Ichor Holdings, Ltd. (ICHR) Earnings Call Transcript & Summary

Great. We're going to get started. I'm Tim Arcuri. I'm the semiconductor analyst here at UBS, and very pleased to have Ichor Holdings with us. We have Jeff Andreson, who is the CEO; and we have Greg Swyt, who is the CFO. So pleased to have both of you.

Yes, thanks for having us.

Thank you.

So maybe to start us off, just provide a quick overview of where Ichor fits within the supply chain and give us an overview of sort of customers and applications that drive the majority of your business.

Yes. So what we are is a critical supplier to the process tool manufacturers as well as the EUV manufacturers. And so our largest customers are Applied, Lam. They're about 70% of our business, about equally weighted, followed by ASML, which last year was 11% of our business and just crested over 10%. And then the fourth largest customer is ASMI. We also have about 5% of our business that's non-semi. It's largely, kind of, I would say, aerospace, defense and commercial air or commercial space like, SpaceX, so is a customer of ours as well. About 60% of our business is gas panel assembly. So it's build-to-print business, gas panels that go on to process tools. And that's inclusive of the gas delivery system we do for the EUV tool as well. And then we have a component of our business that addresses chemical delivery or wet chemistry tools. And this is where we're seeing a lot of the high-bandwidth memory, advanced packaging business growing year-over-year. And then we have a components business. We started with weldments first, which are really the gas lines that go to and from the process chambers to the gas panels. And then we also have a precision machine component business. And now we've kind of pivoted to a strategy where we're starting to develop products that we can integrate directly onto our gas boxes, which is really kind of a margin stacking play, whereas we used to be buying it, now we're starting to make it. So that's kind of in the early innings as we go today. So we're kind of knee-deep. We're very leveraged to dep and etch, which in 2025 is going to be a stronger year than 2024. We'll probably get back to that at some point. But that's where we fit in the ecosystem.

Great. And I'd be remiss if I didn't ask the next question about the 200-plus pages of export control documents that came out yesterday. And sort of how do you assess how that impacts you? I mean, your customers have come out and said, "Well, there's -- AMAT saying, yes, it's like still within the range. Maybe it's toward the lower end of the range but it's still within the range." So how do you sort of think about all that?

Yes. Well, one is late-breaking news always happens right before you come to an investor conference, which is good. But we're not seeing a significant impact today. I don't think it's going to change anything that we've said about the fourth quarter as well. So we'll be fairly aligned with that. We don't see it as a significant shift of what had been done before. I think those controls maybe are tightened down a little bit, but I think in general, it was a lot more muted than people were anticipating. We do not have any direct sales to China. So we don't -- we sell all -- everything that goes into China goes through one of our customers, so we're indirectly impacted. We do not sell anything directly. And honestly, most of our China supply chain is really starting to be migrated out of China as well.

Great. So on your last earnings call, you discussed a number of mix shifts taking place within WFE as we look into '25. And I thought maybe we could first just talk from like a high level, what are the mix shifts that -- and you talked about dep and etch, which is going to be a big year next year, which I agree with. Is there anything else that is sort of a tailwind for your business?

Yes. I think from the tailwind, obviously, we see a much stronger NAND view. Now we're coming off of a very low bottom after 2.5 or 3 years of very little spending. We're seeing NAND demand come back a little bit in the fourth quarter, stronger yet again in the first half of the year is our current visibility. Gate-all-around, I think we're starting to see the benefits of some of that stuff coming around. Advanced packaging, I think, has been beneficial both in 2024 and it's going to continue and get stronger in 2025. Those are kind of our tailwinds. Obviously, there's some headwinds we have to offset. I think our lithography business will be kind of relatively flat for us, maybe slightly down year-on-year but based on maybe the second half, relatively flat. So we had anticipated that to grow a little bit. So some of these tailwinds are offsetting those headwinds. China will be down. I think, in general, everybody thinks China will be down. But if WFE is up even modestly, that just means wafer fab equipment outside of China is stronger and that kind of benefits our largest customers.

