Lam Research Corporation (LRCX) Earnings Call Transcript & Summary

All right. Welcome back, everybody. I'm Joe Moore. Very happy to have with us today the CFO of Lam Research, Doug Bettinger. I think Doug is going to read a brief safe harbor, and then we'll go right into Q&A.

Yes, I'm going to read the brief safe harbor. But I would point out, this won't be a discussion of innovation this year, or maybe it will be, but yes. Thank you. Just real quick, I'd like to keep my attorneys happy. Let me read the safe harbor real quick. Today's discussion may include forward-looking statements subject to risks and uncertainties. Actual results may differ materially. Additional information concerning factors that could cause actual results to differ materially from these forward-looking statements can be found in the risk factors disclosed in our public filings with the SEC, including our most recent 10-K and 10-Q. Thank you.

Great. Maybe we could just start out talking about the size of overall WFE. We started last year thinking it was a mid-60s year, at least for me. We ended up obviously quite a bit higher than that. And actually at this conference, one of your competitors was talking about $100 billion WFE. And kind of hypothetically people we're like that's crazy. And it's something that here we are 1 year later, we're...

And here, that's what we see.

Yes. So it's a big number. You've always felt good about your business and you've sort of felt like there is a rising tide with WFE. But that $100 billion number, can you put that in some perspective? Does it surprise you that we're at that level?

I think if I had been talking to you 2 years ago, it would have surprised me. But as I sit here today and seeing what's happening in the industry, it doesn't. And so what do I see going on? I guess the first thing to get your head all wrapped around the investment that's going on right now is, first, the demand for semiconductors is very strong, right? The secular growth drivers for the entire industry is very, very good. And frankly, I think coming through 2 years of COVID has accelerated some of the digital transformations in the economy and society/and that's part of what you're seeing relative to the demand for semis. It will be an all-time record year for revenue in the semiconductor industry. So that's the first thing. There's secular growth drivers that are stronger as I sit here today than I would have thought 2 years ago, Joe, but this has incrementally gotten stronger. So that's one. You've got to think about how big is the industry and what does the industry need. Two, when you layer on top of that the strong semiconductor revenue, you've got rising capital intensity that's going on and somewhat of an acceleration maybe in process node transitions occurring. So that's why on top of the fact that semiconductor revenue is really strong, you put WFE on top -- or the capital intensity growing, and WFE is maybe even a little bit stronger. The good news about Lam Research, frankly, within that is our share of that WFE because of the transition to 3D device architectures that you've been seeing for a while, frankly, but it continues, puts us in a great position within that part of the industry. And then third, I think, and I know you'll ask me about this later, is we've got a wonderfully growing installed base business. We call it the Customer Support Business Group, CSBG, that frankly has some fantastic tailwinds because of high utilizations in the industry. And so when I think about the sustainability of our part of the business, I feel really good, Joe. So that's really what I see going on.

Great. And maybe on the point that you just made in terms of 3D device structures kind of fueling your performance, I mean, you guys have continued to outperform WFE even though if I look at memory spending, which you are overly indexed to a little bit, has been less dynamic maybe than foundry logic. What are the dynamics of your business overall that you're able to outgrow the overall industry even though you're indexed to the stuff that's going a little slower?

Yes, I won't deny, we have a wonderful presence on the memory side of the business. But I think people under-appreciate the growing proportion of our Foundry and Logic business. In fact, the last several quarters have been all-time records for the company in Logic. And I think a couple of quarters ago, it was in Foundry as well. So there's a share story going on there that I've been telling for a while in terms of we've been under-indexed at 1 large Logic customer, and that's growing nicely. So that provides incremental tailwind to what's going on there. On top of the fact that when you think about yes, we're very well indexed on the memory side, and as stacks grow taller, we get even better indexed in terms of -- they spend more money on the things that we uniquely supply to the industry. So that has benefited us. And then on top of the fact that I know you'll ask me more about CSBG. CSBG is just great, honestly, and exceeded our expectations. So when you put it all together, that's kind of what's been going on. It had a great year last year. I think we'll have a great year this year. And yes, foundry and logic is growing maybe a little bit more than memory. But everything else we've got going on is offsetting.

