It is 2024 Global TMT Conference. I'm Elizabeth Sun with Citi Research, and for this session, we're pleased to have John Lee, CEO of MKS Instruments, and Paritosh, the new IR. Welcome to be with us. So this session is for Citi clients only. So John, thanks for joining us today. I guess we'll turn over to you to have a brief introduction of MKS and provide any, like, opening remarks, and then we'll get into Q and A's.
Great. Thank you, Elizabeth. Thanks for inviting us here. Some of you may know the MKS story well, and some may not, so maybe just as a background to set everybody to the same level. You know, MKS historically was really in semiconductor equipment only, just, and in fact, it was a subset of that. It was vacuum-based semiconductor equipment. And in the last 10 years, there's been a big transformation of the company, and it started off with the acquisition of Newport Corporation, followed by the acquisition of Electro Scientific Industries, a laser systems company, and a laser subsystems company, Newport. And then it followed with the last big transformation, which was the acquisition of Atotech, which was a chemistry plating and chemistry equipment company.
These were transformative acquisitions, and all tied towards the vision that we have, which is enabling and being foundational to advanced electronics. Our thesis is that advanced electronics, while it has been historically driven by advancements in semiconductor chips, that is still necessary, but no longer sufficient. We need advanced packaging to put these advanced chips together so they can talk to each other much faster. I think that's a story that was, you know, something we told about two, three years ago when we justified the acquisition of Atotech. I don't think a lot of investors really got that then. It's like, "That sounds like it might be true, it might not be true." More recently, I think everybody understands that packaging is just as critical as the chips. You know, the example, of course, is AI, NVIDIA, right?
When that happened, and you see NVIDIA's CEO always holding up boards, right, with lots of chips on it. His chips, and some are not his chips. But his solution requires that all these chips work together, talk to each other very quickly, very fast, and that board is what enables that. And that board, while we have maybe thought about it as a PCB, it's old technology, I don't know how hard it is, it's really different now. That's the change that we saw three or four years ago when we said advanced packaging is taking on the same historical attributes as semi: miniaturization, complexity, harder to do, more layers, larger boards, and smaller features. And as a result of that, it's going to be harder and harder to drive the next generation of advanced packages.
And when things get hard, that's what we love as MKS, because we're wired to do these things because of our 60-year history in semi. And so we now have, you know, the major components, in fact, that we need to enable advanced electronics. So that's the transformation we've seen and driven over the last 10 years. And I think we were a little early in keeping on that transition, the acquisition of Atotech, but I'd rather be early. And we actually did something about it because we saw this opportunity in advanced packaging.
Great. Thanks for the introduction. So let's get into the questions. So I want to start with some questions on the overall market. In the June quarter earnings call, like a month ago, you slightly adjusted down your outlook from like two, like now is second half to be flat with first half. So without going into too many details, like, what changed your outlook compared with a quarter ago, and is there any changes compared to the month ago?
Yeah, I think that's a good question. I think as we talked about on earnings call, we had a slightly better Q2 than we had thought we were going to have. A little bit of that, but that wasn't the real reason why we're guiding second half to be similar to first half. It's really that the recovery we thought would be occurring in the Q3, Q4 timeframe, though muted, that has pushed out, and I think that's not news to most of you and most of you following WFE, and that's consistent. That's what we're seeing. Our visibility also is slightly lower now because our lead times are back to kind of that six-to-eight-week lead time. Our customers know that, obviously, and so they order knowing that they're going to get products from MKS in that six-to-eight-week lead time.
That's within the quarter, obviously, and that was a reason why we have less visibility, but also a reason why we had a bit of a upside surprise in Q2. There was more book and turn within the quarter we could execute on. We guide our best for Q3 from what we see, and, you know, we make an assumption on book and turn within the quarter, and that's our guidance, but what we see now is, you know, the upturn that we were hoping for in that Q3, Q4 timeframe is a little more muted, a little more pushed out. We probably were hoping for a little more NAND to come in sooner. That's not happening in 2024. DRAM is happening, but maybe not as, you know, as fast as we would've liked, and logic continues to motor along.
So that's really a slight adjustment from first half, second half.
Got it.
Yeah.
