KLA Corporation (KLAC) Earnings Call Transcript & Summary

Good morning. I'm Tim Arcuri. I'm the semi and semi equipment analyst here at UBS. And very pleased to have KLA. Very pleased to have Bren Higgins with KLA, who is the CFO. Thanks, Bren.

Yes. Thanks for having me.

Great. Well, Bren, let me start by asking you what I'm asking everybody and just to talk about the wafer fab equipment market generally. You're sort of pointing to the first half of next year as are most being modestly up, flat to actually modestly up is what you said. And then having an acceleration into the back half of next year. Can you talk about the drivers? Does it have to do with DRAM fab readiness? How does the loading of each vertical look next year?

Sure. So a lot there. I'll first start with just 2025 has been a really good year for the company, right? We've seen our share of market go up. We've seen mid-teens, if you take the midpoint of our guidance, mid-teens revenue growth, high 20% EPS growth, stronger share of overall WFE. So we really feel very good about how the company is executing, how it's positioned from a competitive point of view and really excited about some of the drivers that we think that continue as we move forward. We've seen some of this over the last year or 2 in terms of what's happening with high-performance compute and what it means to process control. The nature of a very robust design environment has always been good for process control. We went through a period of time in the last decade where limited number of designs at the leading edge. Customers could debug a process very easily with one particular design instead of trying to debug across multiple designs. The follow-on designs weren't happening, so there was a reuse element. All this was not great for the process control market or process control intensity. But from 7-nanometer forward, we've seen a really strong design environment. And so each product has its own, I'll call, kind of ramp cycle. Customers have to be efficient with how they're managing their capital. They don't want to start too much inventory. They don't want to start too little. They have to deliver to a tight market window. The node gets consumed. So as they move to invest in the next node, don't really use a lot of the previous node. So a design environment that's robust can be really good for us, particularly as you think about the whole kind of node maturity over multiple years because of just the challenges of managing that kind of environment. Large die. Large die benefit KLA in a unique way. It doesn't really change the process world very much, but large die has a defect density challenge. Your yields are lower with larger die. The value of the die doesn't change a process market means a lot for process control because our customers will invest more to ensure that, that value is captured and managed. Process variability is a challenge. You can't -- for very rigorous performance requirements, you have to have a very predictable process. So large die can be -- is very good for KLA, and we see that changing, particularly as it relates to high-performance compute as a driver for our business in logic. High-bandwidth memory has its own requirements that have been good for process control intensity. And then you've got advanced packaging, which is growing as an opportunity. I'm sure we'll talk about some of these things. As you look, though, at what's driving leading edge, and I think it continues into next year, is you see a broadening of investment at the leading edge, but you see 2-nanometer node investment, you see a broadening investment, as I said. High-bandwidth memory continues to happen. I think that drives growth into next year. I think it probably grows a little faster than overall logic/foundry grows next year. NAND flash will grow a little bit, but off a low base. I think China is kind of sluggish, at least in terms of the parts of the market that we fully understand. There's parts of the market we can't serve. So as we look at next year, we haven't quantified it, but we feel very good about the growth across all the market segments. Like others, we feel that first half is lighter than the second half. I think the first half grows versus the second half of '25. I know I said flat to modestly up. I think it's probably more in that kind of modestly low to mid-single-digit kind of growth rate into the first half of next year. So we're seeing some acceleration from customers in terms of engagement. I still think the second half is bigger than the first half, but certainly feeling more confident with some of the engagement. We're seeing lead times extend. So pretty excited about what's in front of us here as we look at next year.

So it seems to me like the next 2 years, whether it's next year is a great year, whether it's 2027 is a great year, it depends on when the fabs are ready and whatnot. But it seems to me like maybe $150 billion in WFE is not an inconceivable number sometime maybe in 2027. I don't want to pin you down on an actual WFE number, but would you generally agree that the next 2 years should be sort of a -- I mean, these should be the best 2 years we've seen just in terms of the year-over-year growth rates in quite some time. Is that a fair statement?