Great. And how about what's going on in NAND? Because I -- so I wrote a big note on what the upgrade opportunity is, and it seems to me like there's $40 billion over the next 3 years, 4 years. I mean, it's a lot. And you're obviously levered to Lam and so that is probably a second derivative, good thing for your business. So can you kind of talk about that?

Yes. One is, I agree with your note. It's a good -- I mean, I knew a lot of the industry was under 200 layers. I think a lot of this is being driven by solid-state drives, maybe not going into PCs because we know those aren't too strong right now are recovering, but it's really going to be around the storage infrastructure. And I think we're starting to see the beginning of that, and I think that's what's driving a lot of this. And I think once the industry, and maybe you can agree or disagree, but I think once they start making the move where everybody is moving above 200 layers, 256, some are in the low 300s, others have to keep up. Otherwise, they're not cost competitive.

What about advanced logic? Obviously, gate-all-around is more dep and etch-intensive, so that by derivative is also good for you, I would think?

Yes. So gate-all-around has about 30% more process steps so that can only mean more process tools. So it's going to be good for those of us in that particular area. But you're also seeing additional applications. ALD is probably the one that has more and more layers, selective etch. Now they're doing horizontal etching as well. You're seeing a much more epi. Epi is pretty low volumes for us. We do generally all epi gas boxes because both of the largest epi providers, we have 100% share. And so all of that, I think, is going to be net beneficial. And it's all just tying into the fact that dep and etch is going to be a much stronger year than we've seen in the last 2 years.

So when you look at your business relative to your biggest customers' businesses, have you been growing in line with them, not every year but over time? Or have you been outgrowing them, meaning that your capture rate of their revenue is actually going up?

Yes. I mean, I think if you looked at it year on year on year, you really have to -- we kind of look through to shipments. So obviously, coming out of the pandemic, there was a bunch of deferred revenue at one of our customers. And so the build rates were actually down and we mirrored that largely. But we've been gaining share in different areas. A lot of share gain has been outside in the components business. And what you guys won't see on the top line is the fact that we're now, instead of buying a valve or a substrate, we're making them ourselves. So that doesn't change the top line. That just allows us to bring in the margin stack. And so from the share of wallet, we're actually gaining. The market is about $2.5 billion for gas boxes is our best estimate. And within that, we would have bought it and put it in. So our revenue doesn't necessarily change but the size of the wallet that we're getting is getting bigger every quarter. So that's been ramping through 2024 and it will continue to go into 2025.

And one of your major customers is a bit of an outlier talking about they're pushing pricing higher. Does that -- when you hear that, does that create you say, "Well, look, they raised price, then there's -- they have more of a margin envelope that I can then maybe price into as well?"

So about 60% or 65% of our business is cost-plus. So we would have to go back and renegotiate those rates. That might be a little different. Where we could benefit on that is in the component side of the business, to some degree, and the manufacturing costs. I would tell you that it's like the pressure on base materials and things like that is really -- it's kind of muted now. So we're not seeing huge cost increases. They're still primarily focused around the labor dollars and we'll generally try and pass that on at some point. But they don't generally take their finger off a price/cost reduction or it's even if they're raising prices.

Right, right, totally. Another area of strength you mentioned is advanced packaging. How does Ichor participate in that part of the market?

Yes, largely around our chemical delivery side of our business or plastics, essentially. So we participate in ECD tool at one of our customers and their spin cleans. And this is where we're seeing the ECD side go up quite a lot year-on-year. And so that's where we benefit today in high-bandwidth memory.

Great. And then can you also talk about EUV? I know you said you made a gas box for litho.

