KLA Corporation (KLAC) Earnings Call Transcript & Summary

Great. Thank you. Welcome back. I'm Joe Moore from Morgan Stanley. Very happy to have with me today are Bren Higgins, the CFO of KLA.

So Bren, let's just jump right in. The -- can you talk a little bit about process control in a kind of more challenging environment like we're seeing now. You guys have traditionally seen process control generally rising over time anyway, and certainly in these types of environments, accelerate. Can you talk about the drivers of that? And how you feel about -- you've talked about this year as a kind of 20% WFE decline. How are you guys feeling about your prospects around that?

Sure. Sure. Thanks, Joe, for having us. Appreciate being here. And I'll make forward-looking statements, and those are covered by our safe harbor statements. You can find our risk factors on our website, and I encourage you to take a look at them. So 2022 was a strong year. And if you just look back at what was really the strongest upturn we've seen over the last few years, you look at WFE growth rates that were 21% CAGR from '20 to -- or from '19 to '22. So against a backdrop of an overall industry even with semiconductor revenue at 2x GDP, you would expect to see some normalization. We had a great year in '22 as a company and I know a lot of focus here going forward. But as we look back and we effectively took the company from $5 billion to $10 billion in -- from 2019 to 2022. I think to testament to the scalability of the operating model. Over that time frame, our incremental operating margins were 51%. So we saw a nice leverage and drop through. We saw our share of the overall market improve, and our relative performance last year was very strong. Our service business continues to grow and all this new equipment and all the demand at the legacy nodes is driving more service utilization, which is also good, and that's been good for service as well. And then our EPC business, which is the more than more part of KLA, I like to call it, and it includes our specialty semiconductors, but our packaging businesses, our flat panel business. We had a decent year in '22. It's closer to consumers. So we felt some of the softness that we're experiencing now a little earlier there, but we still had growth near double digits in the system, part of that -- part of the business. So we saw overall share of WFE grow pretty meaningfully over the course of '22. And to your point, over the last couple of years, we've seen it move in part of our Investor Day presentation that we did last June was talking about the drivers of how we would see that continue to grow. We're ahead of that pace, and we certainly saw, again, the strong performance last year. As we look at '23 though, so we're in a more challenging environment, WFE down roughly 20%. And or so. We feel pretty good about our ability to maintain our share of the overall market. If you look at in down environments, first of all, every down WFE year, KLA has generally done better than the overall market because customers in all segments continue to invest in their technology road maps. They adjust their capacity plans, they invest in technology road maps. And so that tends to be good for us. What cycles more is memory. We have less exposure overall to memory. And so you start to see the process control intensity tick up because it's more logic/foundry centric in terms of WFE with roughly 70% of WFE is logic foundry. The introduction of EUV and the scaling that is being driven by that is driving certain products within KLA that are very strong products in reticle inspection and optical pattern inspection, that are also inflecting through this time. And so that's certainly a driver for us as well. Lots of talk about China and China export restrictions. The impact on KLA versus some of our peers is I think a little bit less, and we'll see how that ultimately plays out, and there's a range of impact that we've talked about. And then finally, we also are seeing investment in wafer capacity and reticle capacity in China, and these are for legacy nodes. And so that's a part of WFE that we're exposed to. So I know there's a fair amount of noise with some of the challenges around supply chain last year with some of the companies around. And so the deferred revenue and how that pulls forward or fast shipments and how that pulls forward. In some ways, it's kind of the same aggregate amount of business over a 2-year period, but there's some timing elements that are part of all that. But because of the reasons I cited, I feel pretty good about our ability to maintain the share of overall WFE somewhere in the mid-7s. It depends again on your denominator. But we feel pretty good about that overall. Our service business will continue to grow. It will grow this year. It will be impacted by some of the China restrictions. It's also impacted by some lower utilizations in memory. So we'll still see growth in service. And then I would expect the EPC business, because it was weaker in '22, strength in semiconductor or specialty semiconductor because of power semi, I think EPC will be down less than the overall WFE level. So when you aggregate it all, I think we set up to somewhere if you look at our consensus estimate somewhere in the high 8s and a share of WFE that's roughly consistent with what we saw last year.

Great. Well, there's a lot to unpack there. Maybe starting with the tactical stuff around supply chain. And hopefully, we're getting close to the end of having asked about this stuff. But you guys have a lot more semiconductors in your supply chain than really anyone in equipment and yet you were more unscathed by these challenges than everyone. I guess why is that? And to the extent that we do see, I think, other companies benefiting materially from the deferred that they built up you probably have a little less than in the first half. Can you just talk to how you've navigated the supply chain so effectively compared to others?