I think the next 2 years based on customer conversations, and I think some of the facility constraints that exist this year, you're right, we'll see -- I think if they had opened facilities might be able to add more in the current year. But I think over the next couple of years, it feels based on customer conversations today to be a pretty strong environment. There's the WFE level and then there's the advanced packaging market. And what was a rounding error in WFE up until just a few years ago is now a pretty meaningful market. And I think as we think -- talk about the size of the market, we have to kind of think about both of them now, right? So like if you look at this year in 2025, WFE is, we'll call it, somewhere in the -- between $105 billion and $110 billion probably. But advanced packaging is another $11 billion on top of that. So sometimes when the numbers are out there, I think sometimes people -- it's not clear whether they're including one or the other. I think now you have to include it because it's a growing TAM. I think it's going to grow faster in WFE, and it's a meaningful number now.

Yes. Okay. And let's talk about your share of that WFE TAM. You gained another 60 basis points last year. Do you think you can gain share again next year given all the dynamics that you talked about?

I think we feel very good about the trajectory of our business, what's happening in these markets. The broadening of investment, as I mentioned, the leading edge will be good. The leading edge has been very efficient over the last few years. But I think as you move into next year, leading and near leading edge will be a higher investment level. So I feel very good about that. Process control intensity, high-bandwidth memory is higher than conventional memory. And so you have some conventional memory investment that's going to happen, too, to support that market as we've seen pricing change meaningfully in that part of the market. But high-bandwidth memory is also pretty positive, I think, in terms of its effects on our business. So over the last couple of years, we've seen the mix shift a little bit from -- it was close to high 60s percent in 2024 of WFE that was logic/foundry. 2025, I'll call it, kind of mid-60s. I think as we look at 2026, it's probably somewhere around 60%, maybe low 60s, but in an environment where you would see memory as a greater percentage of total, you think they put pressure on our relative performance. But I think what's happening at leading-edge logic plus what's happening in high-bandwidth memory is mitigating some of that effect, and we've had pretty good relative performance over the last couple of years as a result of all that.

And how do you think about foundry? I mean, this year, you've monetized a lot of the N2 dollars, not to say that you won't also get N2 into next year, too. A14 starts maybe at the end of next year, so you might start to monetize some of that at the end of next year, but that's more of a 2027 thing. So how do you think about each of these nodes? And is there a significant step-up in process control intensity, say, in A14 versus N2 versus N3? Or are they all stepping up at a similar rate?

Well, so N2, and we've been pretty vocal about just our overall view of KLA's share of WFE at the 2-nanometer node has been at least 100 basis points higher than N3. And as you think about the first 3 to 4 years of a node, the first 150,000 kind of wafer starts is how we do that modeling. Over time, you see higher levels of investment typically on the front end and then it tends to moderate a bit. Historically, it used to moderate a lot. But now because of this design dynamic I talked about and the shift in mix in terms of value of the parts, we're seeing that rollover effect is more minimal now. So as I look at next year, the predominance of investment, again, I talked about broadening investment, but you're going to see a lot of N2 investment. You'll see some N3 investment. And you're right, you might see next-generation investment at a very kind of small level towards the end of the year, which is kind of typically how that customer will start to develop and work on the next process node. So as I look at the year, I think that the logic/foundry part of the business steps up in the second half, and I expect growth really from all the customers that are investing in leading edge.

So I've always thought of it as over the course of a node, you get roughly 70% upfront and it falls off in year 2. It's like 70-30, whereas for a films company, it's more the opposite. Are you saying that it's more evenly balanced for you now?

I wouldn't call it evenly, but it's much closer to that. So I would say that I expect overall from my customer that this year will have an up year versus last year.

Great. Let's talk about China for a second. So you and all your peers see China down next year. I don't agree. I think it's up. But what are the drivers for it to be down? And then as part of that, can you talk about the BIS affiliates rule, which was put in place and it cost you $300 million out of your December quarter revenue. And can you talk about the dynamic of that as you allow those companies, you give them a year reprieve, they're going to probably try to get as many tools as they can as fast as they can. So that seems like it could be a real upside driver for you next year.

