All right. Good morning. Welcome to twenty-sixth annual Needham Growth Conference. My name is Charles Shi. I'm a semi, semi cap analyst here at Needham. I cover Ichor. Here with me on stage is the CEO of Ichor, Jeff Andreson, and thank you for coming.
Thanks for having us.
So no need to really introduce about Ichor, so let's just dive directly into the fireside chat. To kick off this discussion, really just want to start with an observation about Ichor's 2023, right? Based on your 4Q guidance, I'm not asking you to give any more additional color here. Looks like Ichor's revenue is going to be down year-on-year, roughly by high 30%.
Yeah.
WFE looks like it's probably gonna be down single digit, depending on whose estimates are you looking at, right?
Yeah.
If you look at mine, that's single digit. Really want to be forward-looking and not looking backwards too much, but any chance we can start with a little bit of a post-mortem analysis of what happened last year?
Yeah. It's, so as we entered the year, if we go back to the beginning of 2023, and I'll kind of chronologically go through what we saw during the year is, as we entered the year, we were coming out of a period where we were chasing what we call chasing demand. We were barely onto our customers' MRP by the third quarter, and we all know in the third quarter, China embargo got announced, and then the memory guys slowed up tremendously in the fourth quarter around December.
That was 2022, actually.
2025. So, in our Q1 of 2023 was $225 million. I would tell you, our outlook entering the fourth quarter for 2023 Q1 was probably $425, something like that. So in essence, we didn't believe that our customers could have built that much inventory in excess of what they needed to build. So they are holding, our estimates is over $100 million of inventory, and so we don't get to replenish that until they burn it off. Obviously, you can look at our top three customers are all at the highest levels of inventory they've seen, probably ever. And so we underestimated that, and that's well over $100 million.
The other thing that you have to think about is the amount of that we're leveraged to dep and etch. I mean, most of what we do goes to dep and etch. Even whether it's wet chemistry and our chemical, it's generally a deposition process. So, our largest customer system shipments are down 40%. 3D NAND is down 70%-75%, depending on the estimates. And so that side of the market was down tremendously. And keep in mind that our estimate of just the amount of revenue that occurred this year in 2023, I shouldn't say this year, last year, 'cause it's January, we forget. But it's about $5 billion that rolled over.
When you kind of add up what Lam talked about and Applied talked about, and really, ASML talked about having to recognize revenue in the following year because they were missing parts, we would have delivered all of those parts to them. So we didn't get to participate in that portion of the revenue as well. The other part of our business with ASML is we have a 100% position in low pressure gas distribution for the EUV tool. And when you look at the ASP of, obviously of an EUV tool, say it's $150 million or something, our ASP, while it's significantly higher than an individual process tool, is a fraction of a process tool.
A process tool's gas distribution, kind of all in, is between 5%-6% of an ASP, and you'd have to go like 4 decimal points out on an EUV tool. So while that business for us is gonna grow 20%, say, 25% year-on-year, and they're pretty transparent about their growth in EUV as well, so it's very aligned to them. And we'll crest a 10% customer, it just wasn't enough to offset a lot of the inventory burn and the fact that we were so leveraged to dep and etch during the year. You know, having said that, we also gained share during the year, but, you know, you just can't push it through as fast as those big-ticket items are.
Mm-hmm. Mm-hmm. Got it, got it.
Yeah.
So maybe I wanna ask you to, I mean, if you can provide a little more color about the behavior of your customers right now, right? I know you don't always know why or why not your customers are buying from you, and, but any sort of a guess, by your product categories that what, in terms of the inventory level, that your customers are holding, how big they are? What's the status, or how, when are they gonna deplete and then start restocking?
Yeah. I think the short answer is we truly don't know how much they have. And how we know that they have it is because we can see parts that are ordered pretty routinely that are getting ordered at very low volumes, and they basically have told us that they're gonna be burning theirs down before they can replenish from us. At some point, they'll have to cross over because you can't necessarily ramp that quick. If the back half of 2025 or 2024 is going to be the growth half of the year, and we'll get to that, I'm sure, going into a larger 2025, then they're gonna have to cross over at some point. I would say in gas delivery, it's very little, and it's probably insignificant. It's mostly in weldments and our machine component areas.
Is it because the gas panels are like, it's a custom-made, you can't really-
Yes
... put that into inventory?
