ON Semiconductor Corporation (ON) Earnings Call Transcript & Summary

Hi, everybody. Let's get started with the next presentation. We're very pleased to have ON Semi up on stage. We have both the CEO, Hassane El-Khoury as well as the CFO, Thad Trent. I'll be asking questions of them. I don't know if we're doing the raising your hand if you have questions. If you do, feel free to raise your hand. So why don't we start by -- I think, first, thank you for coming. I know it's been a busy week for you.

Let's start on the cyclical side of things? Hassane, you've talked about an L-shape recovery pretty much all year. In my view, in the first half of the year, people were disappointed with that. In the second half of the year, it's looking to be more correct than not. Give us the update. Are you still seeing the same sort of muted recovery or muted…

Yes. I don't think since our last earnings, we haven't seen any signs that will give us a different outlook. And people ask what am I looking for? First, what did we see to call the L-shape? And what I would see in order to change that outlook? And what we need to see is a demand and demand recovery. And what I mean by that, it's not -- you're not going to see me react to, "oh, our backlog is getting better or we're getting more orders in end and end." Because remember, we're going through an inventory digestion. When that happens, you do expect back to start -- if you're undershipping demand, you will click back up naturally when you start shipping to demand. Until you see the end demand, meaning consumers buying items out of inventory, whether it's a car or it's an electronic module or anything, then you're not going to be in a recovery as I would refer to it as a recovery. And because I look at the end demand and the consumer demand itself, you don't want the false start of oh, backlog came up. And then you ramp back up, everybody gets that positive. And then if it doesn't sell through, now you have the inventory again, and you have to take it back down. So we have to look at a steady state recovery for us to go from the L-shape to an actual recovery. That's not to say that you're going to see some positive green shoots, and we've talked about them. We've talked about some of our markets starting to see green shoots coming up. And that's because it's not -- it's back to that lumpiness that we're going to see. So that's overall the macro we look at when I talk about L-shape, and we haven't seen any changes. The things that we talked about are positives, but not yet to call a trend that will lead to recovery.

Talk about a little bit what those positives are and where you're still seeing some negatives.

Yes. So we talked about automotive overall. We think -- the big positive from -- if you think about it at the beginning of the year, when you said people were disappointed, we see stabilization now. In this environment, I think stabilization in general is a positive. So there is stabilization. We did see -- if you look forward-looking, we do see growth or, a, call it, flat to up for some of the automotive areas. We did see some recovery in the energy infrastructure. That's been a very strong market for us. Medical is stable. So we are seeing these things that if you think about it started early, industrial, some of the industrial, if you recall, we called the industrial softness in Q4 of '22. That's when we started taking down utilization. We got where we are here today with the structural -- the results that we are delivering because of the hard decisions we've made not last quarter and 2 quarters ago, I'm talking last year and 2 years ago. That's the structural changes that we made that are leading to the results and the sustainability of our results today.

You guys have been very aggressive on the inventory side. Like you just mentioned, you started this process much earlier than most. Do you believe the customers are going to make the same mistake they've made in the past of over adjusting inventories on the downside and creating volatility, maybe not to kind of pandemic levels of volatility, but do you think anything has structurally changed? Or is the industry just going to be cyclical because it always is?

That's not -- I can't answer with a comment of the industry or a segment because it's at a customer-by-customer level. There are customers that it's amazing to me how quickly they unlearned the lessons from before. Time will tell if they will make it long term in the next uptick when they bring down their OEMs, fool me once. We all know the saying. The OEMs were going to just start shifting and then it's a matter of time before they disappear. So to me, it's not a market commentary. It's really at a customer. And we look at it today some of the customers are taking inventory down to -- in the number of weeks, like low single-digit weeks. And that's making the OEMs nerves today. Some are taking a much more disciplined approach of, hey, we have elevated inventory. We're going to burn it down. We're not going to burn it down to a painful level because if there is an uptick in demand, we need to service their customers, which is the OEM. So that's why I'm saying it's not an industry commentary. So is it going to be cyclical or always cyclical? That's going to depend on how many go one way versus the other. But today, when people ask us, how long do you think the inventory burn is going to be or what level is it, it's hard to give a number because every customer has got a different target.