Great. And then you've had these supply chain issues of the last couple of quarters. I know you've guided that, that will continue to be a little bit of a challenge here. Can you just talk a little bit about that and where you are in terms of resolving some of those issues?

Yes. So I'll remind those of you that have followed us what we said, and then I'll talk a little bit more about it. First, as we went through the December quarter, late in the December quarter, we got a little bit of a surprise from 1 supplier that impacted our ability to recognize revenue from shipments that we made in the December quarter. So that was a little bit of what happened in December. When we released earnings, we described what I see in the supply chain is -- the challenges are broadening out, I guess, is what I would describe. Part of it is the growth of the demand for equipment has gotten stronger. But the broad-based supply chain issues have gotten a little bit more challenging, if you will, right? It's not just 1 supplier that we're working with, and it hasn't been for a while, but it's gotten a little bit more challenging. And by that, I mean, we're tactically managing challenges. One supplier has a certain challenge. Another one has a slightly different challenge and so forth across a lot of the complicated supply chain that we have. So what I would tell you is we're working our way through it. It's, I guess, I call it a game of whack-a-mole. You go fix one thing and then you've got to go after the next one and you've got to go after the next one and then something else pops up and you have to go after that. It's a fixable challenge, but it's not easy is how I've been describing it today as I've talked to people. I wouldn't say it's getting a whole lot better. It's not getting worse either. We're managing it. It's difficult and challenging. And if anybody in my supply chain organization is listening, I hope you're not because you've got work to do, darn it. But it's heavy lifting. But it's problems that you know how to fix once you go figure out what the root cause is, and you work your way through one and then you've got to go manage the next one. We're good at managing this stuff, but it's some heavy lifting. The good news is demand is very strong, right? And if demand wasn't strong, that's not a problem you can go fix, but demand has continued to be very, very strong. So we've got to get caught up with what customers want from us.

Yes. I mean I guess I've been surprised in the last 6 months at how severe this has been, not just for you, but for everybody in equipment. Just given the multiplier being so large, I mean you're talking about a relatively small bill of materials for semiconductors, $5, $10 chips keeping you from shipping $10 million tools. I guess if there's an auction, you win, so.

But that's very much the situation we're in, as you know, because we've talked about this, right, which is it's a very complex bill of materials when you look at our equipment. Probably thousands of individual components going into a tool. And a lot of it, if you don't have it, if you don't have that important semiconductor, the tool doesn't function and you can't recognize revenue without it. And so you've got to go work your way through that. You've got to go manage the subsystems, you've got to go manage freight and logistics challenges. It's just a complex set of things going on.

Yes. I mean and I don't know if people really doubt you guys that demand is strong. But I mean, every check I have is the installation of tools is at a really high level.

It is.

And you just can't get the last sort of module to kind of complete the sale. But it seems like there's still very strong underlying demand, and it's really just a matter of getting revenue recognition on what you've really already shipped.

Yes. In fact, I talked about this in December. What you saw, if you pay attention to our balance sheet, not a lot of people actually look at the balance sheet all that carefully but you should, you saw deferred revenue grow. And so what we have going on today, more than in the past, although we've always done a little bit of this, which is we ship a tool that might not be complete and then do the final component assembly in the customer's location. The customer, actually, when we're constrained the way we are, like this because we can start the installation process a little sooner than if we were just shipping a complete tool. But it also means you're a little bit at risk that, that component doesn't show up when it was supposed to, and that's very much what happened in December. And so we're doing more and more of that. One, because it benefits the customer, and we can start installation. And then we do final assembly more often, perhaps, than we used to in the customer's location. And they get the tool quicker than if we waited to ship the completed tool, if that makes sense.

Maybe if you could just talk a little bit about that deferred revenue and what -- it seems like that's going to go bigger this quarter and then what your ability to work that down.