Yeah, to some degree, and keep in mind, the order horizon as we were going through the tail end of COVID, everybody's order horizon to the supply chain was probably a year... and so was our customers. POs, everything was let out. So as you bring that thing down, then, you know, then that's when it surfaces that they've got excess, right? So there's reasons, because most of us didn't anticipate the drop-off as quickly as it happened.
Got it. So now, onto the forward-looking stuff. One of the things we as analysts want to avoid is you telling me what you read from my writing, and then creating an echo chamber that doesn't help a lot of people. But I'm not going to ask you about your WFE forecast. I want to ask you, how much visibility you actually have right now? How do you characterize your visibility, and how is your 2024 order book telling you right now about 2024?
Just, that's a good way to frame it, 'cause we all have different perspectives. They coalesce to the end of the year, so at some point, we'll all be right. But the... Our visibility, you know, if you think about it, I would say our visibility really was kind of 12 months out as we were kind of going through the pandemic and all of that, period of time. Demand right now has been very stable, as stable as I've seen it. I wish it was higher, but my outlook for the first half of the year is pretty similar to the back half of the year or the fourth quarter. And that's, we've been pretty consistent now for probably 2 quarters on that outlook.
So, but what our visibility is, is most people have, again, pulled in kind of. We don't run our business on PO backlog. I would say in the component business, it's a little more prevalent, but, you know, our lead times are 3 and 4 weeks for gas delivery, with the exception of ASML, which is much longer. So we get great visibility on the ASML side of the business for pretty much a year. I mean, they're very, very transparent. And, but the other ones, I'd say 3 months, it's pretty, it's pretty good. I'd say the next 3 months is directionally accurate and changes a lot. So it's back to where it was pre-pandemic, actually, so.
Okay. So can you kind of clarify a little bit what the one thing you just said, you don't run on PO backlog. What do you mean? What kind of backlog you're running on?
The backlog turns very quickly, so we get forecasts pretty much on a weekly basis out, and like I said, we trust three months of it, and we load that in and run our business on that, and then directionally, we have to kind of predict where it's gonna go after that. So we run the business primarily; two-thirds of the business is our integration business, and that's run mostly on forecast.
Mm-hmm. Got it. Okay. So Ichor has historically outperformed OEM customers, and especially in the first year of the WFE upcycles, right?
Mm-hmm.
I think in 2020, right, for example, WFE revenue of Lam, Applied, I mean, that includes not just systems, but upgrades, everything that counted towards their WFE revenue. I mean, we're up about by mid-20s%. Ichor revenue was up to close to 50%. Yeah, I guess everyone-
46, 47
... here in this room wants something to happen again like that. But
So do I.
Hypothetically, right, if WFE really bounces back, how confident are you this time that you will outperform your customer by as much as that again this time?
Yeah.
And why and why not?
Okay. Well, one, I remain confident because in downturns, we focus on market share gains. And whether you win that business and you start to materialize it within 2023 or it comes in 2024, once you win it, you want it. And so if you go back to as we came out of this in 2019, we talked about a large amount of market share wins, which is what drove the 46%. A big portion of that was our second largest customer outsourced the remaining gas panel assembly in Singapore. That won't reoccur. So the magnitude of it will be less, but when you layer on some of our cost or market share gains, some of those are gonna be top-line gains.
A lot of what we've done this year are actually for internal sourcing, so it's using and leveraging our machining business. Largely, what we've had is internal supply of our weldment business, and now we've extended that a lot more extensively across our machining business. Having said that, because of the inventory burn down and other things, it's moving much, much slower than we had hoped, but it's there to be had. So you're gonna get both revenue outperformance as well as earnings outperformance when you combine those two, and those are the things that we've really focused on this year to do it again. But to say we're gonna go 40-46% versus, I think it was 20 in WFE, I don't think we can do that because the large outsourcing of the gas panels isn't there to be done again.
Got it. Got it. So Jeff, I think, well, top of the mind of a lot of people really is about when the cycle is gonna turn, right? We see signs of bottoming, but when it's gonna actually turn positive, right? That's the top of the mind for a lot of people here.
Yeah.
So what were the kind of leading indicators for a potential upcycle recovery that repeatedly showed up in the past cycles in your observation?
Yeah, uh-
Please don't tell us indicators that we can also see, but the kind of indicators you uniquely can see.
Right.
And what do they look like?