Right. What about the pricing side of the equation? There are -- I'll have some structural questions later. And so ON Semi is a little bit different. If you went cycle to cycle, it's not a fair comparison versus your peers. But if you talked about to the extent there were benefits because of the shortages from a cyclical perspective, are we just going to return to kind of the normal low single-digit decline on a year-over-year basis? Or do you feel that there's a stepdown because there was a step-up?

Yes. I'll talk about ON Semi specifically. What other or my peers are doing, I can't -- we don't price on a cost plus. So what we've done when we talk about the pricing action, it was in the middle of our kind of portfolio rationalization, if you recall, independent of the pandemic. The pandemic accelerated or not the pandemic, but the post-pandemic demand accelerated our shift of the portfolio, but it didn't change where we were headed to begin with. We just got there faster. And what that means is that value -- price to value discrepancy, we were below market even in a steady state. Forget about the pandemic-driven demand. We would have gotten exactly where we are over time. So to compare to a baseline, our latest pricing action, if you think about pricing embedded and the price to value discrepancy was in early '22, mid-'22. So the last few years, you can talk about that's the baseline. And you see where our margin is. You see our margin was held up. When I say held up meaning with the underloading and I'm sure we'll talk about the loading factor, but that's the overhang on the price -- on the gross margin. It is not the pricing. Now expectation of is it going to go click down and so on. I would differentiate the 2. Everybody refers to historical pricing, which is on January 1 of the new year, all of a sudden, pricing goes down 4% to 5%, whatever it is. Well, I don't believe in that. I never did regardless of now or before because the value of the product doesn't decline on January 1 after New Year's. That's not how it works. Now there are efficiencies that are gained, but not on a product-by-product basis. If we have a new generation of products that we are smaller die, cheaper and so on because we advance R&D, there may be a, hey, Mr. Customer, can you convert to this new one? The price is lower, but our margin is higher. So the cost offsets the benefit. So the customer gets a benefit because there's a conversion cost for them. And we get the benefit because it's a gross margin expansion. That's the comment Thad and I make up. Our new products are at or above our corporate. We do want to get those products out and designed them. So it's not a pricing cut or pricing reduction on a schedule per se. That I don't believe in because if that's the case, then you're not pricing on value.

I would also add, we've got long-term agreements with our customers. These go on average 4 years. Sometimes they're going out to as far as 10 years. These have got pricing locked in based on some road map. In a lot of cases, we're aligning engineering resources and those strategic engagements to talk about where we're going. But pricing is locked. So in today's environment, if a customer comes in and their demand is lower, we can have that discussion and come up with a win-win, but if the volume is going down, the pricing isn't going down, right? It's not that discussion. That gives us that visibility on pricing, but also on demand. And that's why we saw industrial get soft before many of our peers back in Q3 of '22. We saw auto get soft in Q4 of last year ahead of many of our peers. And it's because of that visibility. So I think it gives us a better demand visibility and better pricing visibility longer term as well. So we get asked a lot of long-term agreements. Is it a structural change here? Yes, I think to some degree, it is, I don't think it's across the entire portfolio, but there's definitely a benefit to us and the customers. It's worked for the customers and to make sure they get the supply when they want it as well.

That's a good segue and you might have answered this question what you just said. But the pointed question people ask are LTAs worth the paper they're written on because they've fallen apart and for many companies when times have weakened. So does ON Semi view the value of them in kind of the relationship enhancing information sharing aspect? Or is it the strict you agreed to buy X number of parts at Y price, and you're going to be held to that?

Yes. Look, we're never going to just shift to a customer's overship their demand. That's never good, right? You're just causing a problem down the road, right? What it's allowing us to do is engage in a much more strategic level, right? So these are not done at a procurement level. These are done at an executive level, highest level, CEO to CEO type level because they're multiyear and sometimes billions of dollars. So it's a different type of engagement. So I do think they're here to stay, and I think they have value. Are we willing to make a change on it? Of course, we are. But we're going to come up with some win-win situation with the customer.