Could you remind us your end market exposure to memory and logic foundry? And I think you previously mentioned, like, I think it was last year here at Citi Conference, you said you were about 50%,
Yeah
... exposed to memory, with, you know, memory down for the last two years. What's the end exposure right now?
Yeah, so the comment we shared with you last year was in 2022, our memory to logic exposure was more memory, 55%-45%.
Yeah
... 55%. And a lot of that was driven by NAND. Now obviously in 2023 and even in 2024, NAND is, you know, in a correction for sure, and DRAM was in a correction that's kind of, you know, coming out of it with, driven by HBM. We also think NAND will eventually come out of it, because when you're doing AI boards, you need solid state drives to enable those AI servers as well. So in 2023, it was probably more like, two-thirds, one-third-
Okay.
more logic
Okay.
- than memory, just because both DRAM and NAND were down, and '24 is kind of similar-
Okay.
to that, and even though that one-third of memory is still one-third, it's still mostly DRAM.
Okay.
And so, you know, we still are happy with our position in NAND. When it comes back, you know, we'll enjoy that market leadership that we have there.
Okay. Yeah, so I think you're known like for kind of over indexing NAND, like, in the past. So, like, sorry, more DRAM right now. Like, what's a normal level of your exposure like?
Yeah, you know, I think, you know, it probably depends on, you know, your view of where NAND will be long term-
Mm-hmm.
- in terms of its percentage of WFE. We have a view that it's probably more around fifty-fifty memory logic long term. If NAND takes off again, of course, that will be a little more, overindex.
Yeah.
I, I'm okay with being overindexed. It's okay, especially given the market leadership position we have in RF power and enabling NAND. But more importantly, you know, beyond all that, you know, we've had this long history of outgrowing WFE, and that's really about getting design wins in areas where we're underindexed, such as-
Mm-hmm.
- Lithography, metrology, inspection. And then taking advantage of inflections with a broad portfolio.
Got it. And do you see any green shoots in, you know, NAND recovery, or
Yeah.
What signs are you looking at?
Yeah. Not yet, frankly, in NAND. You know, obviously, we've talked about inventory burndown is finished.
Mm.
For everything but some of our key products in NAND. And so, certainly some of it's being burned down slowly, but it would take a, you know, kind of a change in the NAND CapEx to burn down inventory and then, of course, new orders for the NAND equipment. So right now, NAND still feels muted, and, you know, a lot of folks are thinking twenty twenty-five.
Yeah.
That's moved around, early 2025 to late 2025 to middle of 2025. So, right now we're certainly keeping a pulse on it. Obviously, we have very close relationships with all our customers there. Not just direct customers, OEM equipment OEMs, but also the chip fabs. And, you know, they know who we are. They know that, if they're gonna have a plan to, you know, build out a NAND fab, they've got to get their whole supply chain lined up for it, and we'll, you know, we'll get those notices.
Okay, that's fair. So, you guys have a wide range of products, and you have this strategy called, like, Surround the Wafer, if I'm correct?
Yeah.
You've talked about you are, like, exposed to 85% of WFE equipment. Could you talk more about, like, where are the main areas of strength for MKS and what will be the main growth driver?
Yeah, you know, I think just to remind everybody, we've talked about being exposed to 85% of all WFE. So every fab in the world that's built, the 1,000 tools in every fab, 85% of those tools, we can walk around that fab, and you will see multiple MKS pieces of equipment on 85% of the tools. No one, no one else can say that, right? And frankly, it's simple math. Just take Tokyo Electron, Lam, Applied, ASML, KLA, the Big Five, all our customers, and that's 85% WFE. And then we have some more with wet clean as well, so we don't even count that. And so this exposure allows us to see inflections much faster. We're in talking to all the customers about, you know, their new challenges. And so where are the opportunities for growth?
Fundamentally, you know, sometimes these inflections happen, and no one really, you know, expected it, like V-NAND. And then, if you have a broad portfolio, you can do something about it. If you don't have RF power, that's gonna be the critical subcomponent for enabling V-NAND hole etch. Even if you understand you need it, you can't do much about it, right? But I have RF power, so I can just decide if I want to do that, and that's occurred over and over again. And then, the other area for growth, I believe, is our world-class optics initiative. So world-class optics means we have been underindexed in terms of market share in lithography, metrology, and inspection, and that's fundamentally because the company we bought, Newport, had those customers. They brought ASML and KLA and those kinds of customers to the MKS portfolio.
