Good afternoon. Welcome to the 27th Annual Needham Growth conference. Thanks for joining us today. We're going to sit down with a discussion. I'll stand up, and John, and we'll sit down with the CEO of MKS Instruments, John Lee, and the CFO, Ram Mayampurath. My name is Jim Ricchiuti. I am the Senior Analyst in the Equity Research Department at Needham, covering companies in the advanced industrial technology space. MKS, I think, as many of you know, has been a leading subsystem supplier and laser and photonics supplier to the semiconductor industry for years. But in 2022, significantly diversified and broadened its reach into the electronics and packaging market through the acquisition of Atotech, which has certainly proven to be an interesting acquisition that positions the company for the growth in that market.
You know, before we get into some of the questions today, how about a quick snapshot of the business following that acquisition, which is a couple of years old now?
Yeah. No, thanks, Jim. Thanks for having us here. Yeah. So we closed Atotech in August of 2022, so a little over two years now. And remember, the thesis of it was packaging was going to matter. Packaging was going to enable the next generation of more than Moore's Law. Semiconductors were necessary, hard to do, but no longer sufficient. And you're going to have to put these things together. And of course, we were a little early. I don't think people really appreciated how important packaging was going to be. But I think with the advent of AI, which couldn't have happened without heterogeneous packaging, I think everybody understands that story now. So how's Atotech done under our stewardship? Well, I would say a couple of things. Certainly, the larger PCB market has been a little muted. So that was a little headwind.
So PCs and smartphones, that's been a little muted. And the burden of being market share leader is you have everything, right? At the same time, AI, PCB, and packaging for AI has certainly been a great growth story. Those customers of ours are levered to that. They're doing really well. I think if you look at that thesis, that is really something that has played out. I think when you look at synergies with Atotech, there's cost synergies. We kind of did that after 18 months. That was not that hard to do. And we continue putting the MKS approach to operational efficiency onto every company we buy, including Atotech. The product synergies, revenue synergies between our laser group, which we already had, and the chemistry group, I think that's really also really played out as we expected.
A lot of customers see the power of having all those steps together in one house. Atotech already made that strategic decision by having chemistry and chemistry equipment. They're the only large player that has both already. What we did was add the laser step in between. Of course, we made some investments, and Atotech already had some investments in tech centers around the world where you have chemistry plating equipment in-country with those customers. You have chemistry capability, and now you have laser tools there. We can co-develop with all these customers with 70% of the steps that they need. That's really powerful for them, especially if you think about advanced packaging and the roadmap for advanced packaging.
As you put more chips on top of an NVIDIA server, you've got to connect it to the outside world with more complex interconnects. Those complex interconnects require smaller features, harder to do. They require more layers, harder to do, and they require bigger boards, all harder to do. And our thesis is you need more knobs to optimize that difficult roadmap. You need chemistry knobs. You need equipment for chemistry knobs. You need laser knobs. And the only company that has all three is MKS. And so I think we're uniquely positioned to solve those really tough problems for our customers. And that's resonating with lots of customers.
Good overview of that. Before we dive a little bit more deeply into it, obviously a lot of questions coming up at this conference and leading up to the conference just about tariffs and about China export restrictions. If you could maybe bring us up to date first, the direct, indirect impact on MKS, whether it's minimal, I think you've talked about it in the past, and whether there's any uncertainty as it relates to these restrictions, the ones that are being proposed that would affect your large OEM customers?
Yeah. I'll let Ram take that one. Yeah.
Yeah. So hello, everybody. My name is Ram Mayampurath. As Jim said, I'm the CFO, been with the company for about three months, learning fast. The most significant impact, I would say, happened in fall of 2022 for MKS when we were restricted from shipping to certain direct Chinese OEMs. That was about a $200 million-$250 million of top-line impact, a very profitable business, which is out of our numbers now. So 2023, we took that impact. We still have some direct shipment to Chinese OEMs, but much smaller numbers, very small, very small. As Jim said, the primary exposure is to indirect exposure. We sell to all the large OEMs that are outside China, and they sometimes sell to end users in China. We don't have a way of quantifying that.