But one thing I do want to clarify, though, when you talk about some of our peers that didn't really work out for them. Sometimes, it didn't work out for the positive, but sometimes it was too negative because the customers did take and now you're dealing with a much more extended inventory with some of our peers. The difference is our LTSAs are multiyear. You talk about 3 to 5 years. If you have a corporate LTSA or an NCNR, right, order, the order turns NCNR when it hits the backlog. And you tell the customer, sorry, you signed up, you can't cancel it and you force it. The customer says, okay, great. I'll see you in 12 months. Now I have 2 years' worth of inventory, 12 months later, 0 backlog for 2 years. So you stop placing the backlog and the LTSA -- then there is no NCNR to enforce because you already enforced it. Having 3 years, you can't kick the can down the road and say, yes, fine, I'll take it. I'll see you next year because you have 3 more years of the LTSA. You can't be doing that. That's why from a customer perspective, we got the call early because they have to deal with it. Now they can't push it out because it's not going to end. There's no cliff in sight. It's 3 to 5 years. Then we have the conversations. And if you recall, a lot of our conversations either on stage or on calls, I've always said, if nothing else, we're going to get the call. And that's where it starts. And we got the call in industrial, and we got the call in automotive. And both of them, nobody liked what I said that there's softness. Yet 1 or 2 quarters later, everybody is singing the same tone. That, we got the call ahead of everybody, and we were able to manage the company better than most of our peers.

So last question on pricing and inclusive of the LTAs side of things. Hassane, you mentioned that the cost of the product and the value it adds doesn't change on January 1. But you also acknowledge that historically, that's how people did it. Out of your customers, do you think that relearning, training that you -- the change in behavior is accepted by the majority of them now and people get the joke or are you still having to fight that out in January is not as much fun?

It's not a fighting it out because like Thad said, we talk about volume, we talk about new products, we talk about conversions and so on and we talk about higher volume. Those are reasonable asks from a customer, hey, I'm going to double your volume, take this. Okay? But it's not a, oh, there's a ton of capacity, there's no more shortages, cut the volume in half and cut the price. It's got to come a win-win. If the win is a bigger market share, okay, that's a win, too. So that's what I want to differentiate. And those conversations don't happen at the beginning of the year. They happen when they need to happen. If there is an opportunity for more share or an opportunity for a new design that they're going to go, hey, we'll use the same part and double the volume, those are all conversations, what we call a win-win anytime. But it's not a no change at all. Then what's my win. What's ON Semis win. There has to be a win-win. And customers understand that because over the pandemic, every time we've had a win-win, they want. They got the volume. They got the quality. They got the ramp. So when customers see that, people ask, have we signed any LTSAs lately, right? Yes, we have. Because there are some customers that didn't have an LTSA the first time around because we ran out of parts to allocate. Now they're going, you have capacity, I want to see it at the table. So the next time this happens, I'm not left behind.

Right. So let's switch gears over to a little bit more of the secular side. And silicon carbide is always a hot topic. You guys have talked about, I think, growing 2x the market rate and the volatility, at least year-to-date has been more on the market rate than kind of your penetration above and beyond that. Talk about what you're seeing from the end market from an EV perspective in aggregate, silicon carbide and then ON's penetration within it?