Yeah, I would tell you that there's we look at external and internal, and I'll talk about internal drivers first. The place we generally see it had we not still been kind of working through this inventory consumption of our customers, which I really think will take through mid-year this year, is our component side of the business. Their lead times are 10-13 weeks. We've got to see POs before then, right? Whereas 3 or 4 weeks for gas panels, they'll follow. So that's usually a trigger. We haven't seen that yet, and which is why we kind of see the front half of the year being fairly similar to our back half of 2023. And so, externally, really, I continue to just look at the memory market.
I mean, obviously, dep and etch is highly influenced by the 3D NAND market. I would say most people cannot be wrong in saying that 3D NAND in 2024 is going to grow because it's coming off of a $3 billion-$4 billion level. So you can say it's going to grow 70%, and it's $42 billion. So, we need to kind of see that memory recovery because of the exposure that our largest customer has to memory, where our second customer is pretty balanced between the two.
Mm-hmm. Got it.
Those are the things that we look for.
Got it. I think you might have touched upon this a little bit earlier when you talk about internal sourcing, but, I'm going to ask, from a financial perspective, right? I believe you said one of the oftentimes underappreciated stories of Ichor is about the profit margin. And maybe let me first ask a little bit backward-looking question again. Why is Ichor's margin performance seems to be slightly worse this downturn than the last one? Yeah, and let me throw out some numbers. In the last downturn, in the last reported September quarter, 2023, your gross margin was 13.1%. The lowest in the last cycle, I think at some point in 2019, was actually slightly above that-
Yeah
... 13.5%. But, but how should we think about margin improvement targets from there?
Well, maybe I'll talk a little bit about what's happened since the 2018 trough. Maybe it was Q1 of 2019, I believe. But is our infrastructure, as we were kind of peaking in 2016 or 2018, I would say we were almost over capacity. I mean, we had to add capacity. We were running factories full out, most of them 24/7, to do that. We added a lot of capacity to support the next wave of growth. I would tell you that probably external revenue dollars, our capacity is probably north of $425 million, maybe somewhere around $450 million a quarter, okay? Including the internal supply that we planned in it, it's probably another $50 million or $60 million on top of that.
So when you're thinking about doing $500 million a quarter, you've got a lot of fixed costs, and a lot of what we added was really in the machining business, probably in the neighborhood of $50 million. So you can't stop that depreciation. So that's part of it, is just the fact that the industry has rolled back to maybe not similar levels, but we just can't absorb all of the capacity that we've had in the past. And generally, when you start to come out of these upturns because of the inventory reductions and stuff, you generally see the integration business kind of get going first. The other thing is that everyone in the industry, including Ichor, our customers, our peers, we have the highest inventory levels we've had ever, historically.
I mean, inventory turns for us, you know, we'd be in the five times range, maybe on average, through a cycle, and we're sitting at under three. That comes with some risk, and last quarter, we had to take some higher levels of inventory reserves, and even baked into this quarter's guidance, we said it was still a little elevated from what we would consider normal levels. And then, that's just a function of the amount of inventory you have and the prolonged nature of this kind of down cycle.
Okay, got it. Any thoughts on how should people think about the margin improvements from last year? Maybe that was the trough in the second quarter. How should people think about that?
Yeah, I mean, I think with our strategy around trying to drive more internal sourcing, we try and get between, you know, say, 22-25% growth in gross margin. And obviously, we had a hiccup in Q3. It's a challenging environment when you bounce around the bottom, just, you know, you don't. You've got to see the inflection of our component side of the business to really get up into that 25 basis points. I think, you know, when you extend. Now, I'm going to talk a little more long term, which I think you're going to get to eventually, is when you talk about our next, we call our next generation gas box, this is, it's, this is us.
We've developed enough component and IP-centric parts that can be integrated into a gas box to move the procured piece down from 90 to about 25%. In other words, 75% is Ichor content. And so we're having some pretty good success with the first wave of evaluations. We're qualified on three applications. And if these are now over the next, say, three or four months, they're moving into customer qualification at the device manufacturer. That is the next big hurdle for us. And so when you add that much content to a gas box, you don't need to flip it to 50% of our share or something like that to move the needle.
And to get to that 20%-ish gross margin in our long-term model, we really need to get, you know, $40-$50 million of this stuff flowing through the P&L. And it'll move because the margin on a gas box is gonna move up, you know, multiples of what it is in dollar basis. But on a percent basis, that shift between, you know, buying 90 and buying 25, you're basically putting easily 25% gross margin into the same gas box that we get to keep in the margin stack.
Mm-hmm.
