ON Semiconductor Corporation (ON) Earnings Call Transcript & Summary

I think we'll get started. I'm Tim Arcuri. I'm the semiconductor analyst here at UBS. Very pleased to have ON Semi. With ON, we have Hassane El-Khoury, President and CEO; and we have Thad Trent, who is EVP and CFO. So thank you to you both for being here.

Thank you. Thanks for having us.

Thank you.

So I have to ask the first question. It's been a little bit over a month since the earnings call. Anything that has surprised you from other companies that have reported or any developments in the market that are -- that you would want to call out relative to what you talked about in your earnings call?

No, it's not a big deviation. I think the biggest thing is inventory digestion is persistent. We did see, obviously, a lot of the headlines coming out of Europe. So there's a market challenge in Europe that -- not just in automotive, but in general. There's a lot of headlines on the tariff that add to a lot to the uncertainty. So, therefore, it goes back to -- we're not expecting a recovery. What we said last time, we remain consistent where we're not expecting a recovery in 2025. So back to the L that I talked about a few quarters ago. But we do expect below-seasonal quarters, at least for the next couple of quarters, kind of consistent with that. But there are headwinds that we have to acknowledge and focus on what we can manage, which is our own execution keeping the cash flow where we said it would be focusing on inventory management. You've seen us take very proactive actions on utilization to make sure we manage inventory that even through this downturn, the base inventory remained in a very healthy level. Our cash flow remained in a very healthy level. These are things we can control. I can't control the macro, but consistency and execution is what we can control, and that's really where Thad and my focus has been.

Great. EVs in China are a great piece of your story. I think you said 50% China EV share exiting this year. Can you speak to that and sort of how you see the China EV market evolving for you?

Yes. China EV has been a bright spot. I mean we can see a lot of the new models coming out, very competitive models. But because of that competitive models and the aspirations to go beyond the China market, so it's not just Chinese EVs for the China market, but you can start to see Chinese EVs in Europe but also in a lot of the emerging markets. What that puts it is just good enough is not good enough, which means they need compelling innovative products in order to start winning on a global market. They need innovative products and efficient silicon carbide even to compete locally. There's over 150 EV brands in China. So you have to be an outlier in the market from a performance and really from the luxury perspective to be able to get the share that our customers need. That's why they come to us. I've always said you're always going to win because of the technology you provide. And a lot of people say, how do we compete in China? It's the same thing we compete in China, the same way we compete in Europe, the same way we compete in North America. You have better products, customers value these products, customers will design the products. It doesn't matter what geography it's in. So our focus is on deploying our R&D to get those best products that customers value and that's where we get the share. And that's been our last 2 years R&D investment. So when I went to the Beijing Auto Show, where the 50% comment came from is I walked the floor, I knew exactly every car that was our 800-volt car with our 1,200-volt device. It was already on the show floor, which means our products are designed in, qualified and ready to ramp. And that's why we've seen that as a bright spot. We're starting to see those come through.

Great. Can you talk about the Trio platform? You just announced this. It combines analog and mixed signal uses 65-nanometer to integrate high and low voltage on a single chip. And I think you said the goal is to add $1 billion in revenue over the next 5 years. What is the main selling point here for customers? And is this not something that your competitors can also do?

So I'm sure since we've made the announcement, some of them claimed that it's not a big deal, and they can do. So let me tell you why they can't do. Because you have to look at the -- literally the Trio everything that it provides. So just to break it down. The Trio platform is an analog mixed signal based on a 65-nanometer from 1 volt to 90 volt, monolithic with a 65-nanometer, 175 degrees C. What does that mean in the market? And why am I comfortable saying we leapfrogged the competition and the unequivocal answer is no, they can't because they needed to do make a trade-off. Some customers, some competitors, some peers will say, "Well, we have 55-nanometer BCD." It goes to 25 volts at 85-degree C. So it's consumer-driven, okay? Some will say, well, we do have 70-volt BCD, so we can do high voltage. Yes, at 130 or 180 or maybe 90. Not a single technology exists in the world today, whether foundry or with our peers, that does 65-nanometer 1 to 90 volt at 175-degree C, which is auto, industrial and AI. So that's why we say it's a highly -- it's competitive and it's leapfrogging the competition because of that. So why do customers care? Because today, when a customer has, let's take the case of AI, for example, or the case of auto or industrial and they say, "We need a high voltage what we need to control together." Today, it's a 2-chip solution. You can get discrete power or you can get a high-voltage BCD, like I talked about, on a 130 or 180 to get the 90 volt. And then you have some other chip to do the control, logic or other. What we can do is do both of them in a single chip. And our ability to -- with our approach to design, which is a block-based design, which is an SoC-like design, we can go from a concept with the customer to a sample at the customer in 6 to 9 months. That's faster than an ASIC. So you're getting the best of both worlds. And when you are able to combine products into a single chip, people talk about, "Oh, yes, that's great." But the board gets smaller. The packaging gets smaller, the plastic around it gets smaller, the heat gets smaller. Everything gets cheaper, the customer wins. That's the competitive advantage of having this broad range in the Trio platform.