Yes. So from an overall -- obviously, we all know we read the headlines on the EV. Now is EV a growing market? Absolutely. Even in the short term, EVs are a higher percent of total vehicles. So even if you get the lumpiness, is it going to be a straight line all the way? The answer is no. It's going to be lumpy throughout. Now within EVs, there's another aspect of growth, which is silicon carbide specific penetration because EV can be both silicon and silicon carbide. So you have the EV is one factor. You have the silicon carbide penetration within EVs. So if you think about it in terms of numbers, if you take out the early adopter of silicon carbide in North America, the penetration of silicon carbide into EVs is 6%. So it's -- we're at the early stage of silicon carbide within that. So those are the double kind of triggers. You got a silicon carbide into EVs, then EVs into SAAR. That has not changed. The slope may be different in the short term. But if you think about it over the long term, which is where we're investing and where we're capturing share, that doesn't change our strategy. And from a performance perspective, there's a short term and long term. From a short term, we talked about European models ramping in the second half of this year. We said that in the last year's Q4. Last quarter, I talked about post Beijing auto show, over 50% of the China models that were on stage made had ON Semi products in it. So our progress into the penetration of both in silicon carbide and in EVs is exactly where we worked on to be. The key remains how many of these cars are going to sell, which goes back to your demand commentary. That's the short term. If you talk about the long term, is this the new norm of EV growth that we should look at? Or is it going to recover back to where it was 2 years ago. It's not going to be always the same slope. It's going to change, and it's going to change with what I call positive catalysts. What does that mean? The first wave of EV purchases that people made are early adopters. People that want an EV, they had a charger, they can live with it. They have a commute and so on, early adopters. Second wave, as soon as we have charging network as broadly as gas stations, more people are going to get the next phase of adopters because now I'm not worried about finding charging stations. Then battery technology is advancing very rapidly in the last few years. Fast charging. As soon as you get to a broad-based 15-minute -- to an 80% charge in 15 minutes, you're going to get another wave of adoption. People today, if you can charge in 15 minutes, slam dunk, way more people will get EVs. We're headed in that direction. So when I talk about lumpy, that's how it's going to be. It's not going to be a straight line between now and 2030 or 2035 or whatever it is. That's the key. That's what I watch for from enabling technology externally.

What about the slope on the silicon carbide penetration going from that 6% to something more is that...

That's accelerating. If you look at today, most of the RFQs and most of the new designs being either RFQ'ed or being in development, silicon carbide is the primary. So it already flipped into silicon carbide. Because a lot of -- because it also matches primarily the battery voltage, most batteries are going to 800 volts. 800 volts is a 1,200-volt device, 1,200 volt device is way more efficient in silicon carbide.

We're also in a good proxy for that as well because we sell both silicon and silicon carbide. We've got IGBTs and silicon carbide. So when we walk into an account, it's what's the problem you're trying to solve. We're not necessarily just pushing the silicon carbide, right? We're happy to sell IGBTs all day long. So I think the fact that the vast majority of the wins over the last couple of years have been silicon garment. It's just a matter of those getting into production. That's the penetration rate. Now the question is the sell-through.

And the adoption of more of a hybrid that does a little bit of both. Do you see that happening?

Yes. So there are 2 types of hybrids. Just to clarify. There's a hybrid where it's primarily a dual drive where you drive on an internal combustion and you drive on electric. That's -- you can think about it in content in about $350. You compare that for ON Semi. You compare that to a $50 of content in a pure internal combustion. And those are drivetrain content, not vehicle content. So $50 going to $350 in a hybrid drive system. There's also a very big market, and that's specifically driven by China, making its way to the rest of the world because it is a path which is a battery -- extender hybrid -- range extender hybrid. What that means is you have an internal combustion engine that is only a generator to charge the battery. It doesn't drive the wheels. So it charges the batteries over long distance. For our content, that is as pure as a BEV. It's a battery electric vehicle or a range extender. Both of them is a $750 range content for ON Semi. So that's why for hybrid, you have to look at both. There is a very big case of the range extender. China is primarily a range extender.

The hybrid, that's actually one definition of hybrid that was super helpful. I was actually referring to the use of IGBTs and silicon carbide and kind of trying to cut costs as we've heard some OEMs talk about. How do you see that kind of more blended approach?

That's going to be -- that's more regional, I know North America, that's the hybrid they go with. Some of the European have that hybrid. That's where I talk about the $350 worth of content. When I say worth of content, back to Thad's point, we do both content. So it's going to be IGBT because it's not a primary drive, you don't need the efficiency and the power. It's more stop and go traffic and so on for the hybrid, primarily, it's an internal combustion drive for the hybrid as we know it. So you don't need the efficiency in the power of silicon carbide. So that's the IGBT side of it. And the fact that we do both will benefit with the customer choice.

So on the silicon carbide side of things, it's a point of big contention, not just because EV demand has slowed, still growing, but slowed, but also because the fear of commoditization on the silicon carbide side in and of itself, largely more on the substrate side than the kind of solution side. Talk a little bit about that commoditization risk and why I would presume you're not concerned about it from a ON Semi point of view.