And so it'll move the needle as fast as anything. And then, we don't have to have that big bang kind of approach... We've developed valves and substrates, and even flow controllers. And so anywhere in between, you know, 10 and 75, we can manage that in the legacy product lines, too, as we go and get qualified. Customer success at device level is really gonna be key, because that'll help with the pull-through.
Okay. Okay. So maybe one last question, then we wrap up by the fireside chat part, and then we can open up for Q&A. Jeff, I think the audience is eager to jump in to ask you some tougher questions.
Okay.
I definitely want to ask you about the long term, right?
Yes.
I know the results of the long-term strategy, I mean, probably won't be in your next quarter's numbers.
Mm-hmm.
Maybe not even in next year's numbers. Who knows? But what do you feel the most excited about Ichor's long-term growth, I mean, going to the next 3-5 years? Any of the... I mean, number one, number two opportunities you're seeing right now.
Yeah, what's exciting to me and the team and the company is that we're part of this march to the $1 trillion semiconductor industry. We've done models on what that means to total WFE. I'm sure you've done models. Opinions will vary, but it's nearly doubling what we're seeing this year to get to $1 trillion. No industry is gonna grow that fast. There might be ones that grow that fast for a period of time, but this is a sustained growth industry, and as we continue to grow, there's opportunities out there beyond our four customers. Increasing internal supply, having success with our new gas panel, all of this stuff is gonna drive top line and earnings per share growth above revenue growth, and so that's what we're excited about.
You know, downturns are not much fun, as you guys know. For you or for us in management, they're not much fun, but they're opportunities to drive. You get more time with your customers when they're not chasing parts and ramps and things to get qualification, and I think we've done a pretty good job. We wish more would have come through this year, but it's gonna start coming through next year as well. So those are the things that excite us as a company, and I hope excite you guys as shareholders or potential shareholders.
Great. I think that's a good ending to the fireside chat portion of the session. Maybe let's open up for Q&A, please.
Yeah, hi, Jeff. Good to see you.
Hey.
So, I'll ask the China question: To what extent is the embargoes and that kind of restrictions on China encouraging maybe overspending in China? Because I understand that's happening, and and what extent could that provide a little bit of a boom bust from that end of the globe? And secondly, offsetting that is, how much do you think AI is contributing incrementally to demand for wafer fab?
Two good questions. I think on China, I think when you kind of look at what people thought WFE was gonna be maybe a year and a quarter ago to where it ended up now, the probably the biggest change was the fact that China, the decline of $8 billion-$10 billion after the embargoes, they found a way to spend that money on trailing-edge technology. We don't We're getting indications that it's not gonna see a big, you know, continued growth. We're not getting indications from our customer base that it's gonna see a significant decline. So I don't know where it's gonna end up, but I don't think there's a big drop-off. I think that's a strategy of the country.
If I can't get leading edge, I'm gonna get lagging edge, and I'm gonna dominate there." AI, I think AI is gonna be beneficial for the industry. Most of what we've seen is through our chemical delivery side. So, you know, we've kind of seen it on the advanced packaging initially. I think there's enough wafer capacity out there to handle any of the incremental DRAM or logic that's being built. I mean, you didn't see a big uptick in TSMC for the NVIDIA. You know, the portion of their cost of sales that's related to the actual device is pretty small. And so there's a lot of capacity out there, but this is gonna - I think what, what, what. By the way, maybe I'm a neophyte or something on this, but the more I read about it, the closer it's getting to our handhelds.
To your phones?
To your phones. Phones, edge computing, all this. The further out it gets, the better it's gonna be for the industry. And so those are things that I'm kind of trying to, to understand better because if you, if you do get AI in the handsets, then that's gonna drive more memory, more NAND, you know, faster processors, longer, you know, battery life. All of that stuff is good for the industry. As we kind of continue to the next logic, foundry logic gate-all-around, you know, that's gonna be probably 15% incrementally more capital intensive than before. So we're not really pulling off that 10 or 15% capital intensity on that side of the business. So I think, I think it's good for the industry, obviously. I just don't think it's a needle mover right now, so. But it's, it's, it's, it's incremental, that's for sure.
Okay, any more questions? We still have time.
Lots of time.
Lots of time, yeah.
Maybe I'll just talk a little bit about capital allocation.
Yeah.
