ON Semiconductor Corporation (ON) Earnings Call Transcript & Summary

Great. Good morning, everybody. My name is John Vinh. I cover semis here at KeyBanc Capital Markets, and pleased to have Hassane El-Khoury, CEO; and Thad Trent CFO of onsemi. Welcome, guys.

Thanks.

Hassane, maybe just kind of high level. I feel like you're always really good about just putting things into perspective. You guys put up some really strong numbers, but we're obviously starting to see demand, especially in the consumer side start to slow down a little bit. Inventory is increasing in the channel and at end customers. It feels to me we're headed for a some sort of correction at some point, maybe the debates either softlanding, hard landing. But I feel you'll fare better than your peers given your limited exposure to consumer and focus on electrification and kind of EV trends within autos and industrials. But I'd like for you to just give us your view on whether the cycle is different this time? And do you think you're more likely to see a softlanding versus hard landing?

Yes. Look, I mean I can speak for ourselves where we are, at a high level, I look at the market in 2 different buckets, especially from our perspective. You got the auto and industrial and then you have the consumer compute. And we're definitely just like a lot of our peers that have much more exposure to the consumer and compute. You're starting to see signs of softness. And for us, specifically, that's not actually the tractor to our strategy, it's a potential accelerator because you'll hear Thad and I talk about exiting some noncore businesses. Those noncore businesses are the consumer and compute. So as that market, you start seeing softness and that accelerates our exit. It's actually favorable to our strategy because on the flip side of that, back to where we will fare up better is the auto and industrial. So as that growth happens in auto, industrial to one, offset the losses, which we have done a good job of over the last 4 to 6 quarters. That's going to be now the net-net cost of the company. And that's really -- I don't want to say immune, but that's more disconnected for anything that happens from a demand environment. And let me give you kind of an example of why I feel that way. If you look at automotive, and I've always said it's a content story, not SAAR. And people say, okay, everybody says that. But let me give you an example from an OEM perspective. If an OEM is building the 5 million units, 5 million cars. Let's say they get cut 20%. That's 4 million vehicles. That's a pretty big cut. If 1 million vehicles out of that 5 million is EV, when they go down and reduce 20% and they built 4 million, they're still going to build 1 million of EVs. No single OEM is going to pull back on their EV aspiration because demand of the SAAR fluctuates because that puts them at a disadvantage over a long time. So for us, having tied to that strategy from an OEM, which is the customer demand, end demand makes us that much more immune to any fluctuation. Now I don't know I can judge the "softlanding correction", whatever people are calling it in this conference, too, it will be better for us than it was for us historically, but also a lot of our peers.

Great. That makes a lot of sense. I think the other kind of topic that investors are kind of wrestling with is on LTSAs, right? You talked about your LTSAs are covering 66% of your revenues versus 62% a year ago. I wanted to get your thoughts on how you think things play out here if we go through a more meaningful correction? I think there's a lot of skepticism from investors that when things truly do soften that these things hold, I think -- over the last week or so, we saw a few of your semiconductor peers announced that they're trying to decimate and renegotiate some of their long-term wafer purchase commitments that they timed just last year. So your thoughts here would be great.

Yes. So just one correction, John. The 66% is the percent of auto and industrial percent as part of total revenue. I'm not talking about how much a percent is under LTSA or not. But the question remains. What does the LTSA get us? I focus a lot on LTSA and people are like, okay, well, then what, what do you do with it? So let me give you what an LTSA is. LTSA is a legally binding document between us and the customer that discloses on a part number basis, multiple years that extend into 24, 25 and sometimes 29, that has volume and price in it over that period of time. And it's legally binding. I use it to put capacity where -- when I know the demand is, and I'm not going to build capacity or invest on hope. When I say on hope "is for what people call forecast in our industry that can get canceled 30 days before I ship it." So the LTSA provide that foundation and the visibility, clarity and visibility. Now that's all fine where we are today. The example you mentioned, let me give you our attitude, what would I do? Well, one, and go back not too far, in '19, backlog got canceled. Nobody got a call. I know a semiconductor got a call and saying, yes, we're going to cut our backlog. While the LTSA guarantees somebody is going to pick up the bat phone and call me because there's a liability in there. So first, I'm going to get a call, and I'm going to get a call ahead of everybody else to figure out what are we going to do about it. Now if it's a demand issue, where they say, I thought it's 1 million, now it's 800. There's that 200-unit gap. If I get it ahead of time, I can manage my business differently. But it's not an impact on pricing because pricing is in there also. So yes, there's maybe some change in volume and so on that I would offset with that customer where I say, okay, if this is not ramping as fast as you think, well, how about the other one that's ramping, I'll take that share. But that gives us that visibility and the dialogue way ahead of just backlog disappearing. But people are also concerned about what happens to the margin when that happen -- for our story with the margin, people always ask, okay, what about the margin? The margin is also in the LTSA in the form of pricing. So that's not going to change because I also don't want to shove inventory with the customer saying, here's LTSA, take it. That doesn't favor anybody in the long run. But the pricing is firm, and the pricing is an obligation on both party. So that gives the sustainability of the margin, the credible outlook of the revenue and the better capability to manage to that output seeing it coming. And that's the difference.