Yes. So if you think about the situation I just described, you're shipping in a complete tool. It needs a final set of components to show up and be installed. And so as soon as that happens, you recognize revenue more quickly once it happens, right? And as long as it's getting better, then you've got revenue that will happen in the future that's already somewhat complete at the end of the reporting period. And that's very much what you saw in the December quarter. But the challenges continue to be there in March. And so you get this stuff from last quarter and then more comes in is kind of what I described when we released earnings, Joe, and you know that. You heard us talk about it.

Okay. Great. So maybe shifting a little bit. Geographical trade tensions and the role of that. I mean you've kind of always emphasize that the most important thing is supply/demand for semiconductors and how much customers need to build. But as you look at these sort of geographic factors driving localized spending in each region, it seems like a good trend for you guys. Like how pervasive do you think that is? And how big a part of this year do you think that is?

It's a piece for sure. I mean you're seeing different new locations announced across all of my customers, right, places in Ohio, in Europe and so forth. But you're right, what matters most of all is the demand for semiconductors, the strong demand for semiconductors. And what I've observed happening is some government dollars flow into the industry. It's redistributing perhaps where those factories otherwise may have been built. But I'm not sure it's meaningfully changing in demand, and that's what matters most of all is the strong demand for semiconductors. Maybe it does impact demand a little bit on the margin if you think through elasticity of demand and whatnot. So there's that piece. And the industry likely ends up being somewhat less efficient because they're running multiple fab locations. And so there's a little bit of incremental equipment. You should think about it that way. But all of these government dollars that people talk about isn't necessarily incremental to WFE. Some of it will be, but a lot of it may not be. What matters is demand for semiconductors.

Got it. And then as you think about those drivers, can you talk about your business in China with China sovereign customers? I think you described it as $10 billion of WFE a couple of years ago and growing each year.

Yes.

How broad is that for you? Is the perception it's just NAND? How broad is that to you guys? And how do you see that growing?

Yes, maybe let me step back just a little bit. So we have strong presence in China. There's lots of fabs in China. I think it's important, and I remind people, not all -- so when you look at our reported revenue by geography every quarter, some of that are global customers with fabs in China, and some of it are local Chinese customers. So it's -- I always like to remind people, make sure you understand that bit. There's relative balance between the 2 buckets. Maybe it's skewed a little bit more to local Chinese customers, but understand when you see that percentage, it's a combination of both. It's a ship to fab location number. But you're absolutely right, you remember everything we say. I said in 2010, the local China WFE was, call it, $10 billion. And then in '21, I said it grew. It's growing in '21. I didn't say how much, but we said it's growing. And then in '22, I see it growing again. So it's pretty strong. It's a broad set of customers, got 1 big NAND customer, got 1 big DRAM customer, a couple of foundries that everybody knows and then a broad set of foundry investment that's occurring probably from a handful of customers most of you haven't heard of before. It's a broad set of kind of lagging edge investment. Think about analog, power devices, IGBTs, stuff like that. And that's a lot of customers investing in a lot of fab locations, Joe, for IoT, power analog kind of capability. We call it specialty nodes. I mean it's not that it's -- well, lagging edge necessarily specialty communicating. There's some uniqueness about the equipment needed for it. Image sensor is another example.

Yes. Specialty is a good word. We've been looking for euphemisms for that. But can you talk a little bit about the tension around potential export controls in China? And I mean, I guess it seems to be fairly unlikely that you would have anything like that at a time when we're already wrestling with shortages. But just how do you see the risk of that moving forward?

Yes, I haven't seen that change for a while, frankly. It hasn't necessarily gotten better, hasn't gotten worse. For a while, it felt like it was getting maybe a little bit worse incrementally bit-by-bit. It just feels steady-state to me right now. There's certain equipment we need license for if it comes from the U.S. and still where it is. We've applied for those licenses, still waiting to hear back in a lot of cases. We have gotten some. But in a lot of cases, we haven't. We're in the process talking that through, but it feels pretty steady state. Not getting better, not getting worse.