And is the $1 billion, is that truly incremental or does that cannibalize some of your existing revenue?

It's -- majority is incremental. Of course, we do have, like some of our peers, the 180-nanometer legacy stuff that in auto and industrial runs for decades. As those sunset and customers refresh their designs, we're going to be designing them on the BCD 65 on the Trio platform, but it is not something we're actively going after our own business. So I would say majority of the $1 billion is fully incremental. There's some that progression wise, we'll get on the platform, but otherwise, it's incremental.

Got it. Can you talk about the CapEx plan as it relates to silicon carbide? You recently sort of cut your plan and at the same time, your tone around buying Chinese wafers has changed a little bit and you say, look, we're -- I mean you've always been willing to buy Chinese wafers. But the quality of Chinese wafers is actually getting better. But that doesn't seem like it's why you've cut the CapEx plan. You've cut the plan because you're getting more output with the same CapEx because of higher yields. So can you just talk about that is -- what -- sort of what's changed that's allowing you to spend less money?

Yes, we've done better than we expected. So the performance of our existing silicon carbide footprint is much better than we expected at this point. But it started getting better than we expected a few quarters ago when we actually started taking down the CapEx for silicon carbide. So that's the reason we took down the CapEx, we were able to get the output. So it's not about the Chinese substrates. And my tone with that is still the same. We've always said, we're always looking. Some qualify, some do not. So quality is improving, but it's still within that 20% to 30% that we want to get from the outside, that percent has not changed because that's more a geopolitical rather than a supply capability. You can't go 100% in China, although maybe you could because of your geopolitical concerns, where you have European North American company, that's not a supply resilient plan to have that, especially in this macro and geopolitical environment. So that's the reason we took the silicon carbide or CapEx overall, majority of it was silicon carbide down, and we put our cash flow target higher than it was before. So it's a fall through to cash flow. Another aspect of that is not just the performance, but last quarter, I talked about we started to move to 8-inch silicon carbide and the yields are the same as 6. So now you're getting 70% more out of an existing one with a very small CapEx to convert from 6 to 8 the furnaces that we have. So our ability to get more output out of an existing installed base that is today, 6 inch going to 8 is why the CapEx is what we call CapEx slide because it's only conversion of the furnaces. Our fab is already 8 inch. So that's behind us. So that's why we feel very comfortable to maintain the growth trajectory that we have for silicon carbide, but a much lighter CapEx.

Great. I wanted to ask about the LTSAs. And I don't know if, Thad, you want to answer this or Hassane you do. But LTSAs have been coming down about $1 billion a quarter. And there's still a pretty significant part of the 13.7 that's noncurrent and I assume that's parked beyond 12 months. And so I would think that as that becomes current, that it will be the noncurrent that's coming down, but it's both for the noncurrent and the current that are coming down. So can you just talk about sort of how the LTSAs are working and how the noncore...

Yes. I think you got to look at it. What we report is it's a rolling 12 months, right? And then we do a lifetime. And as you said, both have come down. That's a result of, one, shipping every quarter, right? So we ship a portion of the LTSAs and another quarter kind of rolls in. If you think about the environment of what's going on with the softness, LTSA dollars have come down as we were not just going to overship customers, right? We're not going to force the inventory on customers and deal with a hangover there. So LTSA volumes have come down as we've negotiated win-wins with our customers, right, going through that process. But if you think about going forward, right now where you've got customers looking at renewing LTSAs or extending LTSAs, in this environment, they're just reluctant to commit to a volume right now. And that's okay, that's temporary. We have customers that are worried about a recovery happening and what's that going to do and concerned about supply assurance when that happens, and they're talking to us about it. But right now, they're being very cautious. So when you look at those outer quarters, yes, they are lower just because the slope is lower, right? But I would look at it generally, whether it's short term, 12 months or longer term, just volumes are lower just given the market dynamics.

Got it. And then one investor concern is around pricing and everyone says well ON was the one who raised prices the most during the during the peak. That wasn't really you raising prices as much as it was, you walking away from business that you didn't want and mix got better. But there's this theory that as the LTSAs roll-off, that there's just a delayed effect on your pricing, that the pricing will start to roll off as the LTSAs roll off. Can you just talk about that?