All what we go through is look at their demands or their orders to us and their guidance. We haven't seen any meaningful decline in their guidance. We ship to all the OEMs that you know, you're familiar with, so I would say that there is no real known news compared to what we have said before with regard to China so far. It's an area we are watching very carefully.
Yeah. And I would add too that we've talked about 85% of all WFE have multiple MKS subsystems in them. So if you go in any fab in the world, 85% of that sea of equipment will have multiple MKS subsystems in it. So we care about where it is made, where the tool ships, and we don't, right? We really care about WFE. It could be made in Arizona. It could be going to Arizona. It could be one customer or another customer. It could be a foundry customer or a DRAM customer. When you're 85% of all WFE, you kind of don't really care. You just care about WFE, whether that number goes up or not. We're a little more indexed in NAND. That's true because of our RF power. But in general, we have multiple subsystems on every piece of equipment, every fab in the world.
I'm not sure you'll answer this, but any reaction to some of the public comments from some of your bigger customers in early December, notably AMAT and Lam since the publication of these rules? I think one of them even suggesting back in December that WFE, they expect growth in WFE in 2025. I'm just curious what your reaction might be to some of the comments.
Yeah. I think you should ask them how they got their numbers, but I would say we look at all the same things everybody's looking at. We see potentially China being less. We see DRAM driven by HBM potentially being more. We see TSMC, of course, their business is doing well. Our visibility is lower now fundamentally because our lead times are back to normal. Lead times have traditionally been 6- 12 weeks for our stuff. It's been a relatively steady bouncing on the bottom, steady. Our customers know that, so they know that if they order something and it's six-week lead time, they'll get it. People aren't over-ordering, aren't ordering 12 months out. I think in general, our visibility is a lot lower. It's not lower than it always has been, right? It's actually just going back to what's normal.
I'd be the last person to try to predict WFE because if I did, I wouldn't have to do this job, right? I could be an investor like you guys and make more money. I think, though, that we are always looking at design wins and how do we position ourselves relative to competition. We always have one foot on the brake or one foot on the gas pedal. That's how we run. We've had to run like that because, as some of you know, in semi, it can go up 30% in two quarters and can go down 30% in two quarters. And you've got to be able to be flexible and agile in order to operate financially responsibly during those kinds of periods of time. So this current downturn has been maybe six quarters, maybe eight of kind of bouncing on the bottom.
The great thing is it's stable. It's never been this stable. The bad thing is it's stable on the bottom, right? But the longer it is, the sooner it'll come back, the faster it'll come back. So we do subscribe to the longer view of WFE, which is that in 2030, it ought to be a trillion-dollar semi market. And in 2030, CapEx intensity ought to be in that 14%-15%. And you can do the math as easily as I can. And that's a pretty good long-term market to be in, especially if you're addressing 85% of it.
It might be worth reminding us, too, of the areas because this part of the business has changed a little bit. It's going back a ways from the Newport acquisition. But it might be worth talking about where MKS today is over-indexed in certain parts of the business and generally how that might play out as we see some of the things that we've been talking about.
Yeah. Maybe a few questions about semi. I'll address that.
Yeah. That's right.
Semi, right? So semi, we were traditionally a Dep/Etch company, right? Applied, Lam, Tokyo Electron. And then with the acquisition of Newport, that brought to us metrology, lithography, inspection. And however, the share that Newport had of those kinds of WFE equipment was much lower by design, right? We are changing that. We made investments in what we call World Class Optics, better capabilities of CapEx and process engineering so that we can build more complex optics, more complex optical subsystems, integrating the lasers, for instance, as well. And with that decision four years ago, we started seeing those design wins take place, and then the numbers started to show. And so four years ago, our revenue in semi for World Class Optics was about $150 million, give or take, cycles. Now it's $300 million, mostly in the last two years because design wins take a while.