Yes. So we -- it's a gas distribution system. So it's a little different than a process tool where we're not putting process gases to change the character of the wafer. This is really around keeping the reticle clean. So we sit right on top of the reticle. We're single sourced. The customer generally does that because the volumes are relatively small, but the complexity is infinitely bigger than the other gas boxes we do. The average, like for us, when you look at our content versus the ASP that continues to go up, it's a fraction of a percent, whereas process tools are 5% or 6% of an ASP for us. So when you see that balance starting to shift, we actually benefit a little bit more on the other side, given the higher content of gas delivery and the process tools.

And how successful have you been? I mean, I always think of the company as being levered to the top two customers. But there's always this other group of customers that were always sort of potential for you. So how do you view the diversification of your customer base? How is that?

I would say, if you went back 5 or 6 years ago, ASML was low single-digits millions and it's grown to 10% or 11%. So that has helped us balance because we were -- as Tim remembers, we were 90% Lam and Applied at one point. And so now they're about 70%. That may increase with the recovery of NAND, and Lam has traditionally been our largest customers. ASMI is our fastest-growing customer. And there's a couple of the smaller players that we're starting to now have the bandwidth to go and pursue that are the smaller equipment suppliers.

So given that the increases in NAND, gate-all-around and advanced packaging are good for your business, how do you think about WFE next year? And really, I mean, you're not -- you just take your cue from whatever the tool company think. But like how do you think about the market growth, and then what your premium because you -- I think we agree that the vectors are pointing in your favor? How do you think of your growth relative to the market next year?

Yes. I think if you were to go out and obviously, we get a lot of input from our customers. We take your input and others and we look at it all and maybe kind of single digits, some are high single digits, some are low single digits. And for us, when we look at next year, we think we can definitely outgrow the market even in a flat environment. Some of it is as we are coming through kind of the inventory correction of our customers and then us on our supply chain, that's going to largely be over. It's largely over for most of our customers and some we still are working through some of that so that's a net benefit. And then dep and etch are just -- it's a very consistent message that they're going to be a stronger year than they have been for the last 2 years. So that will also -- because of that outgrowing the market. And I think in general, our customers are outgrowing what most people think are WFE. And then we benefit from that as well.

And how do you think as multiples in semiconductors have gone up and semis are now the largest group in the S&P. And all of the excitement around AI, there are a bunch of global contract manufacturers that have become very big. But I've always thought that if there ever really is a fire that is lit under semiconductors that some of these large multinational contract manufacturers would want to acquire assets to gain more exposure to semiconductors. And I always thought of Ichor as being one that would maybe fit the bill. So how do you think about like your business and your competitive advantage relative to like one of these large global contract manufacturers? Could they come in and do what you do?

Without acquiring us, I would probably say it would be a very big change in their business model. The reason that there's very few of, let's just say, the gas box integrators in operation today is because we're low volume, high mix. And so we all carry inventory that's at a much higher level. And if you're going to come in and be a contract manufacturer, they're trying to turn their inventory, say, 15 to 20 times, we're trying to crawl back to 4 and get to 6. So it's a bigger level of investment. It's harder to do. Over my tenure at Ichor, a few people have tried but nobody has really penetrated to any significance. So it's a very difficult business to do. And plus, it's very difficult -- it's a very sticky demand. I mean, our customers are not looking to add somebody that has to go through a massive qualification effort as well. So what we have to do is, one, be nimble; two is, adjust to changes in supply. So it's really operational execution that is where we kind of have our secret sauce. And then now we're developing products and we're bringing some IP to the table as well, which is going to raise our gross margin in the long term.

Does this investment, we always think of semi equipment being leading edge, leading edge, leading edge. But the lagging nodes are now 60% of Applied's foundry logic business. There's been a ton of investment there. How does that play into your business?