Yes. Look, if -- when you talk about piece apples-to-apples pricing, there's stability there, meaning all the LTSAs expired, all of a sudden, there's an X percent drop. While the value of the product did not change. And I've always been very consistent here in these forms, but also with customers that we price on the value of the product. Whether it's LTSA or not, the value is the value. So we're consistent with that. Now, are we making improvements from not just the cost structure, but also think about the conversation we just had on Trio. Something going from a 180 nanometer going to a BCD 65 nanometer on 12-inch East Fishkill with a much smaller footprint. Well, that costs less. So are we going to -- is the customer going to also benefit from that? Yes, that margin of the product is 60% to 70%. But from a customer that was paying this much before and now paying slightly less, that's a benefit. Is that a price reduction? No. There's just a value on a new product. That helps with our margin. So you're going to see that, but it's not the good old days of January 1, everything goes down 4%, set your clock to it. I don't subscribe to that.

Okay. Got it. Can we talk about silicon carbide? I believe you did a little more than $800 million last year. This year has been a little bit lumpier. You said that for the year, it will be up low to mid-singles this year. So can you just talk about how you see the market headed into 2025? I know you're not guiding silicon carbide revenue per se. You're thinking about it more relative to the market. But what are the factors to consider in where do the current third-party estimates think that the market's growing next year?

Okay. I'll start with the third-party estimates. They're wrong. So I don't look at them because they're still wrong. We're December already, and they're still wrong. It's like the magic is not going to happen in the last 3 weeks of the year. So I don't put a lot of value. It's a good indicator that it's positive. That's as far as I'll go. In 2025, we are going to see growth. And the growth is come in 2 ways. One is we started ramping some of the models that we set in the second half of this year, although not to the rate that we expected, but it did ramp specifically in Europe and some in China. So the stuff that ramped in Q3 and Q4, now you're going to get 4 quarters of it. So just that progression of the ramps that we started this year that will continue for a full annualized basis, you're going to start seeing that growth. And we have some market share gains in some existing accounts where we've taken share from some of our peers that are not doing as well. Both of these, you're going to see growth for us. Now I don't want to peg it to market or give a percent growth yet, it's too early, but we are expecting growth in '25.

You are. And are there any different dynamics in the industrial part of the silicon carbide business, is that growing? And auto is just a little lumpier because of what's going on in auto?

It's also lumpy because both of them are megatrend. The megatrend is the energy infrastructure, whether it's energy storage or energy conversion. And I'm not talking about residential solar that we don't play in. I'm talking about utility-scale, commercial-grade conversion and energy storage systems. We're #1 in that market. That market has got stellar growth the last few years as a lot of capacity got built up -- capacity, what I mean by conversion capacity, gigawatt conversion capacity. That we saw very good growth in -- 60% to 70% year-on-year last few years that took a pause as customers digest the inventory and get the deployment. We see some green shoots there. Where it's going to land in '25, that's really macro. But from our penetration in that market, we're very comfortable, we're in a leading position. So as the market recover, our market share is going to pull that recovery. But it is lumpy, just like any new adoption of a new megatrend, you're going to see that. What we can control is designing and market share. That is exactly where we want it to be. Now, it's purely an end consumer or end demand lift.

Great. Thad, I wanted to ask you about gross margins. East Fishkill has been a drag on gross margin. Can you speak to how that's going to roll off through '25? And sort of how you plan to ultimately get leverage from owning that fab? And maybe if you can just generally talk about some of the puts and takes? I don't think you're expecting a lot of gross margin recovery next year. It's more of a 2026 thing.

Yes, that's right. Look, let me just kind of start with where are we with gross margins, right? The #1 driver in gross margins for us in the short term is utilization demand, right? Today, we're about 65% utilized, holding a mid-40% gross margin. Every point of utilization is 15 to 20 basis points of gross margin improvement. You think about going from 65 so we peaked out around 83%. You can do the math on the gross margin expansion there. So that's going to be market-dependent, right? So just as demand comes back, we'll increase our utilization. The other headwind that we have today currently is 100 basis points from East Fishkill. This is the foundry business that we're doing for Global Foundries as a part of the acquisition of that fab. Contractually obligated, it's going to be -- it will wind down in '25. So by the end of '25 that 100 basis points will be off in the headwind. So you can think about that as fairly linear throughout next year. And then as you go further out, you've got the divestiture of the 4 fabs that we did in 2022. It's $160 million of fixed cost that comes off once we start manufacturing that inside our network again. Now again, some of this is going to be demand driven, right? So as the demand comes back, when we start manufacturing, we get the benefit. Don't think we're going to see a lot of that in '25 just based on what we're seeing in the market visibility of not much of a recovery. So we'll start to see that in '26. And then to your question about East Fishkill, so we're loading more products in there. This has been a long transition for us of qualifying products in there. Trio is going to be manufactured in there. So it's a matter of bringing that utilization up. That kind of goes into that equation that I gave you. And then the last element is just the mix shift of the company as you go into higher-margin products as we've exited the low-margin type stuff, you get a favorable mix, and that's a nice tailwind with us over the long term as well. But short term, it's all about utilization demand.