Well, so we have different supply chains in a lot of cases. And our supply chain strategy historically, and I spent some time on this at the Investor Day, has been much more about strategic supply chain versus transactional. We have a lot of suppliers that were single source through that work with just us in our space. And so that allows us to in some ways, optimize we're not sort of trading capacity with some of our competitors, which tends to be a positive. We saw our trading capacity across KLA products, which can always be a challenge, particularly if you're a general manager who's not getting your allocation and somebody else's. But we -- in most of these cases, it's fairly clear about what we're going to get in terms of volume. And then at that point, and the lead times are incredibly long. If you think about optical components, you can go have lead times that can be 2 years. When you start thinking about lead time on raw material, calcium fluoride, the lead time on equipment and then through a full production process. So -- but you get to a place where it's fairly predictable, adding increments beyond that is a challenge. But what you're going to get is fairly predictable. Now we start procuring material for shipments sometimes 12 months in advance, and we do a lot to hedge that. Sets those volume levels, and then we do all the work that everybody else does to go and figure out all the different short, more commodity-like parts that you need. And there's a lot of semiconductors there. Now my ASPs are higher than some of my peers. So I need maybe fewer sets of the core electronics and mechatronic equipment, wafer handling equipment and the components that are within that. So I need less of it. So you could argue maybe I'm a little less exposed than some of my peers. As it relates to high-end image computing and data processing, AI and machine learning capability, we weren't challenged all that much in those areas, right? And of course, that's our high-end systems. Again, there's a volume aspect to it. But the biggest challenges we had was just trying to go and find all those microcontrollers and some of those parts that drove a lot of the subsystem movement in mechatronic systems. So look, we did not ship what we wanted to ship. And we still have huge gaps in some of our key products, some of our enabling products like high-end wafer inspection, but also in reticle inspection, where we have this fundamental gap, and we're trying to resolve it. We're getting new capacity, but it's going to take some time to come online. It is a driver of our business today and it some -- a part of the business that I would expect to grow even in this weaker environment where you feel pressures and you're more capacity-centric parts of the company. And in those areas, the supply chain challenge is more limited. So to your point, we are working through it. We've invested a lot to add new capability. Part of that capability was -- we were hoping if you just go back to early the middle of '22, we thought '23 would be a much stronger year. We knew there was an adjustment year out there somewhere, but wasn't quite clear. But we've been making investments to try to position us for this 2026 plan. We want to be able to get to $14 billion in revenue, plus or minus $500 million. And so a lot of investments that we're making in the supply chain to ensure that we have the capacity to be able to support that over the long run. So a, I think it validated through '22, some of the things that we do. And at the same time, we also need to make sure that these relationships, we nurture them. They're important to us. We need these suppliers to be there. So we take the parts. We plan ahead. It's -- I always go back to what are you optimizing for and what are you expecting? I'm optimizing for our business model or differentiation and the custom nature of the parts that feeds that, I'm accepting that I'm going to carry more inventory. I'm not the company with high asset terms, right, high inventory turns, and I never have been. But it's part of sort of enabling the model that what we are optimizing for versus what we accept and we accept that. And there's also the service streams, which is also part of that.

I think everyone in equipment is going to pull a lot more inventory after the last 24 months too. Maybe you could talk to China a little bit. You mentioned that upfront that it maybe affects you less than others. I was quite surprised by a couple of things you guys have said, one, the original impact of the export controls was probably bigger than I might have thought, which probably speaks to how much those guys were planning on spending. And then you said more recently that China probably grows despite the impact of those controls. What's going on in China? Like where is the relative strength coming from to offset the impact of those controllers.

Yes. So what I've said is that I still expect it to be a decent year and down less than the overall market. So I don't know if [ it ] will grow. And we'll see, depending on where we fall in the range that I laid out of $500 million to $900 million in terms of impact from the October export control announcements. And there are some cases where we're working with the government on potential licenses where it might create an opportunity for us to ship to some of the areas of the business that wasn't caught in the -- where the technology lines were drawn. But there's a lot of legacy activity that's happening in China, and there's also a lot of investment in infrastructure in China. I mentioned it earlier, right? We have wafer capacity that's being added, we're getting capacity has been hard for a lot of the Chinese suppliers. Reticles, a lot of reticle houses are captive the reticle -- the customers rely on merchants to supply those reticles. In a lot of cases, the merchants haven't invested as much. And so there's a desire to invest in reticle capacity. Now these are for legacy nodes. So these are for 28-nanometer and above technologies. And so when you add that up and you add up the legacy investment that's happening across a huge variety of projects, a lot of greenfield projects that I would expect that we'll continue to see strong demand in those areas. The impact is -- so what changed in October was more clarity around the logic restrictions and the impact in logic was less and kind of had already been contemplated. There were new rules that -- around U.S. citizens and so on in terms of their ability to support tools. But in general, the overall impact didn't change. There's a service component, and we're losing some service, which is a drag on service, about 5 points of service growth this year is if you lose roughly $100 million of service, which I've said 10% to 15% of the total range. So there's some impact there also. But the rest is the memory impact. And so some of that is pretty clear that unlikely it will ship and then some of it, there might be a path forward, and we're working collaboratively with the government, with our peer companies about trying to understand where the design rules are, where the rules are in terms of where the lines are drawn. And is there a path forward? There's the clear assurances that we're not supporting anything that would be beyond where the technology lines are drawn.

