Test, test.
Test.
You want to try your microphone? Say test.
Test.
Testing. Do you guys have a safe harbor statement you want to read or disclosures or anything? Okay, perfect.
Lam does that all the time. I'm like, everybody knows.
Right.
Our General Counsel is in here, so feel free.
Okay, good. Great. We'll go ahead and get started with the next session. Thanks, everybody, for joining us at the DB Tech Conference here in Dana Point. I'm Melissa Weathers. I am the Lead Analyst covering semicap equipment and memory here at Deutsche Bank. For our next session this morning, we are pleased to be joined by John Lee, President and CEO, and Ram Mayampurath, CFO of MKS. I had to remind myself to call it MKS and not MKS Instruments this time. Before we get started, if anyone has any questions from the audience, feel free to raise your hand and we can get a mic over to you. To kick us off, I'm going to keep it a bit more near-term focus just to kick us off. You reported earnings a couple weeks ago.
A pretty strong quarter in my view, which is very impressive given some of the other trends that we're seeing in WFE and in the semicap equipment space right now. Could you just recap us on what are the main takeaways from the quarter you just printed and what are the main takeaways you want investors to have walked away with after that print?
Ram, why don't you take that?
Sure. It's interesting to see what top line can bring to the overall P&L and cash flow, right? We had $743 million as top line, and it's been a while since we saw that kind of a top line. It's been eight quarters, actually. Semiconductor did very well. Electronics and packaging , which is a very profitable part of our business, did very well. That speaks for the breadth of the portfolio and the reason why we did that acquisition to take us away from the cycles of semiconductor. The flow-through and gross margin was very strong, 46.6% with the 115 basis points impact of the tariffs. It was the first quarter we saw the tariff impact. Without that, we would have been close to 48% gross margin. Cash flow was 14% of revenue, and OpEx scalability we saw. OpEx was in the low end of what we guided.
The takeaways are the steady improvements we have made over the last several quarters to the cost structure. We've been running 47%+ gross margin for seven, eight quarters now. Last quarter is the one where we drop again, mostly because of the tariffs. What the current cost structure will help do when that top line comes back to more normal levels. The flow-through that we saw is, I would say, the key takeaway.
What I would add, the other takeaway, obviously, is our progress on deleveraging. In the last two months, we prepaid another $200 million. The free cash flow is allowing us to do that. As Ram said, in a relatively muted market for us, those of you who follow MKS know how much it flows through in the top line growth. We're pretty happy with where we are today in controlling the costs and making our progress towards deleveraging and then spending ourselves up for, you know, lots of engagements with customers for design ones as well.
Great. Maybe let's dig into that top line performance. The WFE environment has been relatively mixed. I think we've seen some mixed prints from some of your customers in the last couple weeks. Can you just remind us where are we in terms of inventories at your semicap customers? How do we think about MKS's performance relative to the broader WFE market, especially in a world where maybe we don't have as much of an inventory overhang as we have?
Yeah, so inventory is burned off. That's good. You saw our first half performance versus our first half last year was 18% higher year- over- year, first half to first half. That's comparing ourselves to ourselves. There was still some inventory burn occurring last year. Upgrades happened, whereas last year didn't. That was a part of it as well. Growing 18% year- over- year is excellent compared to WFE, which is kind of 5%, give or take. You're right. Going out, it's hard to predict. Some of our customers are saying good, and some are saying not so good. We always look at, you know, long- term because quarter to quarter, even half to half, it can vary for sure. When I look at the macro trends for semiconductor equipment, it is more tailwinds for us now. A couple of them are NAND upgrades continuing to happen.
They can be lumpy, but they are going to continue to happen. We have really strong market share there. Remember, when you're upgrading, you have to upgrade from something that's already there. It's kind of been well- known that we're the only ones there, right? I think also our broad portfolio strategy is to allow us to take advantage of whatever inflections are occurring and how that impacts particular subsegments of critical subsystems. That strategy has allowed us to outgrow WFE by 200 basis points a year for 10 years. If you think about what's going to happen in the next five years at least, chips are becoming more vertical. The last five years is about lithium. The UV came in and we were able to print better. The five years before that, it was double patterning because lithium wasn't ready, right? These things always happen.