Great. Can you talk about how you think about tariffs or export controls? I'm sure that you're war rooming around what could happen from a tariff perspective. Does that affect ON or do you feel like you're pretty well insulated from what could happen from a tariff?

Look, it's too soon to tell because there's a lot of headlines, but nothing kind of -- there's a general commentary about tariffs here and tariff there versus the country and so on and so forth. So it's too early to say what is it that we're talking about? Or what is the tariff and what does it apply to so on and so forth? What I will say the impact is the uncertainty it creates. Uncertainty with customers that are looking at it and going, okay, the same uncertainty I just mentioned, customers are looking at saying, there's uncertainty. So we have to start thinking about revisiting our supply chains, supply chain decisions from a supply resilience, supply assurance and so on. So it is creating a cloud of uncertainty that I hope we can quickly get clarity one way or another because you can disagree with where it's going to land, but the uncertainty is worse than where it would land, favorable or unfavorable. So we need to get that certainty for us to have a proper assessment, but it is creating some angst in the supply chain. Now, the way we're dealing with it, of course, is we have a very broad footprint -- worldwide footprint. We're not anchored only in North America, but we do have North American manufacturing, including fabs. We have Southeast Asia. We have Japan. We have Europe. So from a geopolitical environment, customers do see us as we have the versatile and a pretty global footprint. So they feel more comfort with that, but the uncertainty remains.

Great. Can you talk about inventory? I know inventory at disties seems okay for you. It seems within the 9, 11-week range. What -- how does that inform you to how your shipments are relative to consumption? Do you think you're shipping in line with consumption and it just as consumption that's bad?

No. I think we're undershipping natural demand right now. We still think there's an inventory digestion going out there with our customers. So we think we're under-shipping that demand. When you think about inventory, I think we've done a really good job of positioning the company. If you look at what I call kind of our base working inventory at 113 days, we've got strategic builds for silicon carbide and for our fab transition. So if we exclude that in 113 days. Our sweet spot is 100 to 120. So right in our sweet spot there. So if you think about whenever this market recovery happens, utilization starts to come back pretty quickly. Inventory in the channel, the distribution channel is running just under 10 weeks. We're going to run right in that ballpark for a period of time. We've been increasing -- kind of surgically increasing the inventory in the channel to support the mass market, and we're seeing our customer base go up in that mass market. That's to see that long tail of customers. So I think we're doing all the right things here kind of to support the long-term vision of what we're doing, and we positioned the company. By seeing this downturn coming early, we took utilization down early. So I think we're in a good spot when there is a recovery to ramp back up quickly as well. It's not like we've got to burn through a lot of inventory in the channel and our balance sheet. We can flip things back on pretty quickly.

Can you talk about capital return and M&A? You've been repulling about 70% of free cash flow over the past few quarters. How do you think about use of cash? And is M&A even possible in this environment?

Yes. So from a cash flow standpoint, we've returned over $1 billion in share repurchases since February of '23, right? So we've been very aggressive, 75% of our free cash flow over the last 12 months. Our stated policy is 50% return. We're going to continue to do that. And if you think about what you talked about earlier, with the CapEx declining, right? So if you think about CapEx declining to a mid-single digit, let's call it, 5% capital intensity, that creates a whole lot of cash flow. Even in a down market, it's creating a whole lot of cash flow in 2025 that we can return to shareholders. So I think what we've done over the years is we've really restructured the balance sheet. We're really clean there, right? We don't have a lot of debt. We're returning the capital to shareholders, and I think we're in a really good spot that we're always looking at in M&A. We've done some small tuck-in acquisitions. We're looking at things that are strategic. But any M&A has got to be strategic and financially viable for us. And in this environment, you have to continue to look at things. And the hurdle rate is higher just given the regulatory barriers, but you have to continue to evaluate those things. And I think we're in a good spot.

Great. We focus so much on the power part of your business. But can you speak to the other parts? How are they trending particularly ISG?

