Good morning. Welcome to the 28th Annual Needham Growth Conference. The next presentation will be a fireside with MKS. We're pleased to have with us this morning CEO John Lee, Michelle McCarthy, Chief Accounting Officer. Also in the audience is Paretosh Misra, VP Investor Relations. My name is Jim Ricchiuti, Senior Analyst in the Equity Research Department at Needham, covering companies in the industrial technologies space. I think most of our audience is familiar with MKS. John, Michelle, thanks for joining us at the conference today, so I think MKS over the years has compiled a strong track record in M&A, resulting in the company that we see today, and with that, maybe I'll turn it over to John just to give us a brief overview of the company today, and then we'll dive into some of the discussion.
Yeah, great. Thanks, Jim, and thanks for having us. It's always a great conference to kick off the year. You know, I think Jim is right. We've done a lot of M&A. We've done a lot of organic growth, but our strategy today is made up of being foundational to all the technologies needed to enable advanced electronics. And so when we think about advanced electronics, AI is the latest thing that everybody talks about. But before that, it was PCs. Before that, it was smartphones, et cetera, et cetera. And this was always made up by just semi. Semi was getting better, so you had better advanced electronics. And that was Moore's Law. But in the last decade or so, that model no longer held true. Semi was getting better, but advanced electronics by themselves weren't getting twice as good every two years.
You needed to put these chips together through packaging in order to keep on that road of Moore's Law. And in order to do that, we said in 2021, we needed to decide whether we're going to be foundational just to semi and part of advanced electronics, or to all of advanced electronics. And we decided, obviously, to be foundational, continue to be foundational to all of advanced electronics. And that's why we did the Atotech acquisition. A little early, for sure. Many investors didn't understand it. I think they understand it now. But we address 85% of all wafer fab equipment in every fab in the world. 85% of every piece of equipment in every fab in the world has multiple MKS subsystems in it. No one has that kind of market share.
70% of all the steps needed to make advanced packaging, for instance, 20-layer boards, 40-layer boards for AI, we address 70% of all those steps. So no one's even close to all that. And so this is the reason why we did the acquisitions we did, and the reason why now you're starting to see some of those numbers show up, especially as AI starts filtering into advanced packaging, as well as, of course, semi.
Yeah. And for what it's worth, from where I sit, I mean, it's been interesting to see the questions that I get, and I assume some of my peers get about the company. There's really been a shift toward more of what's happening on the E&P side of the business. And I think for a while, maybe you're right, investors didn't quite know what to make of it. But I think it's pretty clear now that it's certainly it's been E&P is about 28% revenues. It's been your strongest growth market in 2025, I think, up about a little over 20% through the first nine months. If we were to take a step back, John, maybe list the top reasons why we've seen that kind of growth. It's been a healthy market. But just in general, what can you point to?
Yeah, for E&P, for sure. And by the way, not to forget semi. It's also grown well.
We'll talk to semi as well.
But E&P, certainly some of the growth drivers have been AI. And AI has actually driven a lot of the equipment orders for plating chemistry. And that was a bit of a surprise to us because that equipment is to make higher layer count HDI PCBs, as well as MLB PCBs. Because at the end of the day, when you put the chips on top of your package substrate, the highest end, most complex layers of the PCB, you still have to connect that to the rest of the world through HDI, less complex, and MLB, even less complex. And then the chemistries that come with that. So what's driven the growth in E&P has been just chemistry for AI, as well as equipment for some of these parts of the PCB industry for AI. So we're starting to see a lot of that.
One of the most important points we like to make is when we sell our equipment, the chemistry that goes on that equipment is ours. That's great for market share. And long term, we keep 85% of that market share on our equipment. When it's not our equipment, we compete on chemistry, and we are a market share leader in chemistry and packaging anyway. But having this now about four quarters of strong equipment revenue, this is a record equipment year for us. We've talked about kind of first half of 2026, strong bookings. We know we're going to be building through that. And every month, we are talking to customers about what they need for beyond that.
You alluded to AI being a driver. Can you give us a sense as to the E&P business?
Yeah.
