MKS Inc. (MKSI) Earnings Call Transcript & Summary

Let's get started. Good morning. Welcome back to the session. I'm Vivek Arya. I cover semiconductor, semi-cap equipment at BofA Securities. I'm really delighted to have the team from MKS join us this morning, Dr. John Lee, President and CEO; and David Ryzhik, Head of Investor Relations. I'll go through my questions, but please feel free to raise your hand if you would like to bring up something on your side. So maybe, John, for those who are...

Let David say some things first.

Yes, please exiting part of that.

That's what I'm best for. Any forward-looking statements that we may make today are subject to risk factors in our SEC filings. You can find the GAAP reconciliations to any non-GAAP numbers that we may talk about on our IR page of our website. Ric, I'll pass it back to you.

Thank you so much. So maybe, John, for those who are unfamiliar, maybe walk us through kind of the broad MKS story. M&A has been part of the strategy for you. But just how do you look at the company holistically across the 3 segments of semiconductors, right, electronics and packaging and specialty in [indiscernible]?

Yes, maybe I'll step back a little bit share some of those new to MKS story. So MKS was founded in 1961 with one product. This was a pressure sensor that measured pressure inside a vacuum chamber. And that was a critical chamber. It was the etch and depth chambers for semiconductor. And today, that pressure sensor is still over 80% market share. But with that one product, we were able to build on it to make other organic development of the product lines as well as inorganic acquisitions. And over the next 20 years, we did that and became the broadest critical subsystem supplier for semiconductor capital equipment for vacuum chambers, which is about 60% of WFE. And in 2016, we made a big pivot. We acquired a company called Newport Corporation. That did 2 things. One, it brought us to new markets and new technologies, lasers, optics, photonics and different kinds of markets. But it also made us stronger in semi because what it brought to us was up lithography, metrology inspection customers. So today, when you add up the big 5 capital equipment makers in the semiconductor capital equipment world, it's applied in Lam, Tokyo Electron, ASML and KLA. Those are all our customers, and we have multiple critical subsystems in each one of their tools. So every fab in the world today, 85% of the equipment in every fab in the world today has multiple MKS subsystems in it. So every chip in the world is made with a lot of MKS content. But the other part of Newport is not just diversifying to other markets, but it brought us to one particular market that we saw quite interesting, and that was a packaging [ options ] because our lasers are being sold to people using it to make small features with packaging and okay, what's about -- what is it about this market -- that led us to acquire a laser systems company, Electro Scientific industries or advanced packaging. That also then told us and allowed us to see a trend that was happening multiple years ago. And that trend was the need for packaging to enable more than Moore's law. And because of that, we acquired Atotech. Atotech is a leader in enabling advanced packaging with their chemistry and chemistry equipment. So today, we're foundational to the more than Moore's law, right? Moore's Law is an observation and economic observation, and it was about horizontal shrink. Everything was twice as good every 2 years. And that was absolutely true for about 40 of those 60 years that semi has been around. But then they started [indiscernible]. It's going to go off the curve. And as a result of that, people say let's put maybe 2 CPUs together to make that computer twice as good every 2 years. And we did that, and that extended Moore's Law for a little bit. And then that's how we are running out of steam. And then people say, "Let's put GPUs and I've been with memory together and package it all together so that electronic device is twice as good every 2 years. That's what we kind of call more than Moore's Law. And that requires not just the semiconductors, but the packaging as well. You have to have both in order to maintain this more and Moore's Law. And MKS is the only company in the world that addresses foundational technologies to all of those enabling things that allow more of the Moore's Law. Our semiconductor business is about 40% of our revenue. Our electronics and packaging business is about 25% of our revenue. And then we have specialty investments. This is a group of different markets where we're leveraging the technology we develop for semiconductors and for electronics and packaging. They're using it in other markets and applications. That's 33% of our company. And that has been a very stable [ GDV plus ] segment that allows us to have a much stronger financial foundation from which we can use to continue to invest in the semiconductor and the electronics and patent.

All right. So John, as you look at the second half, you called for some half-on-half growth, what are the kind of the leading indicators of that? And how much of that is market growth? How much of that is kind of unique to your product cycles?