Great. And John, I think that helps with the softlanding too, right? We're going to get these notifications from our customers early, allows us to adjust our manufacturing strategy. We're not holding a bunch of inventory. We're not pushing inventory into the distribution channel. So we can manage proactively once we have those discussions with customers versus reacting, which is what our industry has done in the past.

Yes. That's a really good point. Maybe we can talk about silicon carbide. Pretty impressive increase in your committed revenue outlook over the next 3 years from $2.6 billion to $4.0 billion. What were the factors that drove such a meaningful increase in such a short period of time, essentially from a quarter or 2 ago? Was it new design wins you secured? Was it just your ability to get confidence in the ability to kind of ramp supply and capacity here for your customers? What drove that big uptick?

Look, all outlook in demand and the $2.6 billion to $4 billion is all driven by customers. I'm very confident in our ability to scale. And where we are today from where we were 6 months ago, our scale is happening, and it's an execution-driven scale ramp. So a lot of the wins in our capability are new designs, both with new customers where we've been engaged and also existing customers that we've had in the $2.6 billion that came and said your ramp is there. We've ramped -- I mean we have a triple silicon carbide this year from last year. That's a hell of a ramp already. So we've proven the ability to ramp and customer come back saying, okay, you've proven the ability to ramp one, you have the supply assurance because we're vertically integrated now as they develop and they deploy new platforms, they're coming back to us and say, this is the platform we want. And with our technology, you always win on tech. I always say, you can have all the scale and manufacturing and the supply. If you don't have good technology, it's irrelevant. We have good technology. We're able to do have the supply assurance and the vertical integration that gives that assurance. And we're getting more of the designs and the platforms that customers are introducing, but also new customers that we've been working with over the last 2 or 3 quarters have now crossed the boundary where they say, "Here's the award, here's the LTSA, now let's invest in the capacity." Because I'm -- we're not back to the point -- I'm not adding capacity blindly. I'm adding capacity where we have LTSA, especially in silicon carbide.

You had mentioned you're pretty confident in your ability to add silicon carbide. I know some of you have a really technical background, so you understand just the technical challenges about adding capacity. I'm just wondering at a high level, if you could just highlight the differences in terms of expanding capacity for silicon carbide fab versus a traditional CMOS fab? I imagine it's harder to do silicon carbide or else everybody would be doing in and adding capacity. Just trying to get your confidence level and why you're confident of being able to meet these customer commitments?

Yes. So my confidence stems from -- obviously, we've been in the power semiconductor supply chain for over about 2 decades with IGBT. What adds to that confidence is where I'm building silicon carbide is where I've been building IGBTs for a decade or more. So it's not greenfield, new fab build it, got to spend the next 5 years hashing out all the stuff that comes with new fab, whether it's technology or ramp or bottleneck or equipment coming in. So we have an infrastructure. We have a fab and we're expanding an existing capacity of a power fab, right? So that gives me the confidence. And that's what allowed us to get to -- on track to triple this year from last year. And I haven't announced any new manufacturing sites for silicon carbide, including that ramp. Because as we ramp, we're converting existing baseline. There are, of course, gap to -- what we call gap tools where you have to put a few tools in between, but those are just like -- well, they're exactly that gap tool where the technology for silicon carbide is different in one portion. So we'll get that one machine, but pre and post machine is already established. It's already established. It's running, yields -- I mean everything is baseline already. and we're just running silicon carbide. And the same thing, a few quarters ago, I talked about the 200-millimeter silicon carbide introduction. We did the bull, the substrate growth on 200-millimeter and I ran it to our fab. Because I also have an IGBT fab at 8-inch. So I can run silicon carbide through that without having to build an 8-inch greenfield silicon carbide. So as revenue ramps, I start converting more and more, and that's more of an organic capacity expansion for silicon carbide than a -- from 0 to that level, which is I agree with you, would be very challenging. But we're not in that arena here.

Got it. Maybe a follow-up for Thad. Just can you just help us understand as you have this accelerated ramp of silicon carbide, it sounds like that's kind of a near-term headwind on gross margins. What are the factors that are driving that? Is it just right now in the ramp phase, you've just got fixed costs that you're amortizing? And then longer term, once you get through this ramp phase and you're at scale, talk about longer term? Or is silicon carbide going to be a tailwind for your gross margins?

Yes. So we've said that the headwind for the silicon carbide ramp is about 100 to 200 basis points. That's because we're bringing on -- we've got all the start-up costs. We're bringing on that equipment. It will be a little bit lumpy for the next year, depending on how that layers in and the revenue layers in. But we can think about that peaking somewhere kind of about mid next year. We think by the end of next year, we're at parity. We've said that our silicon carbide gross margins at scale are at or above the corporate goal. So it's accretive in what we're doing. We also have a number of things that are offsetting these headwinds, right? So we've got our fab lighter strategy. We've got the exit of the noncore business, which is -- will be accretive as we exit those businesses. We've got new products ramping at accretive margins. So that helps to offset that. So we think we're going to maintain this 48% to 50% range, which is our target as we go through even though we have these headwinds, we think we can continue to offset them. There may be a little bit of noise in there in that range, but we feel like we can maintain that.

Great. Hassane, you had mentioned IGBTs. I'm wondering if you could just talk about how you're thinking about IGBTs for EVs longer term? Well, eventually everything migrate to silicon carbide. And as that transition happens, can you talk about just the financial impact of the trade-off between IGBTs and silicon carbide?