Yes. So people talk about capacity or availability and quickly they tie it to commoditization. I don't look at it as capacity driven. I look at it as value driven. And what does that mean? I go back 3 years ago. People are talking about substrates when we were -- we've acquired GTAT and so on. I've always said, customers don't -- we don't win at a customer because of our substrate. We always win at the customer because of our technology. And I define technology as device and package. You have to have efficient technology. It doesn't matter. The substrate doesn't clinically matter. So having availability -- more availability of substrate is purely a supply conversation. It is not a commodity because it doesn't improve nor detract from the quality of your devices or your packages, if that makes sense. So therefore, it's not a commodity. It's supply availability because think about it this way. Today, nobody talks about substrates for silicon. Why? Because it's irrelevant. It's always been irrelevant. It's what you put on the substrate that matters, right? That is the same as substrates for silicon carbide. Okay? Now you buy it cheaper potentially. Okay, great. But if you have bad products, it doesn't matter how cheap it is, you make bad products. And for an OEM, bad product means lower range of vehicle. Well, they're not going to sell it. So what you're going to save a few bucks on a device and then lose it all in market or have to add battery to offset the range, it's irrelevant. So substrates is not where the value is. So that's why I wouldn't talk about its commodity unless your business is substrates, which goes back to your question, that's not our business. Our substrate, we don't sell in the merchant market, it's purely internal consumption. So having availability outside, okay, great. If we can get it, our margin will get better. If it's cheaper, then what we do inside is first point. Number two is you also have to talk about and acknowledge that the primary source of all of this substrate is China, okay? When you're talking about U.S. customers and you're talking about European customers and you say, I have all my supply assurances from China sourcing. What about the geopolitical risk, right? So we're in the best of both because majority of our substrates come from internal. But that means that the nonmajority comes from the outside. We can change the mix depending on the market. So if there's 0 geopolitical risk and we can get them cheaper in China, we'll shift the mix more to external versus internal. We'll grow into it. So we are in the best position we can to navigate that. Customers see it. Customers know that in the event of you can't get substrates externally, we know ON Semi can ramp quickly because we've ramped GTAT 5X in 12 months or we can go outside, we have the flexibility, and that's the best place to be.

What was the differentiation that allowed you to win VW?

So it always starts with the performance of the product, hands down, there's no shortcuts for this. And what does that mean? It's the device plus the package. So just to get some of the terms, we talk about the device as a die. Then we talk about module, where it has 6 to 10, 12 dies in them, depending on how much power. So multiple die create a module. You need 3 module to run a motor in an EV, 3 phases. Those modules sit on a cooling jacket for cooling -- liquid cooling. That's called the Power box. That's why we're -- that's the win at VW. So it's vertical integration all the way to where we match the OEM where we ship directly to the OEM now.

No Tier 1 in between?

For the Power box. That's a strategic win because it highlights the vertical capability that we have, not just we talk about vertical all the way down to substrate. Now it's vertical from a system perspective. It is to illustrate what it is, recall, Simon, in our Analyst Day held up a metal brick, and he said, we have the metal brick that scales up for all the customer power in the pyramid he showed. That's the win. Because it's VW Group across all of the brands, which have different performance parts if you think about the pyramid and that's the architectural win that we got. So that is, again, the proof of with the right products and the right strategy, you are able to get into the market with high value.

And have you guys talked about either the size or the duration or even timing of when that all happens?

The size -- not -- we didn't talk about size. I mean you can figure out the volume from the OEM, but they have talked about the SSP, timing and all of that. So I'll leave that to their disclosures.

Got you. Last question on silicon carbide. You mentioned about one of the things that made it lumpy, but it would be another adoption accelerant was when the infrastructure was ready. Talk about how ON Semi plays into that part of the market? And has that also been somewhat disappointing? We hear some government statistics about the billions that they've spent for, what, 7 charging stations or something along those lines.