I think we're still in the early innings of gaining that share. So that's one of the areas where, as Ram said, okay, we couldn't ship to Chinese OEMs as much as we did. Okay, that's a headwind. We're not going to whine about it. Let's go do something about it. Let's go see if we can take some share in another subsegment. So our share in World Class Optics is higher, a faster growth rate than the litho metrology inspection WFE has been in the last four years as well. The other area we're a little over-indexed, if you want to call it that, is NAND, RF power for NAND. It brought us a number one market share. So we're not complaining. It's just that NAND is now a little more muted. So when that comes back, that RF power will come back.
But in general, we're levered to everything that goes into foundry, DRAM, NAND, legacy, leading edge, WFE. But we're working on those areas, particular areas where we have opportunity to grow, like World Class Optics.
You mentioned AI. And is there a way for us to think about how much of a driver it is for the packaging side of the business, how much it is for the semi side of the business? Obviously, we're going to be talking about this for a while. But what's the best way to think about that?
Yeah. Maybe I'll start with the semi side, right? So AI is great because chips are more big, bigger, harder things to do in chips. In semi, though, we look at AI maybe a little differently than most. We look at it as the next big thing in a long series of big things. A lot of people think of AI as this miraculous thing just popped out of thin air. And we look at it as, no, the fundamental driver for why AI chips happen now is Moore's Law. Moore's Law is the fundamental driver, right? And if you think about going back to my college days, a long time ago, I was happy to be in college when the PC happened, right? PC, right? Personal computer.
If we wrote the history of mankind in this century and the next century, you might say PCs are a bigger impact than AI. I'm not saying it is. I'm just saying every human being can have a PC, a computer. That's a pretty big change, right? That was enabled by Moore's Law in 1982, and then the PC could be mobile. You can carry it, and then you can carry it in your pocket, and then you can carry it in your pocket and talk to everything on the web, right? Those are all big inflections that are maybe just as big as AI. They were all enabled by something more fundamental, Moore's Law. Okay, so everything gets twice as good every two years. New things can happen, right? New applications. The extension of Moore's Law is more than Moore, some people call it.
And that's the packaging of these chips because at some point, the horizontal shrink wasn't enough. Moore's Law started asymptoting. So then we said, let's put multiple cores on one die. That extended Moore's Law for five or 10 more years until it started asymptoting. And we said, let's put chips together, heterogeneous packaging. That's where we are today. And because we did that, more than Moore's Law kept on this asymptotic curve. And now AI became affordable. The concept of AI and large language models is not new. It's just now affordable that we can all maybe access it. And so what's more important, semi or packaging? Both, my answer. Both, because I think semi is going to continue to be needed to improve, but packaging is also going to be needed. You need both. So I want to have a company that's levered to both.
I want a company that's levered to Moore's Law because that's the fundamental thing that's driving AI today, something tomorrow, and PCs in 1982, right? So that's why I can't tell you which is more important. I hope they're both important. I love all my children, right? But I would say this, 40% of our revenue now is semi. 25% is electronics and packaging. 25% of that is the advanced packaging that's driven by AI, right? But that's growing at high single-digit CAGR. Semi is growing at mid-single-digit CAGR. Big number, right? But I think those are the dynamics we look at. And I think we're the only company positioned to take advantage of whichever way it goes, right? And enabling whichever way it goes as well.
Let's just turn back for a second to the Atotech business. It might be helpful for some folks. You mentioned 25% that represents electronics and packaging, roughly a third of that business coming from specialty industrial, right?
Yeah, roughly.
I guess what I'm asking is if we look at that business, because it's somewhat unique for MKS versus historically, its materials and equipment, give us a rough idea of that mix because it's also been important during this cycle, hasn't it?
Yeah. So Atotech is a $1.2 billion business today. Two-thirds of it is focused towards electronics packaging. One-third is what we call general metal finishing, so automotive and industrial kind of hard-to-do coatings. 90% of it is consumables. So utilization dependent, 90%. 10% is equipment. And that's helpful, right, during downturns. So it's consumables that have a shelf life that's pretty short. So you can't build up inventory. It doesn't live on the shelf very long. So our customers are continuously ordering this. So that's quite helpful from a financial standpoint. So let's talk about gross margin, right? So you might say there might be a difference between the gross margin of the electronics, which is GMF. GMF sounds like maybe that's not so differentiated. It's the same. 55% is gross margin for GMF as well as electronics. It's 10% higher than our nearest competitor.