What, in terms of both equipment and chemistry, what that may represent? Is there a way for us to quantify that?
Yeah, maybe Michelle, you can talk about the chemistry part.
Chemistry, sure. So it's really hard to be precise about that because we sell into kind of high-end and complicated PCBs.
Probably 10% now?
Yeah. So, chemistry revenues, we estimate about 10%. That's about double what it was in 2024. But we don't have perfect visibility in terms of the end use, whether it's AI or otherwise. But that's what we've been able to trace through.
Yeah, and that's just chemistry. The equipment, of course, is on top of that.
You've said it's been a record year.
It's been a record year.
You think the bulk of that growth that you're seeing is tied, you think, to AI?
Yeah, because if you think about dishwashers and washing machines, they're not growing, I don't think. And even smartphones, this is HDI. It's a little better year this year, but it's been pretty flattish, right? PCs, HDI, and some substrates. So all of it is really AI for package substrates, AI for HDI, AI for MLB.
In addition to the backlog strength, the strength in bookings, how do we think about this as far as it extending into 2026, converting into revenues, I guess?
Yeah, yeah. So when we get an order for a piece of equipment, for plating equipment, it takes us four to six months to build it, install it. And then it takes our customers another six to nine months, depending on their business, to qualify it and use it, and then bring it up to ramp. So when we look at a piece of equipment from the day we get the order to revenue, it's in that 18 months, a year and a half to two and a half time frame. And over time, it ramps into high utilization. At full utilization, for every $100 million of equipment we sell, we will get $20 million-$40 million of high gross margin chemistry every year for decades. So $20 million-$40 million for every $100 million of equipment.
and that's just when it's on our equipment, but along our chemistry on other people's equipment.
And so from the time it's commissioned, John, the equipment is commissioned, how quickly do you see that revenue level?
Yeah, that's kind of that nine months to 18 months. It really depends on the customer. And they have to win, number one. And then if they win, likely they order more equipment, by the way. But then they have to ramp it.
So we've seen good growth in chemistry this past year. If we continue to have these kind of tailwinds, and it sounds like we do, is there any reason why you wouldn't anticipate the chemistry demand? And I'm thinking in terms of the equipment that you're putting into the field, you wouldn't continue to see pretty healthy chemistry demand in 2026.
Yeah, we'll see. So yes, I think we will see healthy demand in 2026 because the equipment we've been putting in, in 2025, some of that might start hitting in 2026. But in addition to that, obviously, the package substrate that's driven by HDI, that chemistry is already growing. You did point out, I think we grew 12% or something year-over-year in just the chemistry. And so that is when the industry has gone from 40 layers for an AI server board to, sorry, from 20 layers to 40 layers. And we're already working on 60 as an industry. People are talking about 80. People are talking about 100 layers now. So these layers are made one layer at a time. And they might go through our tool, one layer at a time. So we are a square inches consumables model there. And so that is great.
Consumables also, by the way, have a short shelf life. So our customers can't buy a year's worth and then sit on it. It has to be kind of several weeks, maybe a couple of months. And so this is really, I think, portends well for 2026 for chemistry and beyond that if AI continues to ramp and the number of layers continues to grow.
The margin profile on that chemistry business.
Yes. The margins we publicly disclose, the MSD division, the Atotech division, is in the mid-50s. And that's always a mix of chemistry and equipment. When there's more equipment, the margin goes a little lower. When there's less equipment, the margin goes a little higher. So think mid-50s for your modeling for chemistry revenue. Our nearest competitors are much lower. And so that tells you a little bit about our differentiation. I think, personally, that we have the best R&D team to provide new chemistries and a new roadmap. That's more and more important as you get to higher layer counts. Because you think about it, when you're making five layers, you have a certain yield per layer. So that fifth layer, it's important. Because if you screw that up, you kind of the four below it, that's all gone. Think about 40 layers. You're on layer 40.