In the next five years, it's going to be probably more etch depth centric. We want to be a foundational supplier to lithium and metrology and inspection as well as depth etch. We've made some progress there, but we are still stronger in depth etch. I think that's going to be a relative tailwind for us going forward in the next five years because of an inflection in chip structure.
I want to dig into more of that because I don't think you get quite as much attention on the leading edge nodes that everyone pays attention to NAND and the memory piece. Gate-all-around, I think, is, like you said, it's much more depth and etch intensive, and you guys have decent share there. Any more color technologically on some of the ways that you're able to enable some of those critical depth and etch technologies?
Yeah, gate-all-around is another example of chips going vertical, right? Before, it was just planar, lay down a layer, print it, lay down another layer. Now, gate-all-around means a tube. You're going to surround that tube, and you're going to make a tube, and then you're going to get rid of stuff around it, and then you put the materials you want around it. That's crazy, right? A lot of that requires atomic layer deposition processes, putting things down one atom at a time in all directions simultaneously. A lot of the subsystems that we make enable that. We've been in ALD a long time with ozone systems. Now it's just a lot more ALD-type applications. You can see some of the customers who have traditionally been strong in ALD have really grown, you know, ASMI, for instance, as well as Lam and Applied.
Those critical subsystems, we're seeing that pull as well. The other thing that maybe gets lost is a lot of cleaning. Cleaning of surfaces after you remove some material. Now you're not cleaning just a surface, a flat surface. You're cleaning it, again, three-dimensionally. That, again, requires dissolved gas kind of cleaning. That's when we talked about in our earnings call, a big pull in dissolved gas for 2 nm applications. That's where it's getting harder. We have had competitors in the past, but over time, as things have gotten harder, we've been able to invest in making the next generation dissolved gas systems. I would say now our market share has continued to increase there. Small subsegment of our business, but when you add a lot of these things up, that's how we are able to outgrow WFE by 200 basis points a year.
You talked about lithography. I want to touch on your world-class optics business and specifically the semiconductor side. I think you had talked about a big design win or some increased momentum at some lithography and maybe some metrology and inspection. Can you talk about the progress you're making there and what the market outlook you're seeing is?
Yeah, so it's something that we started five years ago. We said, you know, lithium, metrology, inspection, a great market. We have some capability that limits us in terms of what other things we could do for these companies. They told us that. They said, if you invest in other capabilities, we'll give you more to do. We did more CapEx, more process engineers. We are developing new processes to make advanced optics, and we're putting together an optical subsystem. Five years ago, we were about $150 million in that subsegment, lithium, metrology, inspection, about $300 million. By the way, the first three years is just about investing in design wins until you see that volume. In that space, it's much stickier. Those critical subsystems stay for a very long time, decades, even longer than vacuum-based equipment. It's a little less lumpy as well.
We can see ourselves continuing to do this. I think it's still mid-innings, if you will, in terms of our ability to continue gaining share through investments.
Maybe let's go back to the NAND topic. I think when you first spoke, you said that the NAND upgrades happened past tense, but then you said that NAND upgrades are happening. Can you help us understand, at least from your perspective, what kind of buying trends are you seeing? What kind of upgrade trends are you seeing? What are the key drivers of that NAND upgrade cycle?
Yeah, so if you step way back up to the highest level, you know, what's the fundamental need for NAND memory, right? Non-volatile memory. If you go back, you know, in the 2010s, the reason NAND really rocketed up was because it was replacing hard disks, you know, for smartphones and for PCs. That's kind of done now. The question is, what's the next big application for NAND? You're starting to see some of it in AI data centers, right? It's not as much as DRAM, HBM, but people are starting to talk about, you know, high bandwidth flash. It's pretty exciting, you know, kind of taking NAND chips and stacking on top of each other. Fundamentally, it's going to be, you know, the market driver that requires us to, as an industry, to, you know, put CapEx into NAND manufacturing.