In an industry where gross margin is 20%-30%, all our customers. So think about that. In an industry where some of our customers are growing fast, and they'll probably get better and better, but it's an industry of 20%-30% gross margin, paying a supplier 55% gross margin, we must be delivering something that is unique relative to our competition and also necessary, right? And I think if you talk to many of our customers, these are customers we've had relationships with decades, certainly with the Atotech team. We sell them a tool that's 100 meters long, and they might buy three of them. That is the factory. And we might have chemistry that addresses 60%-70% of those steps. That is yield.
So when there's a decision of whether to go with Atotech or MKS or not, it's a CEO decision for that company. It's also something where when we come in to solve a problem or talk about the next roadmap, CEOs and COOs show up because we're that important to them. We're partners to them, right? We enable them. And so that's really the beauty of the Atotech business, I think. Yes, chemistry, liquid phase, a little different than what we're used to. Consumables, we didn't buy it because it was a consumable company. We bought it because it was an advanced packaging enabling leader company, right? Consumables is a bonus, right? But that wasn't the strategy.
And that business sells to what, the top 30 PCB?
Yeah. So the PCB industry has top 30, but there's hundreds of customers, lots of smaller ones. Just as an analogy, remember in 1990, the number of chip companies there were? 30. I was there, right? Now they're not, right? Fundamentally because the industry has become harder to do, right? The things that the industry is trying to solve, those problems are harder and harder to do. So you've got to have size and scale to be able to invest to solve problems over multiple cycles. So you've got to manage your company quarterly and over the long term. And fewer and fewer companies can do that. And so that's why there's been this consolidation of chip companies. Probably will happen in the packaging world as well as those problems become harder. So it's not going to happen just because people want it to happen.
It's happening because the industry requires it. The problems you have to solve require more CapEx and OpEx over sustained periods of time, and only the best managed companies will be able to do that.
Talk to us about looking at that business, some of the inroads that you've made with these customers with your laser technology, and the drilling business has been a tough business for the last several years. What are you seeing in that market that gives you the confidence that the synergies are going to play out?
Yeah, just as a backup. So we bought Electro Scientific Industries, a laser systems company. They're a market leader in flexible PCB drilling. So that's still the case. But we're trying to grow it in the rigid laser drilling business. And that's what we're talking about for AI and phones and PCs. And we made, I think, good progress. There's, I don't know, eight or so tools in production today of our laser tools, making money for our customer. That's still relatively small market share. But what I see is that those problems are getting harder. Laser tools have to drill smaller holes. They have to drill them faster. They have to drill them more precisely. That's exactly the kind of problem we love. Those are the kind of things where if we can solve it, we differentiate.
The synergy with the Atotech team, as I mentioned before, when Atotech comes, it's COOs and CEOs who show up. We weren't getting that attention when we came with a laser tool because it's one tool, right? Now we get to see everybody because we're coming as one company, right? And the laser tool is getting a lot more of that attention, a lot more demos. And so we've talked about high single-digit synergy wins between Atotech and ESI. When we talk about synergy wins, it's neither of them would have won except that the other one brought them in, right? So if they're already there and they win, we're not counting that. And those are things that will take time to play out in terms of revenue. But as we said many times, we're very comfortable with design and wins taking a while.
We've lived our whole life in semi. We know it's the same in packaging as well.
Is one way to think about both parts of the business, John? The Atotech business is holding up pretty well without the mobile phone driver. PC demand, as you noted, has been soft. WFE, we've been kind of bumping along for a while. But the way you're describing it, it sounds like what you're describing is that there should be some very nice leverage as both these things turn.