Yes. Okay. And what is the role generally of this kind of reshoring activity? And when you see, obviously, the chips act in the U.S. comparable subsidies that are happening in Europe and in multiple regions in Asia. Clearly, China government policy continues to favor investing in areas where they're allowed to. Is it -- is the government subsidy the key aspect there? Or is it the motivation of the customers to have domestic capacity and that money is helping them to pay for what they're going to do anyway. Just how do you view that? Do you view that as something that's incremental to the equipment industry or something that was kind of a...

Yes, it's more of the latter. It's building the capability here and bridging the cost offset between being able to build in the United States or Europe versus alternatives. Semiconductors are fairly cheap to ship and so where you build them doesn't matter that much, just generally, and so it really is motivated by cost. So how do you bridge that cost delta to build here? Look, the industry has gotten a lot of efficiency over the years out of consolidation and consolidation into these big projects and tighter geographic areas. And so that's certainly been a driver of lower capital intensity. And so as you move forward with more projects that are more disparate, it's harder to optimize across all that supply, there's a period of time where you're scaling potentially multiple facilities. And so you have an inefficiency at 1 period of time. But ultimately, our view is that supply/demand over time is generally going to balance. And there might be an increment here of capital inefficiency because you're spread across multiple facilities and multiple geographies, and there's a spread of talent, too. But I think it's -- in the long run, it's probably in the best interest of the industry generally to have supply demand mostly balanced over time. And so I don't necessarily see it as it's certainly not factored into our plan that it's a huge driver of incremental spend over a broad period of time. Again, it could be episodic as you're ramping certain facilities and you're scaling. But it's really about bridging this cost offset to build the capability in the United States or Europe versus where it might have been built before. So we're not planning for it. It could happen. I've heard some numbers thrown around out there that there's some permanent tier of inefficiency. When we look at it is, if I've got the right products, we're sizing our business in the way that we talked about it if it's there, it's not like we won't participate if we've got the right products to solve our customers' problems. And so that's how we're in the company in terms of how we're thinking about that.

And then looking at more specifically at foundry logic, the market conditions seem pretty challenging for your own customers. You're certainly seeing larger companies with low utilizations, foundry companies at 7-nanometer and below, those utilizations have continued to come under pressure. Do you see that changing the spending pattern? I mean, the companies have kind of given their CapEx guidance doesn't seem like it is. But at some point, do you think there's a risk as it does...

Well, I think CapEx levels or WFE levels are going to be lower in the leading edge, right? And so there are adjustments in capacity. We're not seeing adjustments necessarily in technology road maps. And so that's where KLA tends to participate. We're much more exposed to that than the incremental capacity beyond an entitled yield level. And so because at that point, customers have a comfort, they have yield learning, they have yields, and so they tend to invest less in process control because they kind of know what to manage through those environments. So we do see adjustments that are happening. We modeled in that, we think leading edge is down. If you think of the foundry part of the business, it's down less than the overall market, I'll call it, maybe low double digits. I think there are pockets within trailing edge of continued investment, probably down somewhere similar. And in those parts of the market, I think between automotive, which has held up fairly well, and then the China stuff that we talked about earlier that, that's driving that trailing edge part of the market overall. It's less process control intensitive, sort of at a mature state than the leading edge. But in terms of activities of technology road maps, and expectations in -- at N3 and even potentially some investment at the end of the year and into development. I feel pretty good about where we are from a technology road map, but customers are absolutely adjusting their capacity plans here.

And you mentioned the N2 development. I mean it seems like that's a pretty compelling narrative. I mean you now have 3 companies pushing on 2-nanometer-ish technologies in 2025, where historically you had kind of 1.5. So do you see that activity remaining strong? Is there any area where people are pulling back on sort of more advanced [ arms ] race?