Yes. Okay. And then maybe looking at some of the end market spending, foundry logic, the most robust area. And I guess at this point, I would kind of parse that between kind of cutting-edge, 5-nanometer, 3-nanometer types of investments versus what you talked about more specialty nodes. How balanced is your business across those elements? And do you see one being stronger than the other?

Yes. Let me maybe talk through a little bit. Yes, our outlook for WFE this year is $100 billion, call it, and we said mid-80s last year, so up high teens, mid-high teens. Every aspect of that NAND, DRAM, foundry logic growing this year, right, double digits for the most part. But yes, foundry logic is growing more than memory. It's kind of the nature where customers had in the road maps. When you look specifically at foundry and logic and where it is, the majority of the dollars are always at the leading edge nodes, but there's a pretty robust investment occurring at specialty node as well. And in fact, a couple of years ago, we described it as that segment of the business will grow faster than overall WFE, and that's what's turned out to be the case. It's very robust, driven by the secular demand drivers for devices coming off those nodes.

Yes. I mean we've seen your customers guide to something like 100% CapEx growth in those nodes.

Yes. And again, that's our Reliant. You'll hear us talk about our Reliant product line. It used to be just refurbished equipment, although increasingly because demand is so strong, we're selling new equipment there. But it's equipment that's been around for many years, several years. Although we are introducing some new equipment there. For the most part, it's equipment that we've been selling for years.

And is it only Reliant? Or I mean you have traditional -- from your tools business that serves those nodes as well, right?

We do, but it's primarily Reliant is how I would describe it to you.

Okay. And then memory, I think it appears to me to be a fairly balanced spending environment.

Yes, that's how I describe it. Fairly balanced, fairly disciplined, if that's the right way to describe it. But customer base investing for the demand growth, they see relative supply/demand balance is kind of my assessment of what I see going on this year. And that needs incremental equipment.

And DRAM versus NAND, do you see a difference? Because the NAND profitability is quite a bit lower right now and...

Been true for a while.

Been true for a while. There's pretty good cash flow coverage on the spending that we have from what I can see. But do you feel any difference between those 2 markets?

It's a lot of the same customers doing both, frankly. It feels fairly disciplined in both NAND and DRAM to me, to be honest with you.

Okay. And when you look at the other -- if I look at NAND WFE being kind of flat, but I think about more of the spending being technology transitions, layer count increases, you've traditionally captured a higher percentage of revenue, you still grow. I mean you see that as an important dynamic.

It is an important dynamic. In a year when the industry is investing more on conversions than on new wafers, the industry may spend a little bit less but the proportion they spend is more on what Lam Research supplies to the industry. Because if you look at where we're strong in the 3D NAND structure, it's building that alternating film stack. We're very strong positioned in that. It's an etching that etched down the middle of the structure. We pretty much own 100% of that in terms of the application. And it's the metallization, right, the tungsten metallization, we have very, very strong position there as well. So -- and that's what the incremental investment in a conversion year or in a conversion spending environment needs to be is you're adding those critical applications, which is what we provide to the industry.

And I know you commented at one point, I keep -- probably you've -- I know I keep bringing this up, but you talked about a $70 billion WFE in 5 years for NAND. We've been at a level that's a little higher than that run rate, I think, the last year or 2, but it seems like the capital intensity is still growing. Do you think that's right? And do you think that -- are you surprised that, obviously, investing conversion from planar to 3D was expensive, it seems just as expensive to stay 3D and scale the layer accounts.

Yes, capital intensity goes up when you think about CapEx per wafer in the NAND structure. And we gave that, I forget when we first gave that number, 5 years ago, the $70 billion over 5 years to get a certain amount of bit supply growth. It's not what the industry needs today. It's higher, right? And we gave that several years ago to try to make it easier for people to understand. But frankly, it's a stale number.