Yeah. I think 2024 is a nice, complete, clean year where all of you now, you can see our P&L. Every quarter was about the same. Every quarter was bouncing along the bottom. No ransomware, recovery from ransomware, anything like that. It's a full year of muted markets in our semi market and in our electronics and packaging market, so $3.5 billion. 2022, we were $4.5 billion. So that's just market, and what we've tried to do over the last year and a half was get our cost structure in line, get our capital structure in line, work on pricing, commercial pricing, as well as operational efficiency, so we're really happy with where our gross margins are, right? Our gross margins are at the level we talked about in 2022 as our long-term gross margin, 47% plus, at $5.5 billion. We were at 47% at $3.5 billion today.
So we're really looking forward to what we look like at $4.5 billion with this kind of cost structure and gross margin that's different than it ever has been. We're probably more structurally profitable now at this level than we've ever been, right? So I think that's the leverage we have in the model. There's a question over there.
Please.
Sure.
[audio distortion]
Yeah, so advanced packaging is a third of the PCB business, if you will, and that is growing at high single digit.
Why would you say? I'm just trying to understand like where you play in that. Why is it not growing as fast? I'm seeing HBM and everything going out there. It would be growing faster. You tell me. [audio distortion]
Yeah. We are playing in a slightly different part of it. When you think about TSMC's CoWoS, C-O-W-O-S, we're the S, the substrate part. But the S is growing very fast for AI, much higher than high single digits. It's just that AI, as a percentage of the advanced packaging business, is 5%-10%. So the other 90% is PCs and non-AI servers. And we have market share there, and the PCs have been pretty muted. So that's why we say high single digits. A lot of it is driven by this new stuff, right? The AI stuff. But PCs are not growing, obviously, at that rate. The packaging for PCs. Yeah.
You alluded to the gross margin and certainly solid improvement Q3. Was that across all three areas of the business?
Yeah, I can take that. I think there's one more question coming up, but let me address your question now. It is. It is, and John touched upon it. It's a combination of primarily commercial actions and cost control, operational improvements, and we touched upon the chemistry business and packaging and the mix there. The mix is also a big component of it. The addition of Atotech certainly helped the overall mix of MKS margins. Now, quarterly, we might see swings, but bringing in a 55% margin business into the portfolio helped with the overall gross margin percentage. We still have things we can do on the operation side. As we invest in CapEx and look at business continuity and expansion, there are things we can do on the footprint. There are things we can do on procurement. But the biggest step change is going to come from volume.
As the top line comes back to more normal levels, that's when we'll see the next step change increase in gross margins because the cost has been corrected.
Yeah. Just to follow up on the CoWoS. So you said you participate in S. Why not the other area?
Yeah. So we do participate in CoW, the chip-on-wafer. But remember, in that, it's about etching through-silicon via and big trenches and filling with that. So we will benefit from the fact that etch companies, customers of ours, are benefiting from it. So that's the same as more etch tools going out. The chemistry there, though, it's a lot we're not as strong in chemistry for silicon-based kind of processes. So think dual-damascene. Other competitors do that. We're not so strong in that. We're more chemistry for organic substrates or the packaging part. Plus, remember, CoWoS is the CoW is one layer. One. The S is 15-20 layers of package substrate, the most advanced, dense. It could be 5-15 more layers of HDI below that. It could be 40 more layers of MLB below that.
Every one of those things has copper lines and spaces and holes. That's the HDI we're talking about. It's huge compared to that one layer in CoWoS.
[audio distortion]
Yeah. There's three things that are driving that roadmap. It's the same as semi, right? What do we do when we have more transistors than one layer and we got to have them talk to each other? We have more metal layers. We add smaller lines of spaces in those metal layers, and we have bigger chips. All three are happening exactly the same way in package substrates. So there's more layers. There are smaller lines of spaces. So that's harder to do. And then the board is getting bigger, right? And that makes it a lot harder to keep things flat, right? You're putting 40 layers with a bigger board, small lines of spaces. They can't really tolerate a little bit of flex as much as they could when there were bigger lines of spaces. All three are driving a lot of these integration problems for it, right?