So there are different types of governments, put it that way. Some spend way less for way more. But that's a different topic. There are 2 things we have in -- if you think about what we refer to as the ecosystem, the sustainable ecosystem that actually starts with the energy capture. There's energy storage, energy distribution and then consumption, which is the car. So from the energy capture, you're going to talk about utility-grade renewable energy, utility grade and commercial grade. I'm not talking about the lumpy or the volatility in the residential grade. That's not really a business that we have. So primary driver for our business is utility and commercial grade on the capture, then the storage, energy storage. We talked about we have LTSAs with 8 of the top 10 companies in the world that do energy storage. That business has grown 70% '21 to '22, '22 to '23, kind of took a pause. And now that's kind of -- we're seeing an uptick in the second half. So that's growing nicely. And that accelerated beyond our original expectations when we set out the strategy because, unfortunately, the conflict in Europe accelerated that because it was an energy conflict, energy consequential conflict. So that accelerated it. But on top of that, energy storage is necessary because -- and it doesn't matter what country you look at. The grid is an antique grid. It cannot support the charging. They cannot support the air conditioning, depending on what state you're in. And sure enough, it's not going to support the AI drain that we're in and will continue to increase, which means that the way you buffer it is micro griding, which means you put a storage utility in a much smaller than the centralized grid typically we have in North America. So that's the business that we talk about from that. Now -- that's on the storage. Now you talk about distribution, charging. I talked about enabling fast charging will help with the positive catalyst for electrification. At CES earlier this year, we introduced a 350 kilowatt. It is the highest, most integrated module in the world. 350 kilowatt, what does that mean? It means you can get to an almost full charge in 15 minutes. That's a landscape changing approach. If you are able to get to 80% in 15 minutes, you stop at a gas station, you go and buy a bottle of water, maybe use a restroom when your car is charged. That's a very different dynamic than I can only charge it to a full charge at work or at home overnight. Now you just changed the dynamic, and we're enabling that dynamic change. So that's where we plan. We talked about the consumption side, which is the auto.

So in the last couple of minutes, we have going to wrap this up for more financial questions. The structural improvement of the company holding the 45% gross margin at the trough. It obviously be the aspirational peak and now it's the trough.

43% was the aspirational peak.

There you go. I'll give you the 2 points more. So either Hassane or Thad, what are the key drivers of going from 45% to 53%, which is the new target?

Yes. Well, let's first talk about how we get to 45% because I think it's really important. All the work that we did over the last 3 years was to set us up for a market correction, right? And so now you look at where we are today, sitting here 65% utilized at a mid-40% gross margin, historically different than what the company ran at a low 30% gross margin. So as we walked away from $475 million of business, divested 4 fabs, streamlined our operations, that was all the setup where we are today. So anchor point there. So now if you think about how do we go from mid-40s to the 53% target. The primary driver in the short term is utilization. And so I think we're in a great spot today for the recovery whenever this happens and at whatever pace it happens today. So if you look at where we are, we've got lean inventory in the Disney channel. We're at 9 weeks by design. Historically, the company ran 11 to 13 weeks. We've got low inventory in the balance sheet when you look at our working inventory and you exclude our fab transitions and the silicon carbide ramp. And so when you think about that 65% utilization going back up to, we peaked out at 84%. It is for every point of utilization, you get 15 to 20 basis points of gross margin improvement. When the market shifts and comes back on, we don't have to burn through a lot of inventory. We can actually think up the utilization very quickly, depending on what that slope looks like coming out of this. So you'll see that benefit come through in gross margin quickly falling through to operating margin and free cash flow as well. So you got that element. We have the East Fishkill business that we do for GLOBALFOUNDRIES. That's the foundry business as a part of that acquisition. That's 100 basis points dilutive today. That rolls off by the end of '25. So you should think about in '25, is that being linear. So that's not market dependent, it's just time. It's just -- we just have to roll through this contract. And then you've got the monetization of the 4 fabs we divested in 2022. So as we move that production into our network, that's $160 million annualized, and we'll see a benefit. Now some of that will be market-driven because that will be depending on the demand. But we'll start to see some of that in '25, probably a lot of it in '26. And then everything that we're working on in R&D, all new products are at or above the corporate average, so you get favorable mix. So if you start stacking up all those improvements, you get pretty close to that 53% target. And we've always said the 53% is a milestone. It's not a destination. So the structural changes are allowing us to keep the higher lows and hopefully higher highs as well as we go through this.

Well, guys, unfortunately, we're out of time, even though I have so many more questions, but thank you so much for coming. We really appreciate it.

Thank you.