Yes. Okay. You mentioned the installed base business, and I think you've talked about your numbers, the chamber count growth of 13%, revenue per chamber up over 20%. Can you talk a little bit about how the things that you're doing to increase that revenue per chamber?

Yes, sure. A couple of things. One, I mean, we've got a great spares business. And as process nodes move forward and there's more power inside the chamber, the consumption of spares tends to be a little bit more for a leading-edge chamber than a lagging edge one. So that's just it is what it is. That's part of what happened certainly in an etch tool, that it happens that way. I guess the second thing I would point to, Joe, is we've had sort of a way we're thinking about delivering service today different than it historically has been. And what I mean by that is, historically, service has been an engineer shows up in the customer's fab and does an activity for the customer, all right? Kind of engineer on site, show up, do the maintenance that the tool needs and that's okay business. But frankly, the customer oftentimes could do that themselves. And how we've been trying to change how we deliver service over time is more guaranteeing performance or a result as opposed to doing activity. And in so doing, you stop having a conversation with customer base about cost of labor and a cost-plus activity and more towards, I think I can do X, Y or Z for you based on what I know about our tools capability, so let's talk about the performance that we can deliver and have a conversation about a result as opposed to an activity. So that's a richer conversation.

Yes, okay. And I guess you mentioned Reliant. I mean that's a decent part of that business and probably more cyclical, more of a tools business than the other portions. How do you think about that? And I guess when you made the comment in the past that you expect growth every year, maybe not every quarter from this business, but every year, is there -- can Reliant get to the point where maybe you would make that less confidently?

So far feel good about it. Maybe let me step back again and describe CSBG for those in the room and on the webcast maybe that haven't heard me talk about it before. There's 4 things in CSBG, the acronym, Customer Support Business Group, it's the installed base business from chambers that we've shipped in prior years. And so the first thing I would point out is chambers almost never go away. They just get repurposed and reused in a different application, maybe in a different fab, they get moved around sometimes. And so our tools will be productive for decades. And I think sometimes that's underappreciated. Chamber count grows every year as a result of that, right? And it grew meaningfully last year because WFE was very strong last year, and will grow again this year. That provides incremental opportunity for spares, service upgrades and so forth. So that's why I describe it as a business that I can't envision a year where it wouldn't grow every year because chamber count grows every year. And so again, what's in their spares, service upgrades and then the Reliant product line, which you asked about. Yes, the Reliant product line is part of WFE but it's 1 of 4 components of what's in that business.

And all 4 of those components have been growing a lot?

All 4 have been -- when industry utilization is very high and when WFE is strong, that provides a tailwind for the CSBG business, and that's very much the environment we're in.

And it's probably clear to people, but the reason that there's new tools in Reliant is because nobody has got tools to refurbish, nobody is giving you the old tools.

Historically, years ago, we went to market in the Reliant product line almost entirely with refurbished equipment, meaning we'd buy used equipment back, we'd bring it back to our facility, we'd refurbish it, clean it up, replace some components of it and resell it. Think about a used certified preowned car is sometimes an analogy that I will use similar, good business. Right now, demand is so strong that we can't supply the demand with just the refurbished equipment. In fact, there's very little used equipment available in the market for us to buy back and refurbish. So we're actually selling new equipment. It's not the latest generation, the latest model, it's equipment that we've been selling for years, but it's new equipment nonetheless.

Great. So I'll open up for questions for the audience after this. But maybe you could just talk a little bit about profitability. You've talked about moving the manufacturing footprint to Malaysia, principally, but generally expand. Just how do you think about overall your gross margin and operating margin?