Well, I think what -- there's certain areas that need higher level of technology, and I'll just take flow controllers because we're developing flow controllers. So pressure-based side of that $900 million market is really targeted almost entirely at the most advanced nodes in DRAM and in logic. And so what we're also doing is developing parts that can be put on gas boxes that can go backwards compatible. So when you talk about the foundry logic business that's basically trailing edge, I mean we're generally agnostic to the gas box. It might be -- if it's older, it might be a little bit smaller, meaning fewer gases, the more sophisticated you get, like the most advanced DRAM boxes now are like 20 gas lines, whereas like a dep box might be 14, 16 at the most advanced edge. So we're happy with any demand to build it because all of the new components we're doing will be compatible with anything backwards or forwards as they come out.

Maybe turning to you, Greg. So can you talk about your strategies that are to drive gross margin higher?

Sure. So one of the big strategies is what Jeff has talked about is the bringing in our own IP that -- our own in-sourcing of the product and getting the margin stack from that versus buying it from a third-party. So that's one of our biggest strategies is to drive the gross margin. Historically, we were driving around 20% roughly flow-through on the incremental revenue on the margin. Today, we're driving at least 25% because we're bringing our proprietary product into that cost model and getting the benefit of the margin on that. The strategy to drive that even further is bring more of that content into our product, which then, at least in the near term, will continue to drive at least 25% flow-through and then further as we get more of that into the stack, north of the 25% and even up to 30%.

And how difficult has it been to bring that in-house? I mean, that's something that I would think...

It was not that easy. One is during the ramp, it's almost impossible to get the engineering time. And so now as things came down, they all knew what we were working towards. And so we have -- now we had the capacity to fulfill the demand and things like that. So it's gotten kind of a head of steam from Q1. Q2 was bigger, Q3 is bigger, Q4 is bigger, Q1 will be bigger. So it's starting to cut in and so we kind of do it platform by platform. I think the most challenging new product for us is bringing the flow controller fully integrated gas box with 75% of our content. And so there are a number of those out there being qualified. We're qualified on four different types of applications out there. And so once those get qualified, then that OEM, that device OEM is going to then proliferate those as they go through whatever next expansion phase they have. So it's a little more of a '26 impact but that will have another big step-up, as Greg was alluding to '25, and then some, that then some, that will help a lot.

And how involved are the end customers, the chip makers in these sort of qualifications or is it all directly through the toolmaker?

It's all through the toolmaker. So they'll know if it's a new MFC, they generally are worried about the device, the performance at the device level and so long as it works and they'll adopt it but it all has to be integrated together at that level. Now a valve or a substrate, it never really gets to a device guy because they're -- valves open and close. They're more about the longevity of the product than they are about actually the performance on the wafer.

So Greg, you talked about gross margin flow-through. Where do you think -- I mean, like where could drop-through go? I mean, is like 30% sort of like that's an aspirational number?

Well, we have a target of overall longer-term operating model to get to 20% consolidated gross margin, right? So as we move through this process of getting our proprietary products in there, including what Jeff talked about with the flow controller, that will start driving us to that 20% model. But we also have to -- gross margin -- but we also have to start getting on the revenue, right? So as we move through this year at the $200-plus million run rate and get to a $250 million to $300 million run rate, we'll start seeing that incremental margin as well as the leverage as it flows through to the EPS. And so one -- to get to that, it's going to take about a $1.2 billion annual rate and then upwards from there to get additional leverage.

And if you looked at the gross margin flow-through first half, second half, I think it was in the 30% range. And we're coming off of a pretty low, so we don't want to brag on that or anything. But it's possible and you will end up having to get kind of north of that to get to that 19% to 20%, but that's when the flow controllers start to kick in.

Got it. And I would think that there's probably a lot of operating leverage in the model, too, because as this drop-through, as you -- the internalization of these components, you don't have to put a lot of OpEx into that, I wouldn't think.

Correct. Right, we'll get the leverage. I mean, we've been investing in our R&D to get this proprietary product out there, to get the flow controllers developed. But as that gets out there, our investments will continue, but we're going to get more leverage out of our SG&A. We don't have to scale the SG&A. When the revenue grows, we can moderate that. And then we have some other strategies to manage that from a longer term. So we'll expect to get pretty good leverage out of our OpEx.