And that's what we love. That's what the industry is going to. It's going to be a few people who can help customers drive that roadmap. And the best suppliers that can do that are the ones that are going to have the knobs for equipment, the knobs for chemistry.
Maybe on that S, talk about the competitive position or market share. I think you may have touched on it, but I think it's worth.
Yeah. So for the S, chemistry market share on S, the electroplating on electronics. We're number one market share, high 20s% market share. I think our next competitor is probably low 20s%, maybe even below that. We only have two other global competitors, a bunch of small regional guys who are very good at what they do, but they're not global. So the industry already has kind of three big guys. So it's not an industry where a lot of consolidation has to happen. We're already efficient, I like to think. And we're already market share leader for S. Yeah.
Turn to the balance sheet. The company's made significant progress paying down debt. Maybe why don't you remind us what the MKS thinking is in terms of debt pay down relative to what you said back at analyst day?
I'll take that. Yeah.
I can take that and first of all, MKS has done this before and in organic growth, used to be a core part of our growth strategy and we have borrowed money for acquisitions. All our CapEx, internal investments, dividend, all that comes from the cash we generate, so we have a history of borrowing money for acquiring top targets and then paying that down over a period of time, so that's where we are now, so it goes back to the discussion on gross margin too, watching that profitability increase and then tracking that all the way to cash generation and then focusing on debt repayment. That's going to be the mantra for the next several quarters. In 2024, our mandatory debt payment was $50 million. We paid $426 million on top of that. It's voluntary, close to $500 million.
We will continue to invest in CapEx, about 4%, 4%-5%, I would think, top line. We still have a little dividend, but the primary focus on capital allocation, for capital allocation is on debt. It is to get our debt down to acceptable levels and strengthen the balance sheet. So that's going to be the focus for the next several quarters. So we made a good start this year. It's a year where the top line did not really do very well. It came straight out of operational excellence and efficiency and margin improvements. As we see that top line comes back, we should be able to do more towards debt repayment. So our goal is to get leverage down to acceptable levels, which we have communicated before, and strengthen our balance sheet.
I would add to that. In coming into 2024, our annual interest payments were $330 million. Exiting 2024, it's $220 million. And so taking $110 million out of annual interest payments was a lot of work, obviously, in 2024. And so that's the kind of thing we're always going to do, right, to try to reduce those costs.
You shared some other metrics back a couple of years ago, December 2022, just in terms of targets. John, has anything changed with respect to how you see the business today?
Yeah. In 2022, we gave a five-year model, 2027, that was $5.6 billion, 47% plus gross margin, EPS $13 and change, right? And we had that analyst day because we had just acquired Atotech. So here's our strategy. Here's where we think we can grow. Nothing's changed. I think, in fact, if anything else, it's kind of solidified the strategy of packaging being equally important to semiconductors. Also, that semiconductors becomes more and more difficult, and it's a great place to be, right? A $1 trillion semiconductor market. I don't think we could have predicted trade wars and things like that. But those are things we roll with. I think the strategy remains very much intact.
We've exceeded the margins.
Yeah.
Any other questions? Question here?
I do. Just give us a sense of how did our end customers finish their stocking? [audio distortion]
Yeah.
[audio distortion]
There was a really big cycle in 2022, for sure. I would say we talk to our customers all the time. The only area of inventory that's still left to burn of our inventory is still RF Power for V-NAND. That's it. Everything else is pretty much in balance. We're shipping to what they're shipping. We see that. Now, on the upturn, when there's a fast upturn, we usually ship more because they're planning for the future, right, but since it's been pretty flat, our customers are like, "Well, we don't need to. Why would we plan on an uptick yet until we see it?" but once they see that, as it's happened many times, it's just like a rocket ship, and we're all off to the races, but right now, it's been pretty steady. As I said, six quarters are steady. Unfortunately, it's the bottom steady, right?
[audio distortion]
Yeah.
I think coming into 2024, I thought we were done, but there was probably a little more left. But today, it is pretty much done. Yep.
Any other questions? Okay.
Great. Well, thank you again, Jim.
Thank you.
Yeah. Appreciate it, Ram.