Yes. Certainly, let me spend a couple of minutes on Malaysia. Ramping a new factory in Malaysia that over, call it, the last 12 months because it's a new factory. It's been a little bit of a headwind to gross margin that will become a tailwind because it's got better cost structure. We'll have the supply chain moving along with us. So I expect cost improvement there. And maybe more importantly, most of the industries fabs are in Asia, obviously, right? And we fly stuff, inbound and outbound, both from our suppliers in Asia to our manufacturing location, then back out to the customers. Having a manufacturing location in Asia means we're flying a shorter distance. And with logistics spending as elevated as it is in the world we're in right now, that will be a big benefit as well. So when you put all that together, as we ramp that factory, it will be a benefit to gross margin. And that's part of how I see margin improving as we go through this year and into next year. On top of that, we just guided a gross margin that was down from the previous quarter, and that's got everything to do with the constraints on the supply chain. We're working our way through and kind of like, I call it, whack-a-mole bit by bit, we'll work through that.

But if those supply constraints persist and those costs persist over time, I would think you'd be able to start passing those along?

Some. I mean there's an aspect of the inflationary environment that you've got to go get paid for because it's permanent. And there's some of it that you should be able to go beat the cost down over time, right? And so we think about the permanent inflationary stuff is stuff we should get paid for because otherwise, profitability won't improve. And the stuff where it's not a permanent increase in the cost structure, you've got to go manage the cost structure so that it gets back to where it needs to be. So it's -- I think of it in 2 different buckets.

Great. Let me pause here and see if there's questions from the audience.

Do you have a mic in the room?

Is the mic up here? Yes, I can repeat it.

Just talk loud and we'll repeat it.

Why are you not passing along transitory kind of, inflation? Why not pass along temporary cost increases? Your customers are doing it. They're passing on cost and when their costs come back down, then they'll lower their price. Why are you willing to...

Because it's temporary. Those of you on the webcast couldn't hear. The question was, why don't you pass all the cost increase along to the customer because the industry is trying to do that. We've got a responsibility to manage the cost structure in a way that is appropriate, and we have a responsibility to get paid properly for what we do. And we think through both, but to me, we have to go manage the cost and not just be lazy about it. And the stuff that we can manage, we should go manage it. And the stuff that is permanent, we've got to go get paid for.

Freight spend has been brought up and [indiscernible].

Yes. So again, webcast. Freight spend has been up for a couple of years, why aren't you passing that through? And so maybe some of that we are.

I feel like when the semiconductor companies first started having shortages, they didn't start raising prices. And when it became the new normal, then the prices went up. So maybe that ends up happening to you.

Here's what I'd tell you, we all need to get paid fairly for the value we're delivering in the industry or the investment required to keep delivering the value will reduce itself. And we have every intention of getting paid properly and fairly for what we do, and we will. But I also feel an obligation to make sure the cost structure has got enough tension and that it gets to where it needs to be as well.

Other questions from the audience?

That's a good question, I've been getting that one all day today.

Maybe I could ask you something. I mean I just thought this was interesting, an overall commentary on semiconductors. When I asked you a few weeks ago what you're going to do differently when we get past this, are you going to hold semiconductor inventory, are you going to have your suppliers hold it? And how does that compare with how you've behaved in the past?

Yes. No, it's the right question and one I'm not ready to give you the answer to yet. Because frankly, Joe, what we're doing right now is just trying to fix the tactical challenges that we have, and that's what we have to do, right? I've got to get caught up with what customers want and when they want to get lead times back in line and so forth. And once we get to the light at the end of the tunnel, I think we'll look back and say, okay, that didn't feel great coming through that. What are we going to do differently so when it happens again, if it happens again, it's not as painful as what we just came through. Some of it, I'm sure will involve carrying a little bit more inventory, perhaps. I think I see a lot of that happening right now. Some of it might involve having redundant supply chain and different location. So if you're having a challenge in one country, there's an opportunity to get it somewhere else. I think you'll see more and more of that from us. We might do a little bit more ourselves and a little bit less outsourced. I think all of those things will be debates and discussions we have. But right now, it's just too soon to make the decisions because we just got to -- we've got to go play whack-a-mole.

You've got to weather the storm first.

We've got to weather the storm first and then step back and learn from what happened.

Okay. I guess we've got 15 seconds left. So I guess we'll just wrap it up there. Doug, thank you very much.

Yes, thanks for having us, Joe. Appreciate everybody coming today.