Well, if you look at the overall market and part of what's -- China has been interesting, if you go back to '21 and '22, you had our non-China customers, you had the supply constraints and consume generally most of the capacity. You had a lot of greenfield activity. And so when you saw other parts of the business cycle in '23 and '24, you had this backlog of business that you were able to ship to. And so that drove China investment from kind of the low 20s into the mid-30s in terms of overall WFE. Now there's a big part of that market that is not accessible. So when you think about the whole market, you got to -- not accessible to U.S. companies. You have to try to say, okay, how much is that, but it's a little bit hard to figure out. We try to do our best to figure it out. So as I look at this year, I think it's kind of modestly down. I think next year, it's kind of flattish. So I wouldn't say that it's falling off dramatically. I mean the rest of the business is growing. So its percent of KLA is declining. It was 40% in '24, 41%. Over the course of this year, I've been pretty explicit about our views that it would come down to somewhere around 30% this year. And I think next year, it looks like it's somewhere in the mid-20s. And I said this on the earnings call. The BIS rule change did affect about $300 million, and we'll see how that -- and that was $300 million basically of business we expected between November when the rule dropped and the end of '26. And so we're working with that customer to reslot that business. I would expect that to come back. I don't think it changes the overall WFE because I think that was WFE that was likely going to get spent anyway. It just wouldn't be spent with U.S. suppliers. But that certainly changes my view of overall China where I would have thought China would have been down for KLA down 10% to 15%. It's probably down overall about 5% to 10% with the overall market down less than that. So that's how we see it. I think that business will come in, and we'll slot that over the course of the year. And our engagement with the customer has been generally regular engagement, so around those projects. So I don't think there's anything other than trying to work real hard to get these tools slotted when there's a lot of demand and interest in getting slots earlier. I mean we're certainly seeing, I think, more measured investment on the logic side. But on the memory side, it definitely feels like there's more momentum to try to do more earlier.

Yes, we've actually heard that. So during this conference, we've heard a lot of talk about the memory companies coming in and pretty much 24 hours a day asking for pull-ins and when can I get tools earlier, particularly NAND, but also DRAM. So can you actually speak to that?

Yes. So I'm seeing it more on the -- now NAND isn't our strongest market. So I think the engagement level in NAND has been okay. And there's a certain level of business, but it's the smallest part of KLA. But on the DRAM side, we're definitely feeling it.

And your ability -- I mean, given your lead times, your answer to those customers is like, I can't do a whole lot for you because I'm already given away those slots probably. But that probably helps your possibly March, right, and certainly June as they're trying to pull stuff in.

I think we work really hard to try to accommodate all our customers' needs and juggle our slots to be able to deliver to what they require. We have a rule of KLA. We don't lose business because we can't deliver. So we figure out a way.

Can we talk about competition in China? So you're pretty unique because the films companies, they do have domestic -- their domestic alternatives in China. For you, there aren't. I mean there's Skyverse, but they're not shipping that much. So can you talk about sort of the threats that you see in the domestic Chinese market? Are these companies shipping enough to make you concerned?

Well, so because of the nature of process control, the complexity of our systems, this probably applies to lithography at some level also. The level of competition is more at -- and most of the investment in China is at legacy nodes. And so where you do see competition is at very legacy nodes. You see some of that business at these restricted facilities. And then you might see some of that business that will show up as customers in China will buy or try to have some of the competitive offering just as a business continuity measure because they never know what the future might mean, and they've got to run their facilities. The level of business has been pretty modest. If I look at sourcing Gartner, I think Gartner suggested somewhere between $200 million and $300 million in 2025. So we see some of that presence. I wouldn't say it is super formidable at this level. But you're always concerned if your competitors can engage with customers and start to learn and try to leverage that in other places. So we're sensitive to it. I think where we can be directly, it's less of an issue. I think where they're winning is in places we can't serve.

Great. And then foundry. So obviously, there are 2 main foundries, but there's another large company who's making a big push in foundry. So maybe foundry becomes a broader market for you. So what are your assumptions around sort of the breadth of the foundry market? I mean, obviously, more customers who are doing foundries, better for you.

Yes. And that's -- as I said earlier, I think the efficiency of the leading edge, there was a lot of maybe inefficiency in legacy nodes and all the investment that was happening in China and other places. But the leading edge has been very efficient. So I think -- or near leading edge. As you start to move into next year, we do expect, and it's been pretty well publicized that we'll see higher levels of investment off pretty low basis, but it's certainly gives us some confidence as we're looking at next year of this broadening investment. And we think at the end of the day, competition is good in the market. And I think some of that will play to KLA. I think there's a lot of opportunities for us to engage and collaborate at a higher level. And so we're excited about working with those customers and in some cases, new leadership teams to try to make some progress here in terms of competitive positioning in this part of the market.