I mean, obviously, we're carrying just south of $300 million in debt or so. Our focus really is, as we drive our inventory down, we're basically taking those dollars and paying our debt down. And so, you know, our interest rate's in the low 7s. It's higher than anybody thought it would get to. We do believe next year we're gonna see 2-3 rate drops, and further into the next year. So our focus really is around just getting that debt continued to be paid down quarter over quarter, so.
Okay, maybe I can entertain a few more questions then.
Sure.
Yeah, please.
Do you see any opportunities in M&A? You did a little bit, what, two years ago, but-
Yeah. I mean, in 2017, we did two pretty large ones. We did two years ago, we bought IMG. IMG has done really well, actually. I would say their semi business is down, similar to the rest of ours. Their non-semi business is now more than half their business, and so they focus on aerospace, defense, medical, commercial space. It won't be long before their largest non-semi customer is gonna be SpaceX. And so they do a lot there. So they're helping us take some of our machining business into those non-semi markets as well. So those are kind of upsides to things that we see, too. It's just a much smaller piece of the business than the rest of the semi business, so.
Okay. Okay. So maybe let me ask you, two more questions.
Sure.
One is your business with ASML. I think you kinda touched upon, right? It's on the dollar basis, this is probably bigger compared with your business opportunity, I mean, on per chamber basis, per tool basis, but as a fraction of the value of the tools, it's probably-
Very small.
Very small. But I do wanna ask you about what do you see, that part of the business trending, in 2024? I know that's probably a very long lead time. You probably shipped, what they, they need to ship, your customer need to ship this year. But, what about, looking a little bit ahead, 6, 12 months?
Yeah, I think, I think some of what you've-- you, you can read is, you know, TSMC has slowed their fab in Arizona. That's had an effect, a ripple effect on us. I think we're gonna see a similar year. I don't think we're gonna see a step function up in that side of the business. Having said that, it just depends on when they decide to turn it back on. They build, they have to build the tools. And I'm not gonna speak on their behalf, but you can't lose a slot in an EUV factory. It's too critical. And so, so I, I'm sure that they're continuing the flow of, of manufacturing, and whether they slow it or ramp it, that's it. I mean, for us, DUV offset any of the softness in 2023. We just don't have a position in DUV.
It's a focus area for our sales teams. It's not off the table, but it's, it's kind of a long lead time to kind of penetrate that side of the business. There's been long term... When I say long term, I'm talking, you know, that have been ASML suppliers for 30 years, that, that kind of dominate that side of the business.
Mm-hmm. Okay. Maybe the other question, I mean, that's maybe a little bit more topical, is, is about the silicon carbide.
Mm-hmm.
So maybe a two-part question. First off, how much exposure you have in that part of the market? And two, well, I think folks are probably still quite bullish about the long-term prospect about the silicon carbide, what do you think about the business opportunity for you, a little bit longer term?
Well, I mean, I think with silicon carbide, it's, you know, our first customer, we went from first article build probably in February 2023 to volume production by kind of the third quarter or so. We see that business holding up year-over-year. And, you know, there's several places you can play in, in called silicon carbide. There's the deposition side, and I would say we're kind of-- that's basically it's Aixtron and LPE and to some degree, Veeco. And we have one of the larger players in their positions. It's 100%, it's a single source. It's a complex gas box in comparison to a standard deposition.
Much more like an epi box, it is a lot more complicated. If you looked at it, it's twice the size of a normal standard, say, dep box that goes on a CVD ALD tool. So the ASPs are pretty good on that side of the business. The industry, I think, is going to quickly transition from 6 to 8 inch to scale. I think you've seen some words about softness, but when you look at just EUV growth, it's gonna continue to grow. So we're focused on trying to extend beyond our first customer, and the one area that we don't have really any position in is with any of the implants, both either at Applied and/or at Axcelis.
And so, you know, we've tried to enter engagements with these guys too, but again, you know, it takes a while to kind of get in and scale those types of business. So there's more opportunity there.
Mm-hmm. Mm-hmm. Let me just do a check. Any questions from the audience?
It's a quiet group, second one.
Yeah. It's a lot of people, but they're very quiet. Okay. All right. I guess-
Just one last reminder-
Yes
... before we sign off, is that we've kind of refreshed, I think our IR presentation has some new content. You can see some of the new gas box stuff and stuff. So if you wanna take a look at that on our website, it's great. No new breaking news in there, obviously, but some new, new refreshed, updated album.
All right. Good.
Thank you.
I wanna really thank Jeff, thank Claire, for joining us, and hopefully you guys have a good rest of the conference.
Yes.
Thank you.
Thank you.