If I look at today, it's still most efficient for our customers' customers to upgrade. At some point, they're going to need capacity. Even today, some customers are saying, I just don't have a fab to upgrade, so I got to build, you know, greenfield. There are not a lot of players in NAND, so that's why it can be lumpy. Lam has talked about a certain number in terms of the opportunity. Their SAM is some subsegment of that, sorry. Ours is a subsegment of that because we're really talking about dielectric edge. You can see that when it happens, you know, that was the upside surprise in our Q2. It was significant. To Ram's point, it flows through very quickly.
As we think about those semicap customers, now that inventories have burnt down, I think we talked about this on the last earnings call. Their order patterns, do you expect kind of a rush back to refill their inventory, or what kind of rebound do you expect from the purchasing side from your?
I think we're shipping to end demand. I think that's the bottom line. Just to give you a little more detail, we have inventory at our customers. We own it. It's consigned. They always have some inventory to pull very quickly. As they pull, then we refill. There's a buffer, built-in buffer by design. Our lead times now are pretty short, back to kind of normal. Our power decks are pretty complex units, so those lead times are a little longer than our baratron or flow meters, but they're back to normal. Our customers can plan. They know that if they order X, they're going to get it in that same time period. They're able to plan and not need to over-order or order for inventory. I think we're in a good spot as an industry and certainly for MKS' inventory. We're shipping to end demand.
Is that the case across DRAM and foundry logic as well?
It's the case across all of our critical subsystems.
Is there any tightness in particular areas? I know HBM, we can get into the advanced packaging side later on in the discussion.
Yeah, it's advanced packaging that's got the tightness, and then the equipment orders, not in the semiconductor chip manufacturing part for us.
Got it. With that, let's transition over to the E&P side. This has been a nice profit driver for you guys in recent quarters and a nice revenue stabilizer, I think. At your 2022 Analyst Day, you outlined a revenue growth target, I think, for this business of GDP + 300 basis points. There are a couple of different segments within this business that you've talked about growing at different growth rates. Can you just remind us what are the moving pieces within this E&P segment? Yeah, and maybe we can dive a little deeper into each of those.
Yeah, so how we characterize the electronics and packaging market is PCBs, printed circuit boards, right? That industry can be segmented into three thirds. The bottom third we call multi-layer board PCBs. Those are lower technology, if you will, fewer layers, bigger features. Think washing machines and dishwashers. Big industry, though. We have market share leadership there. The middle third is HDI, high-density interconnect PCBs, I think smartphones. More layers, smaller features, more complex, higher ASPs. The top third is what we call packaged substrate PCBs. They're all PCBs and even smaller features. That is what you need to connect large chips through large chips. GPUs together or CPUs together or GPUs with CPUs and HBM. That's really what's enabling data centers and AI today. The top third growth, we think, grows at high single digit, middle third, mid-single digit, and lower third GDP.
We mush it all together, and that's where we came up with GDP + 300 basis points for the entire slot because we participate in all three. We're not changing our model right now today, but I would say some things for us to consider is we have been getting a lot of orders for equipment for HDI and MLB, those two lower technology levels. We ask why. We're not making more washing machines or smartphones than normal. It's all driven by AI. AI boards, the chips go on that higher level packaged substrate, but then it's got to go on to something else, HDI, and then something else, MLB, before it can talk to the outside world. Those HDI boards are many more layers than what you need for a smartphone, so more complexity. The MLB boards are many, many more layers than you would need for a dishwasher.
AI is actually driving the entire industry from the most advanced to the middle to the lower edge. What we're seeing in terms of constraints is a lot of our customers are ordering equipment for HDI and MLB processes, so they don't have the capacity. We've had, I think, four quarters now of very strong equipment bookings. Historically, it's been a very lumpy kind of business, chemistry equipment. It's been four quarters of strong equipment bookings. You can start seeing in our numbers as well.
I guess two questions on that. The sustainability of that spending, are you worried that it could fall off? The second question, how would you characterize utilization levels of the current capacity that's in place today?