Yes. I would say we've said that the second half of 2024, we believe will be slightly better than the first half in both the semiconductor market and the electronics and packaging market. The semiconductor market is really driven by the fact that a lot of the inventory burn down has occurred. We've seen our customers drop in orders now. So green shoots where we're shipping to their demand and what they're shipping. And that's really a great way to a good sign that the inventory burn down is mostly done with the exception of NAND-based inventory that still has some ways to go in terms of burning down. And then, of course, in electronics and packaging, we do have a consumer products component to electronics and packaging. Think about smartphones and PCs. And in the packaging business, the PCD business, if you will, there is this cyclicality, not all of our revenues from consumer products, but a lot of it is. And Q1 is the lowest point. Q2 increases, Q3 is the peak, and then Q4 depends, depends on how well phones are selling and consumer demand. And so when you add that up, the second half of electronic packing is always slightly better than the first half, given everything else the same. And so that's what we guide. But we are seeing green shoots. And certainly DRAM pricing, NAND pricing, utilization rates, certainly in DRAM factories, even NAND factories. And logic and families seems to be staying also very stable. So we have a little optimism in the second half will be a little better than the first half.

Historically, what's been your correlation to just vapor fab equipment trends? Because this year, for example, we are seeing WFE kind of in the low mid-single-digit type of range, so if that's all you knew, how would it help you forecast your semiconductor business.

Yes. Maybe I'll step back a little bit. So we -- it's hard to forecast quarter-on-quarter because things change so much is inventory and that has to burn or not. So the way we look at how we do -- how we measure ourselves relative to wafer fab equipment is longer term. And when you look at WFE growth rates, it's whatever it is, 5%, 6% over the long term. We've consistently been able to outperform WFE by 200 basis points over the last 10 years. That's just in data. Now we expect to continue doing that and can ask why. Well, there are a lot of growth areas that change. So we believe in only doing critical subsystems, things that are hard to do. But everything is critical, you don't know which one is going to be more critical at that next inflection. The best example is [indiscernible] happened. That was kind of a new thing, very new. And the inflection there was that our power was going to be much more important and much more needed. And because we had or we can double down on that, and that allows us to outgrow the market. So that's an example of a broader idea, which is that having a broad portfolio does that allow you to do that. But it also allows you to bring a more comprehensive solution. You're more important to the customers because you're bringing in multiple solutions all the time. And we're leveraging that beyond WFE. But in general, we tend to outperform WFE on an upturn, underperform in the downturn, but overall, on average, 200 basis points better than WFE over the last 10 years, and that's our model.

Got it. We had Applied Materials I presented earlier today at [indiscernible] search 2 days ago, so they outlined a very strong growth in their gate all around business, right. So I think Applied has said almost $2.5 billion this year. Lam said over $1 billion. How does the move towards this new [indiscernible] structure benefit MKS, right? And how would you kind of contrast it, John, given your experience in the industry versus the move we saw from [indiscernible] to FinFET.

Yes. I think it's similar. When we move to different architectures like FinFET or gate-all-around or even [ backside power ] for that matter, a lot of that is driven by deposition and etch. And of course, Lam and Applied are leaders there. They are our 2 biggest customers. Certainly an upturn, we declare their revenues over 10% each. So we've had a long history in partnership with both of them. We're designed into many of their tools. And so if they're shipping more tools for whatever process there is and gate-all-around, we're likely going to benefit from it. And I think that's kind of the same trend. It's a bunch of different steps that Applied supplies, tools to and Lam, and we're all over those tools.

Okay. On the NAND side, have you seen any signs of recovery? Like would you be early to see that? Or would you start to see that in Phase 2 that you first need to see better utilization of what these companies already have?

Well, I think we will certainly see it before our customers had to ship for sure, right? So we preceded our customers. But in terms of -- we haven't seen that yet. I think in the industry, we're all pretty consistent with that. But we have, as an industry see green shoots, the pricing, of course, has gotten better, positive. That's always good to make some money on it. Utilization rates have improved. And I think you've heard of Samsung saying we're going to convert to [ B9 ]. We're going to go to [ B9 ]. That's their terminology for the next node of NAND, right? So those are all green shoots and good. We haven't seen a lot of folks saying NAND is going to inflect in a particular quarter or whatnot. I think in general, we think DRAM will inflect sooner. A lot of that's because of the uniqueness of HPM, driving a lot of capacity. NAND, you need more NAND capacity as well for AI servers, but HPM-DRAM is a little more inefficient to make, which is good for us to put it makers. But NAND, I think we'll catch up eventually. So a little delayed relatively.

I see. As we look into next year, so not on a forecast, obviously, it's early. How would you kind of stack rank where you see the best growth opportunities?