You have 39 layers at work, and you can't screw that up, so yield becomes really important. Chemistry becomes more important. The equipment becomes more important. Think about going 60, going to 80, and so scale to do R&D is really important, and Atotech/MKS has the biggest R&D team for electronics. In addition to that, we also have the biggest footprint of tech centers co-located with our customer clusters, as well as analytical labs and material science labs, so these are labs that are co-located with our customers, so something happens. I don't know if the chemistry is contaminated. Send me a sample, literally in the same city, and we'll analyze it for you with all the analytical instruments that we have. They're consistently tested between all the sites so that we know that the results will always be correct.
There's also material science labs where we'll cut the PCB to make sure the holes are filled, to make sure there's no issue with the various layers. So this is the kind of support that we have that no other competitor has. And I think that's why we're valued. We're going to be not just their R&D partner, but their yield partner. That is very differentiating.
And this really ties to the point you made earlier about the benefits of being the only equipment and chemistry supplier. I mean, I think some people say, well, it's maybe just because they're in the customer's facility, they have the opportunity to continue to sell. But it's more than that.
It's much more than that because, by the way, we don't sell every piece of equipment because not every piece of equipment is differentiating. There's some that are more commoditized, and we don't do that. The kind of equipment we sell is horizontal plating equipment, meaning the panels are going through horizontally versus vertically. And that has allowed us to demonstrate better performance in terms of yield for plating. Just to be clear, we are the only company that has plating equipment, plating chemistry, and laser drilling equipment. No one else has all three. A couple of competitors have plating chemistry and plating equipment, much smaller companies. So really, when you have all those knobs and the problem is more difficult, the yield problem is more difficult, it could be the laser didn't drill it right. It could be the plating chemistry was a little off.
It could be the plating equipment was a little off. Now I've got all three knobs to turn in my tech center. So we have all three components in our tech centers co-located with the customers. It's not just in Berlin. So our customers can come in and troubleshoot with us with all the knobs. And so if they couldn't, they'd have to have three different companies. They'd be shipping stuff between three different R&D centers. And that takes time. And also, things happen when you ship versus all in a line. We're trying to emulate what our customers see and then solve their problem faster because we have more knobs to turn.
I think when we look at the example of the equipment orders that we've been getting recently, one of the changes that we had to make with the equipment was, well, the boards are now, remember, 40 layers, 20 layers. There used to be five. The equipment that makes five can't make 20 because that board is a lot thicker. So we had to make adjustments on that equipment. We could just go do that a lot faster because we understand that if we do this and we modify the chemistry, we can solve that problem much faster. Then a customer asking an equipment guy, can you modify your tool to go from five to 20 layers? And the chemistry person, can you modify your chemistry to be able to do that?
So this is the kind of speed to solution that we think is differentiating for us and more and more important for the industry as things get more complex. When things get more complex, it takes longer. It just takes longer to solve these problems. The industry is not going to wait. Whoever can get there faster is going to win. We want to enable our customers, the ones who want to partner with us, to get there faster.
Do you think you've gained share in 2025 in the chemistry side of the business, or is this just a case of rising tide lifting all boats?
Yeah, we measure market share all the time. Let's put it this way. I think we're pretty comfortable where we are. Let's put it this way. We're market share leader. The next two biggest competitors in chemistry together is our market share. So we don't have dominant market share for sure, but we have the strongest market share, and I think the equipment installations that we've been talking about for the last 12 months, as well as the future, perhaps 12 months, that portends well for market share growth because that is our chemistry only.
You mentioned the laser drilling business, which was the old Electro Scientific Industries business. My impression is so much of that business was tied to handsets and the overall growth of the market. Are there some other factors that are contributing to the growth in that business?
Yeah, maybe you can take that, Michelle.
So yeah, we're a market leader in Flex PCB laser drilling equipment, which predominantly serves the mobile device market. So that market's cyclical, right? It'll fluctuate quarter to quarter, year to year. But we had a good year in Flex in 2025. Beyond smartphones, you can think about wearables, so AirPods, the watch. Those all have flex circuits. But in the coming years, we think the growth is really within handsets, really through increased content or features like the foldable phone.