No, not pulling back on the road map part of it? Again, you're seeing some adjustments in leading edge investment from a capacity point of view. But you have the increase in EUV layers, which drives a scaling dynamic, which is great for the reticle business, but also great for pattern inspection. As you move to new architectures, the gate all around architecture creates new metrology challenges. It doesn't create necessarily new small defect problems, but also -- but create some very defect challenges. And across the portfolio of products at KLA offers, we have the metrology capability. We think, for example, you'll see Gen 4 needed more for -- in optical inspection for gate all around than Gen 5 because it isn't as much of a small defect problem as much as the very defect. And so you got a wavelength challenge or wavelength opportunity through the Gen 4 product that you can help measure and manage the defectivity. So I think it's a pretty healthy environment at the leading edge in terms of the [ arms ] race from a technology point of view, and we'll see how it plays out.

And then the role of EPC, as you think about high-end foundry versus low end? I mean I guess I think of it as more skewed to consumer, like you said. But when I look at the most important silicon at the cutting edge, stuff that's used in HPC. There's an awful lot of advanced packaging and an awful lot of stuff that makes your EPC initiatives look really important. So should I think of EPC is kind of helping the whole foundry logic business or one piece versus another.

Well, it's interesting question. So when you think about EPC, you've got specialty semiconductor, which is being driven by power, advanced packaging and RF, although RF is weaker now with the weakness in some of the handsets. And so when you think about silicon carbide and gallium nitride, that's been drivers for the specialty part of the business. But then you also have products and packaging in etch and deposition, but also wafer singulation and dicing. So within the PCB part, you're seeing the substrate part of the business where the substrate is integrating into the package. And so you're just -- what you're effectively doing is just adding a lot more value. And one of the value prop as we thought about the Orbotech acquisition was the -- inflecting that we're seeing in these specialty markets and our ability to go take the KLA operating model, engage with customers at a higher level and drive process tools as much as process control. But also to expose the company more to areas where we are seeing our customers go for value, right? That it wasn't just about front-end semiconductor manufacturing, but also how you integrate, package the chips. So there are new products that are coming that address this sort of substrate market and how the substrate integrates into the package into the PCB board. And so we'll see that grow over time. Most of the spending in the packaging market tends to be in traditional methods. But overall, across the company, we were, I think, in '22 somewhere around $300 million, $400 million of what I'll call advanced packaging. And certainly, the ability to get chiplets, which makes drives design costs down at the leading edge allows customers to only drive the leading-edge capability where they have to be able to integrate it in a heterogeneous package. So it creates a more robust design environment at the leading edge, which is good because that's an important part of our business. And then more value shifting the package and the value that's in that package driving the need for more inspection because you start feeling things that carry a lot more value with connecting layers and things like that create some of these challenges. So it's -- in some ways, it's validating the thesis we had and why we needed to move it, and we thought it was a good use of shareholder capital to go that direction and position the company to be able to go to all the areas where our customers were going for capability.

All right. That acquisition looks better each passing quarter given silicon carbide given the importance of advanced packaging.

The specially semiconductor part of it has certainly inflected a lot more than we originally thought. And I think our ability to engage with customers, and we've developed the channel and we can get meetings as KLA that maybe smaller players couldn't. The service business creates another opportunity that it has the right place designed for markets they're differentiated. If you look at the gross margins, the clearest sign of differentiation in that business and we segment reports, you can see it relative to general process businesses, sort of highlights that they're able to address these opportunities uniquely, and it's a nice growth opportunity for us moving forward. And we've got some new products in development that hopefully will allow us to expand our SAM in that area.

Okay. And then if we could talk about memory. You mentioned that being an area where you see more concentrated weakness. My thinking has been we would get to the bottom of spending fairly quickly just because the margin went away for your customers so quickly. But it's continuing to get worse. I mean even as a kind of bearish [ personally ] on memory, I've been sort of shocked at where gross margins are going. So do you think there's risk of incremental compression as you move through the year in the memory space.

Well, we aren't leaning into it, right, in terms of our assessment and how we size the year and how we're thinking about that part of the business. You're right, it continues to be a challenging environment. And one of the challenges in that market is that there's more capacity adds to meet long-term bit growth supply. And then historically, shrinks were able to address most of it. And so these -- our customers have the balancing act of moving technology road maps, but also having to be in a position to supply when growth resumes and they're kind of thinking about it in different ways, certain customers. But yes, it's weaker than I thought it would be. I don't think it's impacted our business all that much, given we're less exposed to memory and the technology investment still seems to happen and should happen. I'm having more payment term discussions as a CFO than you might imagine in this kind of environment. And look, we try to do what we can to support our customers.