Great. You talked before about advanced packaging. And I believe you said it was $925 million in revenue this year. It's up 20% year-over-year. I think 1/4 of that is SPTS, which has actually been a bit soft this year. Can you talk about how big of a business advanced packaging can be for you? And there's all this discussion about panel level advanced packaging. How can you play there?

Yes. So it's -- we've -- and we've been upticking over the course of this year in terms of expectations of revenue this year from advanced packaging. We talked about it being about $925 million across the total portfolio, 70% of which is process control. It's actually up versus like a little over $500 million. So 70% year-to-year from last year. And we're seeing most of that happen in the logic part of the business. We're seeing a lot of strength there as our customers are really integrating these new -- basically integrating a GPU and HBM device into what is effectively a system. The amount of processing that's actually happening and the sample rates to ensure that these valuable die and valuable memory is functioning as designed. As I said, the process variability, performance specs are very high. So it's been -- it's driven this part of the market to the need for more advanced capability. Years ago, our presence in the back end was fairly limited what our -- how much value could -- given of the devices our customers are making, could they add in terms of process control make sure that the device was functioning. If it was a mobile chip, not as much, right? And so the investments and the need for higher-end capability wasn't as high. But as you move to high-performance compute, you have very expensive die, a lot of areas of risk. And so how much are you able to invest as a proportion of the value of the underlying thing that you're inspecting has it been a change in the market. And the requirements are becoming more complex. So in some ways, and our customer came to us and said they needed our front-end capability in their packaging fabs. And so we've been able to leverage the front-end portfolio to be able to meet these market requirements. And so we've seen nice share gain on the logic side. We're seeing good momentum on the memory side, some different processes in memory. As they move to different -- because of the size of the die, they move from like wafer level packaging to panel-based. It's -- they're seeing similar kinds of market dynamics and market opportunities. It changes some of the engineering we have to do to be able to handle the larger panels. But the fundamental architecture of the systems doesn't change. And so we feel very good about how we're positioned with the portfolio that currently the lowest end part of our portfolio is addressing this market, but there are additional capabilities and systems that are going to be panel capable that will also be able to meet these increasing requirements. So we feel very good about our market position here as we move forward.

Great. So one sort of structural change, it seems is that your backlog is a little higher than it used to be. So I've covered the stock for 27 years, I think it is. And I don't remember -- I mean, sure, like in the last 2 years, your backlog totally ballooned, but it's back down to 7 to 9 months, and I think it's going to stay in that range. And that's a couple of months more than you used to have in backlog. Is there something structural going on? Is this -- China is building all these new fabs or really everyone is building new fabs. And so they have to get in line earlier. Is that part of the dynamic as to why you have more structurally higher backlog?

I think it's part of it. And certainly, we saw it go way up because you had greenfield fabs. And if you're building a fab, the last thing you want to do is finally finish that big construction project and then not have the tools when you need them. And so -- and if you were a customer that was a relatively new customer to KLA, we would want orders, we want deposits and things like that to make sure that -- to get comfort around that business. So that's come down over time, particularly to the dynamic I mentioned about what happened in '23 and '24. We're now in this 7- to 9-month range, closer to the higher end and maybe potentially going above that, given some of the greenfield activity that's expected out there. But it historically was about 5 to 6 months. So 7 to 9 months is actually a little bit more because there's more greenfield activities, it's more customer -- a broader set of customers. Our long-standing customers, we've always tended to operate with a build-to-forecast model. And so we stay very close to them and understand their requirements. Having the backlog isn't necessarily something I need to get confidence in what I'm building for some of our major customers. So as you see more of the business coming from them, you're not going to have as much backlog related to them moving forward. Now I think there is some discipline given some of the history to understand that the more visibility they can provide for us, the better because we have to supply an entire industry and everybody has their piece of delivering to meet these market requirements. So the more visibility we have, the better we can plan for from a capacity point of view. Our material lead times are pretty long. When you're building optics, it takes some time to have capacity. And so more visibility is better. But at the same time, we're pretty close as -- given our position in the market with these customers. And so it's not so much of a factor in terms of how I think about what we're going to plan for.

Got it. Can we talk a little bit about gross margin? You've done a nice job. You guided 62% for the quarter. What are the puts and takes on gross margin? I mean if we're going to have these banner 2 years, where do you think gross margin can go? And what are the puts and takes on that?