Yes. I think a lot of folks are hoping that 2025 is a great year. I think Foundry/Logic seems to be strong and continue to stay strong. That's the assumption. And that's driven by going to 3-nanometer volume and working on 2-nanometer. And the other assumption is that China stays roughly the same. And then DRAM, DRAM picking up. I think DRAM picking up would certainly drive WFE in '25 and beyond to be certainly materially better than 2024. And I think that's where a lot of the forecasts are coming from. You have some other forecasts are even more aggressive and that would require kind of NAND. And it depends on when that turns up. But if NAND turns up, DRAM turns out and Logic and Foundry stayed consistent, we could have a very good next couple of years.

When we talk with the number of the end customers who are developing GPU ASIC and we asked, where are the biggest supply constraints, they always point to advanced packaging. So what is your kind of exposure to the need for advanced packaging? And how correlated is that business to just the growth in overall AI and cloud deployments.

Yes, that's great. I'll take Colas as an example of an advanced patching scheme. That's the TSMC approach. But the [indiscernible] everybody has their own version of it. But the idea is that there's 2 parts to the packaging. There is the silicon to silicon, chip to chip, chip to silicon [indiscernible]. After that, you have to put it on a PCB. That PCB, we call package substrate, the most advanced PCB. That is something -- that's where we [indiscernible] just the S, right? So we're the S. We do have chemistry in the chip on wafer, that chemistry is very small relative to the chemistry and the chemistry equipment and lasers that are needed for the S, [indiscernible]. And just as an example, so that S, it used to be for the most advanced servers, maybe 15 layers, maybe 12, and now it's pushing 20, 22 layers for the same server, right? So for us, each layer is made one at a time. And for us, it's square inches. And if you have more layers, your SAM just increased by that much. Second thing that's happening is that we're shrinking lines and spaces and interconnect coal between them because you're trying to get -- you're trying to get more interconnects because you're connecting bigger chips that have to talk to each other or more chips that have to talk to each other. So that straight lines in spaces, that's part of to do. That favors people who can -- who have the technology and who can invest in that technology. And then I talked about layers, and I talked to -- and then the other thing is let's make that board bigger -- because I got more chips. And so all that is about technology inflection as well as area. And that chemistry that needed to fill all of those holes from copper and they call those lines and spaces, Atotech is a market share leader. That's why we bought Atotech. We saw this trend happen. We saw more layers happen. We saw small lines of spaces happening. We saw bigger boards happening. And we saw more chips had to be packaged together to continue to extend more and move on.

How big exposure is this within your E&P segment and did you see it going?

Yes. So our electronics and packaging business is about 1/3 of our company. That was 2022. 2023, a little lower, [ 45% ]. The package, it's mostly packaging. The advanced packaging is about 1/4 of that, but that's the fastest growing. That's the high single-digit part of the advance of electronics and packaging business. That's the battleground right now. And so that's the opportunity for MKS especially when we are already the market leader in supplying the chemistry. There's no other competitor that supplies chemistry, chemistry equipment and laser equipment. No one does offering. We think that's a huge differentiator for us because if we can come with all 3 to solve the customer's problem faster because we have all 3 they get to market faster, right? And therefore, should reward us -- we're giving them the solutions faster.

Got it. Does it make a difference John if somebody is using a chiplet approach versus multiple radical size chips used together. Does that make a difference on the packaging side?

Yes. Stitching radicals together making bigger chips. I don't think so because if you made a bigger chip by stitching, that chip still got to communicate. And so that communication bandwidth, if you will, is really -- that's what drives how many layers of interconnects you have, how small the lines of spaces are, how big the entire Board is. Stitching is difficult, right? And that's really why people have gone to packaging chips by themselves of the same size radical, if you will, [indiscernible]. So I don't think it would really matter for us. What really matters is how many interconnects do you need to have on that package substrate.

One other thing that I noticed is this opportunity within E&P for laser drilling or low earth orbit applications. That seems to be like a very interesting and unique aspect of -- not something we hear normally I think in a semiconductor.

Yes, that was kind of a niche application. It's not a big number for us, but it shows you that -- it showed us that our laser tool is quite differentiated because in order to make those types of PCBs for low earth orbit applications, our tool is much faster than the competition. And if you think about low earth orbit, so there's tens of thousands of satellites rolling around the world. And they're not in geosynchronous orbit, meaning the same satellite is not looking the same antenna. They're actually -- they have to chase each other, right? And because of that, you need a lot of antennas on the ground level. Those antennas are basically -- if you see some of them, they are kind of -- they look at one big PCB, kind of 12 inches by 15 inches. That entire PCB is the antenna. And that's what we're making and so it's kind of a proof point that our laser drilling tool is quite even.