Makes sense. Okay. I have 1 more question, and then we'll open it up to the audience. Just if there's any. In terms of services, you talked about that being an area of kind of relative stability, I think when we look across all of the top 5 equipment companies, the services is growing faster than the installed base of tools. So obviously, a lot of that is intent. A lot of that is you're building new services capabilities on top of that. But what's happening -- what does that all look like from the standpoint of customer spending? And is this idea that services can be continue to grow relative to the size of the installed base, is that something...

Yes, it's been almost 3x, right? If you just go back about 5 years for us. In our service business, we all kind of roll it up differently, but our service business is traditional service. There's no equipment in our service reported service revenue. And about 75% of it is contract, which is service contract. And in a lot of cases, they're tool specific contracts. In other cases, it's fab-wide contracts. So it provides -- the other part of that is not just the growth but also provides a more predictable stream of revenue. It's a stickier stream. It carries a profitability level, we believe, is accretive to the overall. So it's a nice anchor for the company over time. It's only had 1 down year in revenue in the last 20, 25 years, and that was in 2009, which was a unique environment. And even then, it was only down about 12%. So you've seen it in a couple of ways: a, the value of the tools and the complexity of our tools. And so the percent of ASP that's tied to service contract is slowly rising a little bit, but you're seeing that the installed base of tools live longer, and that's been a driver. And then you've got the growth of WFE how much we've shipped and then those tools go into service. And so -- all that got us to a point where -- and we talked about it at Investor Day going from a 9% to 11% long-term growth rate to 12% to 14%, and we were at the top end of that in 2022 despite some of the China headwinds and some FX headwinds. So over the long run, we think 12% to 14% CAGR and with the profitability stream I mentioned to continue. And because of all that, that's certainly been a factor for us for decades and getting more and more comfortable about not only our dividend strategy, but how we finance the business in terms of our leverage strategy as well.

Yes. And that's a great business. Can we have questions from the audience.

Quick one. How are you thinking about investing in the product and also investing in new areas in terms of thinking that there's more competition in the industry, definitely from: one of your pure plays like Laser tech, but also ASML, AMAT all been investing a lot into metrology, et cetera. So how are you thinking about the overall portfolio, how you think about making more investment into [ Southeast ].

Yes, it's an attractive market, right? And our margin profile, it does a couple of things. First of all highlight the value that we bring and also that I think highlights the differentiation the company has. We're 4x our nearest competitor in our space. And the relative growth of process control over the last few years is certainly attractive to anyone in the space. We go to market in a -- with a portfolio approach. I mean one thing that's great about KLA, which drives our investments is that we are, in some ways, agnostic of which tools we allow our customers use for meeting their either technical requirements or their economic requirements, which change as they're moving from defect discovery and development that are going through a ramp cycle and then monitoring long-term capacity. Most of our competitors are point product competitors. And so they've got to sort of come up with ways that, okay, depending on where road maps move or the maturity of processes and designs move that this solution has to try to meet multiple application requirements. And it -- I think, validates the fact that having the portfolio allows us to compete in a way that's unique in the market. We also have the ability to leverage a lot of the data that comes out of those tools to have a network effect across the tools, which creates a competitive advantage for us. We do have a competitor in a market where we're investing in the long run. We think that there's a portfolio approach to solving the challenges today, and this is in the EUV reticle part of the market. it's a market and reticle inspection that will probably gain 7 or 8 points of share in over the last year because what's available in the market today that meets technical given the density of the transistor design on the reticle in the layers of the pitch of the lines and spaces. And the economics of running the fab is playing to our optical solutions and our wafer inspection solutions to validate reticles in the fab environment. So clearly, that's playing out in the EUV space, 90%-plus reticles are running through KLA systems. We have new capability in the market that as they start to shrink feature sizes, that's an e-beam, electron beam-based capability that should meet those higher end post writing inspection requirements. And we feel pretty good about the strategy and the introduction of new capability, similar to what our competitor is offering and trying to engineer today when you move into a high NA environment because High NA requires that you'll see more reticles, you'll see more pelletization which is a factor driving reticle inspection. And so a lot of investment in the company to meet that timing. If you look at the reticle space, for EUV, the company. I think at the end of this year, I was adding up the numbers as CFO, you kind of roll your eyes you go holly -- wow, over $1 billion across multiple products to solve the reticle challenge in terms of R&D investment. So this is a core market for the company, and we have some unique capability. We have capability that allows our customers to balance. So we feel pretty good about our road map.

Great. Unfortunately, we're out of time. We have to wrap up our [ hypothesis ]. Thank you, everyone.

Thank you, Joe.