Right. And I was going to bring that up, Michelle. I mean, there's, I think, a widespread view that we're going to see more foldables, including a foldable from a pretty large consumer electronics company getting introduced. And the benefit for you from that change in form factor, it would require more laser drilling equipment.
That's right.
Yeah, for Flex, for sure.
For Flex.
Yeah, yeah. But there's already foldable phones out there, right? Samsung and Huawei. But you're right. The notable missing phone maker, if that takes off, that would change things for the positive.
And we'll shift over to Semi, which is probably time to do. But I just wanted to ask you about the competitive environment just in light of the spinoff of the DuPont electronics material business, which I guess now is known as DuPont de Nemours, or I'm not sure.
I don't know either, so.
Do you see what are the ramifications, implications of that?
Yeah. Well, look, Jon Kemp, he's a great guy. DuPont is a great company, great competitor. But I would say this. DuPont is strong in the semi side of plating, which we're not. It doesn't mean they're not in PCBs as well. So we compete in the PCB part. But that's where we're the strongest. So they're a good competitor. I think they should be more flexible. It's a smaller company. That's the whole idea. But they've been a good competitor for a long time. And we tend to win more, we think, since our market share will tell you that. But we certainly expect them to remain a good competitor. And I wish Jon well.
So on Semi, the semiconductor business, through the first nine months, actually, I think it was up around 15%. I guess the upgrade in NAND, was it Q2 where you got some nice tailwind from that? What were some of the other contributors to that growth in Semi through the first three quarters of the year?
Yeah, I think part of it is the fact that inventory burned down is finally done. And so we saw the full year's effect in 2025 versus there was still a little bit in 2024. So that naturally gets us a year-over-year growth. The NAND upgrade does add to that in Q2. And then I think in general, things are picking up. In general, things are picking up a little bit. I think the whole industry is preparing for what potentially could be a great next two years. And so we have been in this a long time. We've seen these cycles a long time. We know that if you don't get ready and you're behind, you never catch up in a ramp. We've done that. We've seen that so often. So we are preparing, as I'm sure most of our peers are, for the potentiality of it.
You've seen a lot of announcements of fab buildouts for DRAM and HBM for sure and Logic for sure. I guess we're all waiting for TSMC's commentary soon. We haven't seen a lot of commentary on NAND yet, except for upgrades. And so I think the discussion of the meeting is, when will you see that? And I guess I would say it feels like it should happen, like it should have happened. So we're kind of assuming something will happen there. And as I said earlier, we're levered to 85% of WFE. So we're seeing all of that now. And NAND, should it have more upgrades or a greenfield even, that'll be icing on the cake for 2026 and 2027.
Remind us, you're heavily over-indexed to NAND.
We're slightly over-indexed on one particular product, power. But remember, we have 20 different product lines all around vacuum chambers. We also have litho metrology inspection, optics, lasers, etc. So we're addressing the ASMLs and KLA product lines as well. But yes, if NAND takes off, our market share and market position in high aspect ratio power delivery is very solid.
If it doesn't, if we still see a muted recovery, but we are, as you alluded to, there's been a lot more positive commentary around DRAM. How do we think about that in the context of your ability to outgrow WFE, which historically you've been able to with the exception of the inventory adjustment year?
Yeah. So typically, through cycle, we've outgrown WFE by 200 basis points. So whatever CAGR WFE is, we've outgrown 200 basis points. That's just data. Typically, on an upcycle, we outgrow by a lot just because people are pulling our products. I think in the past, when WFE is 10%, we've been 25% or 30% on the upcycle. In the downcycle, there's inventory burn, we underperform. That's why we go through the cycle. We calculate through cycles. I would say this year was 15%, as you pointed out. I don't think WFE grew 15%. So we're already seeing a little bit of that momentum as things start picking up, and we haven't really picked up yet. We really haven't picked up. If you look at our Semi revenue, it's incrementally better, quarter and quarter.
But it's not what we have seen in the past when there's a ramp where that Semi revenue really rockets up. So I think it's in a good place. Our market position is very strong in WFE. And I think as the industry picks up, we're not worried about outgrowing WFE.