Yes. I mean, ISG is going through the same inventory digestion that we have at the end customers. The business did grow, so we are seeing some pickup. And that growth is not just a unit, but we started a conversion that we've been talking about 8 megapixel, which is an ASP uplift in this case. We talked about ASPs going the other way as technology advances. In this case, as you go from an average 2 to 3 megapixels today 8, megapixel actually ASPs go up that in that business. So we have seen that penetration. But there's both inventory correction going both in auto and industrial. If you think about factory automation, a lot of the robots that are running around in factories, they have a lot of imaging that we provide. That has not recovered yet. So we're kind of more tied to the market, not more of a penetration. From a new product, however, we have introduced almost 10 new products on 4 different platforms in ISG, both in automotive and industrial. So that's feeding a lot of the market share gains that we talk about, but also a new product revenue increase that we talked about. As we get these new products, of course, margin comes associated with that, which is a good thing. And the other thing, Thad talked about tuck-in acquisitions. We did the SWIR acquisition, that adds yet another layer of functionality and differentiation to our product, where we've introduced those products in industrial, and that's starting to get a lot of pickup from customers. And as they design in, that's going to get the pickup of new products. So overall, we're in a pretty good position. It's again, back to the rest of the macro. It's an end-demand conversation, but we're doing everything we can control. We're doing it all on time and the products are just out there ready to rent.

Can you talk a little bit about LTSAs in the context of the China market? Are they willing to engage in LTSAs? Is that something that's attractive to them? And are they -- do they stick to the LTSAs when they sign them, the Chinese EV makers?

Yes. It's not just Chinese EV, like I'll take it as general because we have some industrial customers in China as well and the behavior of China customers with our LTSAs is consistent with all other regions because who do we have LTSAs with? It's really the brand name, the marquee names in China. Those -- they do have a reputation to maintain because we're not the only suppliers, one; and two, they value the products and the relationship because we're not a one-product shop. We have a broad portfolio. But the challenges remain, and we deal with the same challenges as we deal with European and U.S. customers, where we want to -- if there is a need for a change in mix or a change in volume because look at the macro, nobody escapes the macro, we're going to talk about a win-win. Same playbook, no behavior change and no terms change between China and non-China LTSA. So from our side, we're very consistent and from our customer side, they're very respectable.

Great. And what's the thing in the silicon carbide market that we might be missing? Because I know, I get these continue to sort of like negatively leading questions about silicon carbide. And so what -- when you talk to investors, what's the thing that people miss about your position in silicon carbide?

So I think a lot of the overhang on silicon carbide, where you think about it, a lot of people mix device and it's related to China. It's related to supply, but a lot of people confuse and make substrates and devices. If you separate them, it's a very different picture. If you clump them together, oh my god, there's oversupply is coming from China. It's got -- and we all know how that movie ends and so on. But if you separate them, it's no different than any other business. And we look at it separate. I don't -- we, at onsemi, we don't supply in the merchant market, substrates in the merchant market. So having a supply available from China on silicon carbide substrates is actually -- could be favorable for us from a cost perspective. They're able to supply it lower than we can make it internally. Well, guess what, our margin gets better. So that's a positive development for us. Now, the macro and the geopolitical environment may dictate otherwise, but it is a possibility. The device side, if you now talk about devices and device technology, China is not close to where we are or some of my European peers or North American peers. So there's a lag there. And that lag is not, "Oh, you can dump a lot of money and catch up." That lag is generational improvements where we've had that generational improvement. We're about 2 generations ahead. And our R&D investment is to maintain that leadership in devices. So if you separate the 2, it is not as dark of a cloud over it. It is still a growing business. If we take it all in, in a market environment, forget about the pause in EVs and the lumpiness in EVs we're seeing in the short term, if you take out the U.S. EV disruptor, silicon carbide penetration in EV 6%. I mean we have a ton to go. Even if EV penetration stays the same, silicon carbide penetration into EV is going to have a lift of demand. So we're still in the early stages. And as long as you maintain device competitiveness and I say device and technology, I add packaging to that. If you maintain that competitiveness, you're going to win the long game. Substrates, yes, if you're in the substrate -- if you're in the merchant substrate business, I said it last quarter, good luck.

And I would just add back on the penetration of 6%, right? If you look at our design win activity over the last couple of years, and we sell both silicon carbide and silicon products IGVTs, so we're a really good proxy for the market. It's the vast majority going to silicon carbide. So it's a matter of these models getting to production, which is a tailwind, right? So as Hassane said, even if EV production goes flat, the penetration rate is going up.

Got it. Well, we've run out of time. Thank you to you both.

Thank you.

Thanks. Thanks a lot.