The current capacity is very high. Otherwise, they wouldn't be ordering equipment. Equipment is lumpy. You know, we come from a semi-CapEx world. We just deal with the cycles. At some point, the equipment orders should slow down. We haven't seen that yet. In fact, after the first quarter of higher bookings of equipment, I thought maybe it's over, right? Here we are on the fifth quarter, and it's still looking pretty good. The most important point is that equipment that we're selling comes with our chemistry, almost 100% market share. That chemistry is, of course, much higher gross margin and consumables and is always going to be used for the future. Even after five years of installation, after five years of running that piece of equipment, we still have 85% market share of our chemistry on our equipment.
The install base of equipment that we're putting in now, and for the next few quarters, that's going to portend to great market share gains in chemistry.
Maybe let's just zoom out a little bit on that topic. Atotech, I think it's now been three years since the acquisition closed. Sorry, Ram, I know some of this predates you, but can you give us a high-level review of how have those three years been? How has the acquisition played out versus what you were hoping for?
Yeah, I'll start. Maybe Ram can add, but I think in terms of cost synergies, we got that. That was pretty fast and easy to do. I think in terms of the strategy, like the reason why we bought Atotech, we thought packaging was going to be more important. We thought packaging was going to be harder to do. We thought it was going to be the same movie as semi. That's all not just been reaffirmed, but even more because when we decided to make the offer to Atotech, AI wasn't in anybody's consciousness. AI just put an exclamation point on that strategy. The other part of the strategy was that having equipment and chemistry knobs was going to be an advantage to gaining market share. We just talked about ordering equipment with all our chemistry. That's another exclamation point on that strategy.
I think when we closed Atotech, the industry went to a downturn. It didn't look good, like, oh, well. Of course, interest rates were much higher. All these assumptions were kind of against us. Now you start seeing the strategy play out and you start seeing what we believed, what the reason of putting Atotech together with the rest of MKS. I don't know if you have anything else to add.
No, thank you.
As we talk about or as we think about some of those new packaging technologies, I agree with you. It's some of the most interesting growth drivers in the space right now. As we think about CoWoS, CoPoS, some of those new acronyms that we all have to learn what they mean, which one of those makes you the most excited? What should we get excited about for MKS on that packaging piece?
Yeah, I think they're incremental tailwinds. CoWoS, we all know, is TSMC's chip on wafer on substrate. That substrate's the PCB, right? The highest level's PCB. That's good for us. The wafer, the chip on wafer, that's a redistribution layer of silicon. We're not as strong there. It's one layer. CoPoS, chip on panel on substrate. Substrate stays the same. Chip stays the same. The panel is replacing the silicon RDL with a PCB. For us, that's great. It's an incremental tailwind, but it's one layer versus 10 more layers in HDI or 15 more layers in MLB. Then CoWoS, chip on wafer on PCB. The P's are different. That idea there is to get rid of the S, which is also a different type of substrate. It's the most advanced S, the PCB substrate. Let's skip that step. Go right to HDI.
If we can do that, it's cheaper and faster for the integrated server board. That requires more layers of HDI, smaller features than we've ever made. I think they're going to come to technology providers with more knobs in order to do that because if we could have done that, we would have done that. We couldn't do it. That's why we went to substrates, a different layer, a different type of PCB. Now the idea is if we can make HDI even better, can we remove that need for a substrate? That's going to require a lot more development. We're, of course, intimately involved with our customers.
Maybe if I could squeeze one more in, glass substrates, can you remind us how you plan on the glass side?
Yeah, people have talked about a glass substrate. There are two, when they talk about it, there are two different applications. A glass substrate core, of which PCBs are put on both sides, allows that package of PCB layers to be more rigid. OK, that's a great idea. It's one layer. We've been working on glass for 15 years. We hope it happens, actually, because we've been working on it for 15 years. If it doesn't happen, it's OK too because that core layer just stays a PCB, which we play as well. The other glass layer potentially is at the chip level, with chips going right onto glass. That's probably a little further out, I would say. There are a lot more technical problems to solve. Glass is something we've worked on. If it happens, great. If it doesn't happen, OK by us.