On the DRAM side of the business and all of the discussion around HBM as it relates to AI, can you maybe help talk to us about which areas of your Semi business will typically benefit the most from that?
Yeah, I think when DRAM takes off, it's relatively more Dep/Etch than litho. When Logic takes off, it's relatively slightly more litho than Dep/Etch for the reasons of shrink versus verticality of processes. So we're still more levered to Dep/Etch as a company, even though we have much more litho metrology inspection now than we ever did in the past. So when DRAM takes off, we should be a beneficiary of that. When chips go vertical, so even Gate-All-Around , which is not DRAM, there's soft etch, there's soft deposition, there's ALD, soft clean. All those things are product lines that we have. So when things are more complex, we tend to see that upside just because of the breadth of our technology. And also the fact that we're highly engaged with the entire industry in terms of what's next.
When you're at 85% of all the equipment, they're probably going to come to you with, here's our next thing. Here's the next thing that's going to need you to do this to this particular subsystem. So I think DRAM and Dep/Etch is what I would call potentially a tailwind in 2026 just because there's more Dep/Etch going for not just DRAM, but also some of the Logic processes.
How satisfied, John, are you with the progress you've made on the optical side of the Semi business? You've touched on it, but I'm curious as it relates to how we might think about the opportunities over the next one to two years in that area.
Yeah, I think we talked about world-class optics. This is our effort to be more important to that supply chain for lithography metrology inspection. To do that, we had to make investments for people, for CapEx. We did that five years ago. We started that five years ago. And we grew that revenue from $150 million-$300 million. So that was great progress. I think this year was a little flatter, kind of because the industry is a little flatter, lithography metrology. But I think as we start looking at WFE numbers in 2026 and 2027, they're incrementing, pretty much a lot bigger. You'll see also the world-class optics revenue grow as well. That's what our expectation is. So we're not done yet. It's still mid-innings, I would say, in terms of our ability to gain more share in lithography metrology and inspection.
Maybe a last thing on semi. This has been going on for several months, but folks seem to have settled in on this stronger second half, 2026 recovery, strengthening. Is that consistent with some of the conversations you were having with customers back in Q3? To what extent have those conversations maybe changed?
Yeah, I would say in Q3, it was more of a second half, 2026 discussion. And I think even in our last earnings call, 90 days later, I think it's incrementally more positive. So I think two things are happening. One is people are worried about 2027 being even higher. And that might add to pull-ins because you want to kind of get ahead of it before you get behind. As I said earlier, you don't want to be behind because during a ramp, you never catch up. So normally, if our customers need to ship in Q3, we would see those orders in Q2. That's no change there. But if they were worried about not just the shipment in Q2, but what they're going to really need in 2027, you could see them wetting the pipeline a little more. And I think many of them are, as we are.
People are preparing for a potentially sustained ramp, and when there's a sustained ramp, you kind of want to wet the supply chain a little more than if you just saw the next three quarters are going to be good, so I think that would be the dynamic that we're checking on.
So reflected more so in the Semi outlook for Q2, we would see it in that.
Maybe, maybe. I think it wouldn't be a surprise. We're not planning on it because we haven't been told about it. We are planning internally to do it. But if our customers pull, then you'll see it.
Maybe shift to the financials just in terms of the target model that you gave back in late 2022. You've achieved these gross margins, the 47% plus gross margins sooner than I think anticipated. At a lower level, I think of revenues, fair to say, that was, I guess, mixed related, some chemistry and some other things that you've done. I'm wondering how we should think about your target in terms of getting to the 26% + operating margins that you highlighted back in 2022.
Yeah. So we're really pleased with the execution on the gross margin side. We also think it obviously demonstrates the value our customers see in our offerings. To your point, we operated at that 47% level all through 2024. That trend would have continued in 2025 had it not been for tariffs. But we've managed through that really, really well. In our Q3 call, we had said by Q4, we were going to offset tariff cost dollar for dollar. But we'd also still see a drag of about 50 basis points because it's still dilutive, right, even if you're passing through those costs. So we're continuing efforts to try to mitigate even that 50 basis points. But that'll take a little bit of time. But beyond tariffs, mix and volume are really the key factors to expand gross margin.