But Newport treated those customers as more of, you know, one of thirty good customers. We don't want to be overindexed to semi. They kind of understood semi goes up and down. We don't want that. Having bought Newport, we were different. Our approach is, if I get designed into those two customers, are they gonna lose their share? Unlikely, right? So let me get designed in. And to do that, though, we had to invest. We had to invest more in equipment. We had to invest in process engineers, because that equipment everybody can buy, you know, equipment to process optics. But the recipes that are developed on those pieces of equipment require process engineers. That's the IP. So kind of like a chip company, right?
You can buy the equipment, but the process engineering is really kind of what the chip company adds, similar to that World-Class Optics initiative. And because we've done that, you can see the numbers. You can see our revenue in lithography, metrology, inspection over the last two or three years has outgrown WFE for lithography, metrology, and inspection. Not just WFE, but just those. And I think you also know that lithography, metrology, inspection, so it's much more stable CapEx revenue because the lead times on EUV are quite long, inspection tools quite long, and so they don't go through some of the cycles that some of the vacuum equipment companies go through. So that's another area where we will continue to invest.
Okay. A couple of more questions on the semi, and then we'll move on to other markets you would like to talk about. So in the semi market, you are expecting 200 basis WFE performance in your target model, and so what gives you the confidence about this outgrowing, and what's your competitive advantage?
Yeah. Well, you know, one bit of data is that we've done it for the last 10 years, but, you know, as investors like to say, you know, past performance doesn't predict the future. So how do I feel? Why do I feel that we're gonna continue to do that? I think, though, when you gain share, it's really because there are changes and inflections that become harder to do. When things become harder to do, fewer companies can actually do that for a couple of reasons. One is you've got to invest more, longer, so size does matter, actually. Number two, you have to be wired to do that. Number three, you have to have a broad portfolio, I believe, so you can see those inflections and act on those inflections much faster than the other person.
And of course, even if you're able to react quickly, you have to have a base of foundational product in whatever it is that's going to cause that inflection. So I believe that we have all those pieces at MKS, and that's unique. There's no one else with the portfolio we have just in semiconductor vacuum critical subsystems, let alone lithography, metrology, inspection critical subsystems. So that allows us the opportunity actually to grow a lot faster than the rest of the industry.
Got it. And just want to double-click into one of the competitors. So, regarding Advanced Energy, I think historically you are stronger in dielectric etching, but they have, you know, talking about, like, these new tools they're hoping to gain some share in that area. So could you talk about your competition environment there?
Yeah, well, you know, we, we don't like to talk about specific competitors, but we respect all our competitors, because all of our competitors, like us, have survived thirty to sixty years of a very cyclical industry. We're all strong. We're financially strong. We have innovation engines, no doubt about it. I would say, you know, that in power, you know, the reason we gained number one market share in 2022 timeframe was because of V-NAND. And, the RF power that we provide there is foundational, critical to enabling V-NAND hole etch. And we're on generation nine, so there's been nine generations of V-NAND. With that, there's been nine changes in the equipment to do channel hole etching specifically on RF power, meaning there's always three power sources on a chamber. Every one of them has changed in terms of power level nine times in nine years.
We've never had that happen before. That's the kind of innovation engine that we have that demonstrated. It's kind of scar tissue we have, too, 'cause we learned with our partner there's a lot of things you have to solve, right? So once we get the RF power put in the chamber, you know, lightning happens, sparks happen. You got to fix all that. That's the kind of capability that we have that I believe is unmatched by any other RF power company.
Okay. Okay, and now just move on to the E&P segment, so in advanced packaging, I think you have, like, 25% of exposure in advanced packaging, so which I think is right now the best of EMP segment. So what's your growth expectation there, and, like, who do you compete with in that area?
Yeah, so in advanced packaging, you know, the PCB industry, if you will, there's three segments, if you will, of PCBs. There's the, we'll call it old stuff, multilayer boards. That's the stuff that is easy to do. It's in your refrigerators, some cars, etc., but it's also changing. It's also having more layers, so there's a growth there, too. But we kind of call that as GDP plus growth.
Mm-hmm.
That's a third of the market.
Yeah.