Yes. And I think in general, our -- outside of the acquisition we do with IMG, I think the next biggest piece of growth in OpEx has really been in our D&E, so -- but you need to develop products. So I think we're highly disciplined, as Greg would say, in SG&A. And we're getting a fully integrated Oracle system rolled out, so we're down to kind of the last region and that will help us again.

So I guess bottom line, Greg, just on earnings leverage, sort of what kind of business volumes and revenue, when you put it all together, what do you need to get earnings power, say, of $3 a share or even $5 a share?

Well, it's what we've just talked about, right? We've got to get the margin. We talked about getting that 20% up to that range, leveraging our OpEx and then getting to -- with our cost structure, getting to that $1.2 billion to $1.3 billion run rate will get us to that $3. And then as we move past that and drive to what the industry is driving to the $1 trillion, getting to about a $1.5 billion run rate will get us to that $5 EPS.

So then of the $1.5 billion to get to $5, you said?

In that ballpark.

In that range.

What does WFE have to be at $1.5 billion? Because if that corresponds to $1 trillion in industry revenue...

It doesn't.

It doesn't, right?

When we look at modeling out to $1 trillion, we kind of think WFE and opinions varies 150-plus. And so that would be, if you think about it, that would be a pretty high number. With the internal content, it's hard for you to see that, but it will be a bigger number, like maybe 1.5% of WFE or something like that. So it will occur before that, we think.

Okay. So like to get to -- bottom line, to get to like $5, you need 140, 150 WFE, something...

Yes, probably. Yes.

Okay. And maybe you can just talk also just about the competitive environment? Are you pretty -- I mean, you're obviously very embedded at these customers and unless you really massively screw something up, they're not going to kick you out. But how does the competitive landscape...

I think we'll see a lot more competition on the component side. But to some degree, we control our own destiny there because we're qualified on our own products, so we get to pick those. There's been several attempts to enter kind of the integration gas box area. And like you said, semiconductor is a very attractive market. It would not surprise me that others will try and come back in. The question is how do they come back in? It's very difficult to build-to-print and then come in and underprice us and then increase your asset base. So I think we have some barriers to entry that are pretty difficult for new entrants to come in. So where you're probably going to see them is on the component side, machining or additional weldment suppliers, which has occurred since the pandemic. But machining, there's a lot of machining. And we've got -- as you know, our customers are all moving to Asia. I mean, they're keeping production in the U.S. and other areas, but expansion is largely outside. So we have to keep up with that as well.

And what about non-semiconductor opportunities?

Yes. So about 5% of our business, it comes through our IMG acquisition that we did in 2021. They have positions in aerospace and defense, some medical applications. Their largest customers has now become SpaceX, which is a very fast-growing part of the business for us. It's going to take another step function in 2025. It doesn't mean that the 5% is going to become 10% but it's crawling along. So we're looking now to leverage the entire operational base across that. And so as you go in and IMG was primarily machining and services. And they've had some headwinds in their business that they have largely overcome to -- because they've grown the other side of their business. And so defense spending is actually going up and so there's been a lot of new demand created in that where -- and then around commercial space is where we're starting to see real strong momentum. And now we're trying to bring that into our weldment side of the business and then a different type of machine. They do kind of midsized and we do precision in our Minnesota operation and our Mexico operations so we can start to bring some of that and to try and continue to expand that side of the business.

And how is the -- are the margin profiles any different in the, like, for example, the SpaceX stuff that...

Well, I would say machining to machining, they're very similar. Yes, so they're not dramatically different than what we would see in just a build-to-print machining operation as well. But you can kind of think of them in the high 30s and maybe sometimes low 40s for machining.

Great. I think that's all the time we have. Thank you to you both.

Super. Thanks for having us.

Thank you.

Thanks.