So we've talked about the tailwinds that chemistry revenue provides, just given high margin. But we've also talked a lot in 2025 about the tremendous strength in chemistry equipment. So that's low margin on the equipment. But it obviously is a good leading indicator that we're going to see future chemistry revenues at that high margin with attach rates of over 85%, so for a long period of time thereafter. So we've provided a flow-through model of about 50% on gross margin for any incremental revenues from current levels. On the operating margin side, we have been super disciplined in terms of managing costs. We kept our 2023 and 2024 operating expenses flat, trying to manage through the cycle. In 2025, we did say the run rate was going to go up 250-260. We've stayed within that range because we had to invest in the business.
But we've really set the cost structure at a place where we believe from here on in, we'll start to see some leverage as that top line expands.
Do you feel on the R&D side there are areas that you might have to step up in terms of the way the market may be changing?
Yeah, we look at R&D all the time, and we're not afraid of stepping up a lot, and we have in certain areas, so we're not starving R&D at all. In fact, we're very disciplined about it. The unfortunate thing for my team is that I'm a tech guy, so if they say something that doesn't make sense technically, they're not going to get the R&D. But we've been pretty happy with our ability to afford that, and I think scale helps. Scale helps. 40% recurring revenue, consumables and recurring service revenue helps at great gross margin and cash generation, so we've been able to meet any kind of customer demand about, hey, can you put more R&D here, and we don't do it for every customer, but for certainly some of the customers that have great opportunities, we've been able to do that. No issues with R&D.
Talk about debt leverage and some of the debt leverage targets. Your net leverage while exiting Q3 was.
3.9.
3.9. And then trailing 12 months. You've made significant voluntary debt payments, what, about $400 million through October. Going back to the analyst day, you talked about bringing gross leverage 2x longer term. A lot of puts and takes. The market has changed. There's been changes in the market. But I'm just wondering, since you set that goal, how you're thinking about it.
Yeah, so we're really happy with the cash generation of the business. It's performing incredibly well. To your point, we've prioritized deleveraging. So we've paid down $400 million this year. That's on top of over $400 million that we paid voluntarily last year as well. Beyond that, we're proactive repricing, refinancing our debt. We've cut our interest expense by almost half since the exit rate in 2023, just taking those actions. From a net leverage standpoint, right now, kind of near-term goal is 2x-2.5x is where we'd like to get. We think that's appropriate given our view of the business right now.
We have a few minutes. Any questions from the audience? Okay.
Do you mind if we talk a little bit more about some of the bottlenecks, particularly as it relates to lithography?
Bottlenecks for delivery of tools, especially in lithography. I would say lithography subcomponents, if you will, subsystems are much longer lead time as lithography tools are. So there's less of that up and down kind of things. So at least on the products that we are supplying, we feel strongly that we won't be a bottleneck for lithography. I can't speak for everything else. ASML should answer that. But I feel that we've got plenty of capacity to deliver whatever is being asked right now.
Any others? Okay. Oh.
Just incrementally more positive conversation versus 90 days ago. Can you speak to how diversified that is? Is it just one or two customers or how?
Yeah, well, we talked to all the customers for sure. But I would say a lot of it is driven by memory, DRAM particularly. So that's really been the increment. And I think Logic has always been big and strong. And we'll see what TSMC says. They might change that view. But it's really been driven by the move from many chip companies to HBM, which is less efficient per bit, and so taking up capacity. And so they're all adding capacity right now with the shell and building shells for the future. NAND, I think, is kind of, well, when will that happen? And I think it has to happen to match the memory and the Logic. So at some point, it will happen.
And I think we just got to get visibility on that because remember, there's got to be a shell first if you're going to do Greenfield NAND versus just upgrades. I think upgrades will happen. But I think at some point, we kind of expect Greenfield NAND has to eventually happen. But a shell takes a year and a half, two years to build. So we're kind of waiting for that to happen.
Okay. Great.
Thank you. Thank you, Jim.
Okay. Thanks, everybody.