The middle third is what we call HDI, high-density interconnect PCBs. They're all PCBs, by the way. And HDI PCBs are, you know, think of smartphones. They're driving the smartphones, and they're growing kind of in the mid-single digit range, CAGR. And then the final third, the most advanced, is what we call package substrate PCBs. These are the most advanced, most number of layers, bigger boards, smaller lines, smaller holes. This is what's driving and has driven server boards and PC, advanced PC boards, and of course, even more AI server boards. So the difference between a regular server board and an AI server board is more layers, maybe 30% more layers, because an AI board, server board is putting more stuff on it.
And so, you can imagine, the board is the highway that allows chips to talk to each other. And when you make bigger chips, we have more IO talking to other bigger chips with more IO and HBM that's going from eight to 12 to 16. They all got to talk to each other through these highways. So what do you do? Well, I'm gonna make more highways. I'll shrink the lines, I'll shrink the holes, and that is what we're doing as an industry, but that's not enough. So how do I do this? So I make more layers. Okay, I'm just gonna make more layers, so I have more highways within this big stack, and that's happening... but that's not enough. I'm making the board bigger as well. So three things are driving the chemistry and the chemistry equipment.
That's the high single digit to maybe double-digit CAGR part that you're talking about, the IC Substrate. That's, of course, the most exciting thing. It's the hardest thing to do. We love that. It's the hardest thing to do, and it therefore requires suppliers like us, who have an advantage by bringing more solutions to our customers. We can bring a chemistry solution. We can bring the laser solution. We can bring the chemistry equipment solution. No one else can do that. And if we co-optimize this with the customer, they get to their solution faster. And so that's the most exciting part of the PCB industry, the electronics and packaging industry that we see.
Got it. And you just talk about, like, 30% increasing in layers for, like, AI server board compared with non-AI. So what's the content increase for MKS?
When we bought Atotech, just like in Travis Bertrand's earlier, Atotech is a square inches kind of business. We plate each board one layer at a time, and if the board is bigger, more square inches, more chemistry. If the number of layers is more, more layers, more chemistry. It's really square inches of the layers for this. When the number of layers just goes up 30%, that's exactly the amount of chemistry needed, 30%. If the size of the boards goes up 10%, 20%, 30%, if everything else is equal, that's more chemistry. That's exactly how it works.
Then, if the features are smaller, it's not necessarily volume of chemistry, it's the type of chemistry, the precision of the delivery of that chemistry to make smaller holes and smaller lines, the delivery of the laser beam to make smaller holes. And so that's really a little different. That's about technology and winning share there. Yeah.
Got it. And you just mentioned Atotech, and I guess, in your June quarter call, you talk about, like, one of the first wins of the Atotech synergies.
Yeah.
Can you remind us what was the area and, you know, what kind of the future opportunities you are looking at?
Yeah. So in the June quarter, we talked about a synergy win, and just to remind everybody how we count it, if Atotech was already in that customer, and the laser group, ESI, was already in that customer, and they both win, or one wins and one doesn't, we don't count that. That's. They were both there. The synergy win is when one is there and the other is not, and the other is brought in and wins, and that's what we're talking about here. And earlier, right after the close of the acquisition, we talked about a couple of laser wins that were synergy wins. So Atotech brought us in. The laser guys never had, you know, a relationship, and they won. I call those bluebirds-
Mm
... 'cause they're so fast that, you know, just happened to work out right. The laser tool was just perfect. They were buying, et cetera. What we talked about in June, the June quarter, was something we've been working on for two years, and this is probably more common. We're working on the next generation roadmap, item, for the next generation PCBs. You know, so Atotech was brought in, they didn't have a relationship, by the laser group. You know, first time, you just have a PowerPoint presentation, and then you have, "Hey, let me try some samples." And then, "Okay, that's interesting. Let me, let me run a lot of samples," and over and over again.
Then you go from development tool record, or development process of record, to production tool record, production process of record, and that's what we were talking about in the two Atotech wins. They were in some of the more advanced areas of PCB manufacturing.
Okay. All right, before I move on, is there any questions in the audience? Okay, let's just move on.
Yeah.
Yeah, so I just want to quickly touch on the specialty materials segment, so it's a relatively, I guess, stable business, and you think about it, it's GDP plus growth.