And then maybe help us unpack the specialty industrial business? What is in that segment, what are kind of the growth drivers?

Yes. So specialty industrial, as I said before, leverage is the R&D we've already invested into the semi market and the electronics and packaging market. So for instance, we've had different types of products being used in industrial and dining manufacturing. So there's an industrial segment to that specialty industrial. Industrial, industrial. We've had lasers being used for eye surgery. So there's a life and health sciences component to it. We've had vibration [indiscernible] used in microscopes. So there's an R&D part of it. There's thermal lenses that we make, that's also part of the [ D ], the defense industry. So it's life and health sciences, research and defense and industrials. And each one of these are niche applications for us, but it's great gross margins, great cash flow, and we're not doing much R&D like at all. We're leveraging it and that's 1/3 of our business now. And that 1/3 is really giving us a lot more stability as we go through these cycles in the semiconductor and the electronics and packaging markets.

All right. On Atotech, you hit your cost synergy targets already, how much more is left? Are there more operating synergies that are left?

Yes. I think it's part of the DNA of MKS. We're always looking at how to be more efficient. I would say this, we hit the $55 million of synergy in 20 months. So on the earlier end of our 18- to 36-month target. Probably not a surprise to those who know us well. I would say this, even today, we're still finding ways to be more efficient with some of the legacy, legacy MKS business, the vacuum side. You saw the gross margins starting to continue to uptick over the years. We've been scrubbing that group for 60 years. Newport, you saw some of that margins improving a little bit too. We bought Newport for 8 years. We've been scrubbing that for 8 years, maybe won't stop. Atotech, we've scrubbing for 2 years. we'll start talking about it as a synergy target from the M&A, but rest assure, we will continue looking at opportunities.

And then, John, maybe moving on to the financial model. I think you've given a '27 revenue target of $5.6 billion, so that's sort of a mid-teens or so sales CAGR from '23 levels, maybe give us a view of how much progress you're making there? Is that a conservative target? Is that an easy target?

Yes. So when we developed a 2027 target and we showed that at the Analyst Day in 2022, the 5-year model. We looked at it, we built in cycles as well. But we kind of looked at, well, how fast is chip revenue going, what do we think is the range of CapEx intensity. So therefore, what is WFE and then 200 basis points above that. So that's part of the model. 40% of our company is WFE plus 200 basis points. And what do you want WFE to be in 2030, let's say, if the semi-conductor market is $1 trillion, pick your favorite CapEx intense, 15%, it's $150 billion WFE, back that up to 2027. So that's one part of our market. The other part is electronics and packaging. And there, you have 3 segments to it, you have the GDP part, which is kind of multilayer boards, kind of old stuff, older stuff. It's a great business for Atotech. The middle stuff, which is high-density interconnects think smartphones, so big market, too, growing at mid-single digit. And then you have the fastest growing top third, this is the package substrate, we just talked about for AI, and that's growing at high single digits. And if we put it all together, we think that we should be able to grow E&P at 300 basis points above GDP, right? Obviously, much faster in the higher end and less in the lower end. And especially the industrial [indiscernible] company is really GDP plus. And that's how we get to that $5.6 billion. And we had some other numbers there, right? 47% plus for gross margin.

Which all of these exceed.

We exceeded in Q1, but we guided a little lower for Q2, just to be -- but we're happy with that progress for sure.

Is there more leverage to the gross margins in that.

We're not ready to say that there's a different model than what we've said publicly and our model is that a 50% drop through for gross margin for incremental dollars of revenue. So -- and 40% incremental margins. So we're really happy with where we've gotten there. And then EPS is in that $13 range. And certainly, we're happy with the progress so far. And we're not guiding a 5-year model every quarter, obviously. But remember, we got to 47% gross margin at $3.6 billion versus $5.6 billion, right? And so I think there's just hats off to our team for really working hard on optimizing the company and operating it and managing it really efficiently in a downturn.

Is it a services component to your business also that is perhaps a little more predictable under [indiscernible]?

