trade-off. The trade-off is what's gonna drive the ASP, not the content of what silicon carbide is in there.
There's time for one more question. Maybe you could just talk to, I mean, you mentioned gross margins being, you know, at your corporate average minus some headwinds. Can you talk about, you know, the headwinds of silicon carbide startup expense, you know, how quickly does that wind down, and your conviction in the sustainability of those margins, you know, in an uncertain macro?
Yeah. Let me take that. The silicon carbide in 2023. We think about 2023 as a transition year, right?
Yeah.
We've got headwinds from silicon carbide. It's 100-200 basis points. We think by the time we exit the year, that'll subside and we're back to We've always said at scale, our silicon carbide gross margins are at or above the corporate average, so we think we'll start to achieve that by the time we exit the year. We also have a 50-70 basis point headwind from the EFK, the East Fishkill fab acquisition, as we provide foundry services to GlobalFoundries. That will go over three years, but it steps down each year. 2023 is actually the worst of all, you know, with the, with it wind up. If you actually pull all that out, we're in our 48%-50%.
Mm-hmm.
-gross margin range, right? We're also really comfortable with our floor, given what we've been doing with utilization of saying our floor should be in the mid-40% range. I think we've got a nice kind of boundary conditions on our margin with nice tailwinds going forward in 2024 and beyond. We also have the exiting of the four sub-scale fabs, where there's $160 million of fixed cost, annualized fixed costs that come off over a 3-year period. That starts to kick in in 2024 as well, so we got a tailwind there, as well as ramping new products and things like that. 2023 is a transition year for us.
Great. All right. Thank you so much. Appreciate it.