Our pricing is much more related to what we're selling than where we're selling it, right? So every customer buys different products. We have a portfolio of products. And so the margins can vary across the different products that we have. So that tends to be quarter-to-quarter, the biggest driver of any variability in our gross margin tend to come from just those mix dynamics. The packaging market has been a great opportunity as we talked about, but it's been dilutive to corporate averages from a gross margin point of view. Obviously, we've had the tariff headwinds, which are closer to 100 basis points today. I said 50 to 100. I think we can mitigate some of that in some of our operating processes around how we move parts. We have a lot of repairable parts, and we do -- we have a factory in the U.S. So as we move parts back in the U.S., you kind of incur tariff exposure. So you think about how can you do that more efficiently to kind of avoid tariffs where it didn't really matter all that much before. And so we just did what was -- what made sense. But over time, obviously, the structural tariff environment cause you to look at some of these processes differently. So I think as we move through '26, we'll see that impact come down. There's the lower end of that range. I'm hopeful as we get in the second half of the year, it's towards the lower end, maybe lower than that. But I think structurally, we're going to have a little bit of headwind related to that. Service grows faster. Service has dilutive gross margin, but kind of contemplated in the models that we have and that we share with investors. So our incremental operating margins tend to be between 60% and 65%. I don't have any reason to think that they wouldn't continue at those levels. And right now, we're going to be somewhere in that mid-62% range this year. And I would expect as we're looking at next year, and we'll have more guidance at earnings in January. But I think we're probably operating probably at 62.5%, plus or minus 50 basis points as we look at next year.

Great. And let's talk about service for a moment because I remember not too long ago where the films guys service businesses were growing faster than yours. Yours was growing high single digits. theirs was growing low to mid-double digits. And now that's kind of flipped. So yours is growing 12% to 14%, possibly even 15%, theirs is growing less than yours. So what has changed in that? Are there different strategies that you're employing? What's driven that?

Yes. It's interesting. Our business is very different, right, because it doesn't really have a big consumable element. It's also all service revenue. There's no upgrades or other systems kind of business in any of our numbers. So -- and it's -- 75% of it is contract, so service contracts. So our service business or our business in general is a very high mix, high complexity, relatively low volume business. And as a result of that, is very difficult for our customers to do self-service. A lot of the parts are custom parts. So the supply chain for those parts is us. And so we're able to manage that well. What customers buy from us is they buy performance of our systems and performance means specs, it means matching and availability. And so while we do break and fix, what we do -- what we really sell is the performance of the system and the availability of it. And so because it's -- customers spend what they want on process control, what they need and then they run the tools at a very high level of uptime, they rely on us to keep them up, and we price the contracts with different elements of our delivery system, and we can kind of monetize that in terms of the support levels we offer. And so that's generally how they look. And there's no -- there's very little redundancy. So if you go into an environment where things slow down, they don't really take process control offline because it's the most capital-efficient thing they can do, right, is to ensure that they have yield and predictable outcomes in the fab. If you had asked me in 2013, what our percent of ASP was then versus where it is today, I'd say we're 150 basis points higher today. So prices have gone up, the price of the contracts have gone up. If you ask me the average life of a KLA system, I would have said 10, 12 years, we're in the high teens now. So we're adding tools to the installed base much faster than they're coming off. Markets like high-bandwidth memory, packaging, given the sampling that's happening and the performance requirements, their customers are now looking at those opportunities differently from a service point of view. So that's a nice tailwind. Acquired businesses, we've done pretty well with taking acquired business and putting it in our system and driving more value out of service than they could independently because we have the scale and the engagement. So all those, despite the impact of the BIS rules as it's related to some market access and certain fab access, we've been able to grow service in that 12% to 14% range. I was looking back 2022, we've set out a growth plan for service for 2026. And as we're doing our planning cycle, I'm looking at 2026, and we're pretty much right on the number despite the headwinds as it relates to the export controls. And it carries what we believe is an accretive stream at the operating margin level. So we feel good about the trajectory here. I think it continues at these levels. And it's a nice anchor and predictable stream for the company that went down here in 25 years. If you look at how we think about capital structure and capital allocation in the company, part of it over time has been the presence of this service business and our ability to rely on it in terms of its cash flow generation. So I think it's a less understood but becoming more understood aspect of the story that's pretty important.

Do you think that it can keep growing at 12% to 14%?

Yes.

Great. Well, thank you, Bren. We're out of time. But thanks again.

Thanks for having me.