OK. I'll just pause here, see if anyone has any questions. If not, we can keep going. On the specialty industrial side of your business, this, I think, is a less appreciated part of your business. There are a bunch of different moving pieces. It's difficult to model. I guess for my first question, what's the most misunderstood part of your specialty industrial business that you wish investors could appreciate more?
I think, you know, we've talked about it, but it's leveraging the R&D we're spending in semi and E&P. It's not like we haven't spent any R&D in specialty industrials. Certain areas where we do, for sure, but much lower relative R&D percentage for the same revenue and the same profitability. It's a more stable business, many markets. You're right, it's hard to model, but when you mush it all together, it kind of is flattish, right? Recently, I think we've had some automotive headwinds and industrial headwinds, like everybody else. It's just gone down a little bit. It's a nice stabilizing part of our revenue stream. It's not that it's not part of our strategy to buy companies to grow specialty industrials. Our strategy is to grow semi and to grow E&P. It comes with a specialty industrial component that leverages that technology. It fits the model.
It's good for the P&L.
Does it even make sense to keep a lot of people ask why not divest the business?
Oh, yeah. I think part of our strategy is always our responsibility is always to do things. When we buy a company, make everything better, whether it's in semi, E&P, or specialty industrials. It's also our responsibility to look at each one of our businesses and say, is there a better home for it? We always do that. We've spun out a couple of things over the last 10 years, not many. Because when you make things better, it's like, why do I want to get rid of this, right? We always look at that as an option, for sure. Some of the things we've spun off have gone to other companies. They've grown it. The employees we spun off, there's 3x more. Their revenue is better because they're in that market. Those companies care about that market more than we would have.
We'll always be looking at that portfolio management.
Let's get Ram involved in the conversation. Ram, I think you're coming up on your one-year mark at MKS.
Yep.
Can you give us a scorecard? How have things gone versus your expectations? What has positively surprised you? What's negatively surprised you? This one-year anniversary.
That's a great question. I am very impressed with the execution the company has done. If you look at just 2023- 2024, the top line was slightly down, actually, flat almost. The gross margin grew 190 basis points. Operating income grew 180 basis points. OpEx was flat. Maintained all inflation, made up for all inflation with efficiency. The financial acumen of the leadership team and how quickly they get to execution has been very impressive. It probably goes back to the semiconductor industry background and riding the cycles and knowing how quickly to adjust. That has been very impressive. Agility, the tariff is a good example. How quickly the teams came together to quantify, forecast, identify mitigations, actions, and implement those actions was very impressive. The whole company got together with a core team from various parts of the business because, as you know, this has been changing too.
It's a very fluid environment, right? You come up with a strategy and next Monday morning, it's old. Then there's a new problem. That is what happened to us in Q2, where China was the biggest impact item identified when we did the earnings call. Soon after that, it became metals. That changed the guidance we had given. Being agile to adjust with the changes in the market, the tariff being one example, I find it very impressive based on where I have worked before. The challenges mostly are adapting to a market that's so rapidly changing and the geopolitics of it. I am sure at our cost structure, like we said before, when that top line comes back to more normal levels, you'll see much more flow through both to the bottom line and cash flow.
Let's dig a little deeper onto the tariff side. You talked about some, can you just remind us what are the mitigation measures that you guys are taking? I was very impressed with your ability to continue to optimize gross margins in such a difficult environment. I think a lot of companies have struggled with that. What are the actions that you're taking to mitigate those risks?
There are a broad area of actions we are considering. The good news is that we are having very productive discussions both with our customers and our suppliers. People are transparent. There's nobody trying to make money off of this thing. It's out in the open. Those mitigation actions include supply chain activities, using bonded warehouses where possible, routing production, the multi-site manufacturing capability we have helps us source products from different parts of the world as necessary, and some commercial actions and pass-through where necessary, right? It's a broad range of actions, and it's being addressed very openly with our key customers.