Yeah.
Just could you talk about the plus and takes within this segment?
Yeah. So specialty industrial is about a third of our revenue, and very stable, great gross margins, great, you know, flow through profitability. You know, a couple of points there. One is that it leverages the R&D that we're spending in the semiconductor area and the electronics and packaging area, our two core growth markets. That's really important because if you think about it, our R&D spend, if you measure our R&D spend, is not as high as you would think compared to, you know, our peers, right?
Mm.
Because we have a third of our business that's leveraging it. So the denominator is a lot bigger when you divide, okay? So we're not starved for R&D. In the specialty industrial market, it's made up of several different kinds of markets. It's got industrial-industrial, it's got Life & Health Sciences, it's got Research & Defense. The one area... and it can be lumpy, things go up and down, but in general, it kind of all averages out. The one area where we do have a little more exposure with the acquisition of Atotech is automotive. And so we're not making chips for automotive, we're making all, everything else for automotive: brake calipers, you know, pistons, shock absorbers, as well as all the decorative stuff that goes around a car.
So this is about putting metal on either metal or metal on plastic. If you can do that really well with some really tough customers, the automotive industry is a pretty tough industry, right? You don't get designed in unless you really can deliver at cost and reliability, so that's a third of the Atotech business. It's been pretty stable, which we get that question a lot. You know, the chip guys in automotive are kind of going through a digestion period.
A... Yeah.
Yeah. And because the automotive industry couldn't get chips two years ago, so they bought a lot, and now they don't need all that. But they didn't really overbuild. Our customers did not overbuild, you know, door handles and shock absorbers and brake calipers, right? Because, you know, those things, there was not a shortage of it. And so it's been pretty stable. So we're really more unit-driven today, and so if you look at the world's number of cars being built, it's kinda been a ninety billion, or ninety million, sorry. Wish it was a billion. Ninety million units, kinda for the last three or four years, which explains why we're pretty stable in the automotive chemistry business for the last three or four years.
Yep, that makes sense. Yeah, and then, on China, to ask that, so, I think your China exposure has been increasing for the past few years. Now, it's like, I think this year it's more than 20% so far. So what markets are you most exposed to? Like, say, memory and logic, or, like, which segments of your business is most exposed to China?
Yeah, so it's interesting. So China, it may not be as clear for MKS, but you know, when the new rules came out two and a half years ago, we said that about $250 million of our direct sales to China was at risk.
Mm-hmm.
So this is selling to Chinese equipment OEMs, who we've had twenty-year relationships with, building up our market share with them. That is drastically reduced. That's, that's the $250 million we're talking about. So in terms of direct ship to the semiconductor market in China, we are very, very small.
Okay.
We do benefit because a lot of the equipment OEMs, as you can see from their numbers, are shipping to China, sometimes 30% of their revenue, sometimes 50% of their revenue, so we indirectly get that. But we don't count that as China because we don't know if it's really going to China or not. So we enjoy that, and, you know, if that changes, we'll be affected for sure. But really, most of our revenue to direct ship to China is in packaging.
Mm.
Many packaging companies' customers are in China, and that's Atotech, and so that's direct, and so that's really why it's 20% now.
Uh-huh.
It's actually all, almost all Atotech. Now, there is, you know, movement in the customer base, the packaging customer base towards Southeast Asia, toward other countries, if you will, just to de-risk, you know, the China exposure, 'cause their customers are demanding it. And so we're following those customers. As they move to Southeast Asia, we are also moving to Southeast Asia to support them. And so, you know, who knows what the China revenue will be for packaging longer term, but, but I for sure know that Southeast Asia packaging revenue will grow.
Got it. And do you see incremental competition from domestic Chinese makers?
Not more than normal. I think there's always been competition, chemistry-
Mm-hmm
... you know, and some lower-end equipment, which that we don't do in China, but we see that same in Japan, we see it in Korea. So that's kind of not really changed the competitive dynamics for the chemistry business. But I would say Atotech is number one market share-
Mm
... in electronics, chemistry, and equipment.
Okay. And just finally touch on, like, capital allocation. I know you got the new CFO, but he's not here, so just what's the top priority? I know the top priority is still gonna be debt paydown, but, like, is- are you still looking for M&A at the same time?