Yes, there is. About 50% of our all total MKS's service. And this is a service of hardware equipment. And so all the fabs that have all of our critical subsystems, things can break, things can wear out. Some things are kind of consumable in a longer time frame, so that's 15% of the revenue. That's fairly stable because it's utilization dependent, fab utilization dependent. It doesn't mean it can't go down, but certainly fab utilization dependent, spare parts, things like that. And so that's a very stable part of our revenue. But the other part of our revenue that's very stable is the chemistry. Chemistry is consumable, right? And that's utilization dependent too. So when you add chemistry and you add services, hardware services, 40% of MKS revenue is that bucket of relatively more stable revenue. And so that's also another foundation, if you will, in terms of financial foundation, in times of cycles.

On the capital structure side, you refinanced your term loan, you have been repaying right debt. How should we think about delevering the balance sheet?

Well, over the next 18, 24 months, that is priority #1. M&A is really still going to be important to us, probably in the longer-term view right now because we want to get that leverage down. Our long-term model for leverage goal is 2.0 gross leverage. And you're right. We had unfortunate timing when we closed Atotech, interest rates had gone up 12 months in a row, both markets turn south. So kind of poor timing. But we've got the operating margins back to 20% even in trough quarters. We were able to pay off $200 million prepay in the last 6 months. We were able to get our tax rates down from 28% to 20%. At the term loan A is off the books because there were some covenants there. We weren't really worried about it, but you always get that question. And we did a little repricing as well, and we will continue to be pricing when that makes sense. And of course, 2 weeks ago, we did a convertible. That convertible probably went as well as it could have. Our banking partners did a great job. We had $1 billion with a 15% green shoot that went to $1.2 billion with a $200 million green shoot way oversubscribed. Stock price reaction, the first day, of course, there's a big dip, but the next day, when the hedge funds get allocated because they're the ones that have to hedge, they probably had less. And so the stock recovered and it's pretty much recovered now. What that convert did is it took out $75 million of interest cost per year, right? We are paying incremental interest costs on the term on Visa of 7.5%. We converted 1/4. We also tried to make the converge, much more bottom line, if you will, by purchasing cap call. So the shareholders are not diluted until 100% of the share price when we launched.

Long term, John, if you look at, let's say, the next 3 to 5 years, is there an ideal mix that you would like to see within the company? And are there M&A opportunities you think that will help you drive towards that idea mix and how does one come up with the idea mix?

Yes. Yes. No, it's a good question. I think really, we have all the major components that we want for our goal of being foundational to more than Moore's Law. No one else has this, right? A lot of people talk about packaging now. We actually did something about it, 3 years ago, right? So with the acquisition of Atotech. And so we have the major pieces. Yes, you're right. There's always tuck-ins, right? There's probably more opportunities in tuck-ins in the photonics space and maybe the chemistry space than the vacuum space because we were the consolidator there, if you will. We don't really have a bias as to more consumable, more chemistry, more lasers, more optics. I think it's whatever makes sense, whatever is actionable, we don't have a lot of gaps right now, but always are, right. But we're really happy with the portfolio now. And if some things come up, we're always debating whether we can do it cheaper organically or inorganically. In the past, we've always done a lot of inorganic. But more recently, in the last 10 years, I'd say we've actually had many more singles, doubles of home runs organically. And I think scale helps. So when you have scale, you can actually make big bets organically. Our power of [indiscernible] was one big bet organically because you don't get paid for 2 or 3 years because you've got to hire the engineers, you got to develop it. You got to partner with your OEM customers, they got to win, and the chips got to go build. It's a 3- to 5-year time frame before you win, we get paid. We're willing to do those kinds of bets, we're courageous enough and we're capable enough. And that really is a differentiator for us longer term. So in terms of product mix, we like all our children, if you will, but they're all critical subsystems.

Must have some favorite.

Well, the [indiscernible], as I said, it is still a market share leader. Competitors have been trying to take that share for 65 years. And we have some secret sauce there, if you will, that has allowed us to maintain that very high market share. Our power, more recent story. We were #2 in market share for 30 years, trying to now in 2022, we took #1 market share. Our investments in world class [indiscernible] over the last 4 years has shown -- hasn't resulted already. You can see it in the numbers, our lithography, metrology and inspection numbers have been consistently growing even during a semi downturn because those customers have long lead times and their suppliers have long lead times. Plus we're taking share because we invested more capability. And then, of course, the Atotech acquisition, we just believe that packaging and package substrates are going to be the next battle for shrinking and complexity and requiring companies who can provide R&D scale.

On that positive note, thank you, John.

Thank you. Thanks, a [indiscernible]. Thanks for having us.