On the gross margin piece, you guys have seen a little bit of a boon from the E&P side, and that chemistries piece carries pretty creative gross margins to your business. It's gotten me questioning really what is the long-term gross margin potential of MKS. Could it have a five-handle? Could it be a healthy five-handle? As we think about the moving pieces on gross margins going forward, what are the levers that we should be thinking about in terms of what could drive gross margin upside?
The gross margin has been, you know, it's something the company has worked on for a while. Like you mentioned, Melissa, it's been over 47% for several quarters. Last quarter, like we said, without the tariff, it had been high, 47%, close to 48%. The fundamentals of the gross margin and the P&L structure and cost structure are very strong. It's these superficial things that happen that we need to mitigate, and there are actions in place to do that. The product cost structure work we are doing in terms of identifying manufacturing excellence programs, procurement savings, and design improvements will continue to improve the margins. The step change is going to really come from volume as that top line comes back. You know, I always say that the road to 50 goes through 48 and 49. Hopefully, you'll get a chance to ask the question again.
I'm getting greedy. On the OpEx piece, can you talk about where are your incremental OpEx dollars flowing through? You've kept it pretty tight in the most recent quarters, but there also are a lot of organic growth opportunities for you. Where is the OpEx priority at this point, and how are you thinking about the trajectory?
When the revenue was not growing in 2023, 2024, we held OpEx flat. It's something which we focus very carefully. As you saw in our Q2 numbers and Q3 guidance, we are at the bottom end of that range, the $250 million- $260 million. The investments we are making are in people mostly, and some infrastructure, but mostly to retain talent and help with long-term growth. Going back to the discussions we have had on profitability and cash flow, the improvements we have made are not by starving the business. We have invested in business in the P&L and in CapEx. Our CapEx is 45% this year. The debt repayment we have done is after making the necessary investments. The careful investments we initially look for always reallocating resources before we look for new dollars. The investments we are making in OpEx are tied to retaining talent and driving growth.
Maybe as a final wrap-up question for both of you, on the capital allocation piece, you are highly leveraged, and I know that's a priority to bring down that leverage. First, can you remind us where you are in that journey? What are your expectations or goals in the next 18 months on the leverage side? And then second, I’ll be greedy again, but you have a very strong history of M&A and inorganic growth at the company. How are you thinking about further inorganic growth opportunities and maybe some holes in the portfolio?
Why don't you start first, Ram, and I'll follow up with the strategy going forward?
Sounds good. You're right. The capital allocation priorities are first investing in the business to support growth and business continuity, and then focus on strengthening the balance sheet, which is lowering our debt. We are at net 4x leverage now. Our goal is to get to 2x- 2.5x over the last, in a period where demand has been pretty flat. In the last seven quarters, including the one we are in now, we paid down $800 million towards our term loan. That remains our continued focus. With our cost structure today, as that top line again comes back to more normal levels, we can accelerate that payment quite a bit.
I think, once we get to that 2x and 2.5x net leverage ratio, M&A will always be part of our strategy. I don't think big M&A makes a lot of sense anymore. We just did one. We have the foundations for our strategy, which is advanced electronics, semi-advanced packaging to make advanced electronics and be foundational to that. There are a lot of tokens that make a lot of sense. Fewer in the vacuum equipment side because legacy MKS did a lot of that consolidation. Probably more in optics and lasers because it's still a disaggregated market. Some in chemistry, but not much because Atotech is already a market share leader. Our standard for acquisition is higher now, the bar, just because we have made some big bets in the last 10 years that have paid off.
The reason you do M&A is to get there faster. Some of these big bets, though, are certainly a lot cheaper than buying somebody. We're getting there almost as fast. I think a bar is going to be a lot higher. Why should we buy this company? If it's such a good idea, why don't we invest? Why don't we do it ourselves? Especially if we're already halfway there. I think tokens might make sense. As we're waiting for Atotech to close, we did Photon Control. We didn't do it. It's perfect. Surround the chamber. Same sale channel. Great technology. Differentiated. Same sets of customers. Those things we'll do all day.
Great. I think that's a great place to end it. Thank you guys so much for joining us. Enjoy the rest of the conference.
Thanks, Melissa.