Yeah. So we announced that, Ram Mayampurath will join us as our CFO, in mid-October. You know, some of the attributes that were very attractive, that Ram brings to the table is, you know, seasoned CFO, public company. He's gone through some, you know, some challenging times, in his previous company, for sure. He knows how to run the organization. He's measured and conservative. That's kind of what you want in your CFO, and very experienced. So those things I think will be, you know, nice adders, to the finance organization. Really no change from our priority. Our priority right now is de-leveraging over the next 18-24 months, as we've said, this is the number one priority.
We've been able to take out, you know, prepay, kind of in the order of $50-$100 million a quarter, prepay, of our debt, even during these trough levels of our semi business and our EMP business. We did a lot of things to help the balance sheet. We did the $1.4 billion convert. We've repriced again, and, you know, maybe the Fed will help us out, you know, next week, as well. But we'll continue looking at any opportunities to lower the interest costs, and de-lever. In terms of M&A, your last question, you know, nothing's really changed. I think there are always going to be opportunities, but they're really gonna be on the...
not on the table until the leverage gets to something closer to, you know, where we wanna be long term.
Got it.
And, you know, I would say, though, that I don't see a transformational M&A in the future. We just did our transformational M&A. We're really well positioned for electronics and packaging already. Sure, there are lots of tuck-ins that could certainly enhance that. And then I think the other point I'd make is, you know, because of the last 10 years of being able to organically take share significantly when some inflection happens, V-NAND being one, World-Class Optics being the second, our criteria M&A is a lot higher now. And by that I mean, you know, we will always look at could we do it organically faster? Now, in the past, when we were smaller, it was hard to add 50 engineers to our power when, you know, the whole company was $700 million.
That would kind of blow the bank in terms of OpEx. Now that we're bigger, now that we've made these bets and demonstrated to ourselves that we can do this. We can make these big bets. I think the bar for M&A becomes a little higher.
Yeah.
And so it has to be something unique, something we don't do, something that really, you know, tucks in right to the portfolio.
And I guess your portfolio is diversified enough right now. You're happy with the exposure right now?
We're very happy with the exposure. As I said, you know, 85% of all WFE has multiple MKS subsystems in it. The other thing I didn't say is 70% of all the steps needed to make advanced packaging is addressed by MKS.
Okay.
85%, 70% of everything that's needed to make advanced electronics. There's no other company that can say that.
Okay. Yeah, and just finally, anything just you want to add, you think investors are not fully appreciating of your company?
You know, one area that, Paritosh and I have talked about is, you know, we try to explain this with the TSMC example of packaging, called CoWoS, chip-on-wafer-on-substrate, right? And we say, "Look, the chip-on-wafer, that's, you know, TSV etching, hybrid bonding, that's great. That's growing really fast from a very low base. And we're gonna enjoy that when a vacuum, a piece of vacuum equipment does, you know, through-silicon via etching, et cetera." But the S is really, I think, where we still have work to do to educate investors. The S is Atotech, that's substrate. I think people are starting to understand that, but that S is where the Atotech chemistry, the laser drilling occurs. That S, for us, in chemistry, is a lot bigger number than the chemistry needed for the CoW.
So when you're doing hybrid bonding and chip-on-chip for HBM, there's chemistry there, 'cause you've got to plate the. You've got to make sure the bumps are, you know, they've got the right coatings, et cetera. There is chemistry there. That's an order of magnitude less than the chemistry needed to fill all those lines and holes for the substrate and the number of square inches of substrates being made. So every time we've got, you know, one stack of hybrid bonding on an AI server, there's 25 layers of S. That's, I think, an area where I think some investors are starting to appreciate that. And I think it's because, you know, maybe a lot of our investors have been semi. They came from semi, they understand semi.
You know, it's not a stretch to understand, you know, through-silicon via etching and hybrid bonding of those chips, but they're not necessarily the investor set that, you know, came from the other side, chemistry, investing in chemistry, investing in PCB and packaging. So I think that's an area where we still have work to do-
Okay.
to try to educate the investor base.
Okay, sounds good. Just before I conclude the session, is there any questions in the audience?
Thank you, Elizabeth.
Okay, great.
Yeah.
Thanks, John. Thanks, Kurt.
Thanks, everybody.
Thanks, everyone.