Lam Research Corporation (LRCX) Earnings Call Transcript & Summary

Okay. Hello. I'm Tim Arcuri. I'm the semiconductor analyst here at UBS.

Listen, I have no idea who [ Forest Nara ] is, but that's not me.

That is not Doug Bettinger. This is Doug Bettinger. He's the CFO of Lam Research. Forest is with AMD. So we...

AMD. So we compete with anyway. I'm not Forest. I'm Doug.

We got that a little confused. So anyway, I think you have a...

I do. I have a quick safe harbor that I don't know if we have a slide, I hope we do put it up real quick. Anyway, I encourage you to go to our website, have a look at it. I don't plan to say anything new. I don't plan to change any forward-looking statements. But in case Tim asks some crazy things, these cautionary statements kind of gives me a little layer of cover for that. So have a look at that. And Tim, let's just jump into it.

Great. Okay, Doug. So let's talk about the 200 and some pages that got dropped yesterday, the export controls. You already said, "Hey, listen, it doesn't impact our current quarter guidance or the guidance."

Yep.

So I don't know if there's anything you wanted to say about that.

No, nothing really new. It was largely what came out was what we expected for the most part, right? And so we actually issued a press release yesterday before we showed up here because I knew I was going to be talking to people. I want to make sure we had that air cover. Guidance is good. No impact to anything we put out in the earnings or anything like that. It was largely expected, what showed up in the 200-odd page document yesterday -- 2 documents, I should say, but 200 total pages.

And part of what was unique in this one was, yes, there were chip makers added to the entity list, but there were a lot of the equipment companies added to the entity list.

There were. Yes, that was new.

That was new. And I think you and others have been saying, look, the playing field is not level because the incremental growth is in China. The government is enticing them to buy from domestic sources, even though it's inferior product, and so they've been gaining share.

So listen, here's the thing, right? There's prior customers that we can't sell to any longer, right. They got put on the end user list. So those companies haven't gone out of business necessarily, but they're doing the best they can with what they can get, that's where those Chinese equipment companies are showing up more than any place else, which is you can't get equipment from Lam Research anymore because the government restricted it. So get what you can get and make do what it the best you can. That's the nature of that.

But it's good. I mean you would say it's competitively -- you have been saying that yes, in China, the domestic equipment companies have been gaining some share. So the fact that some of them got to add to the list is net-net good for you.

It's going to make it harder for them, yes. To continue to do what they're doing, right? They can no longer source from a supply chain outside of China.

Doug, so can you talk about -- I'm not asking you for a '25 WFE number, but I'm asking you for some puts and takes.

Kind of how we think about it, yes. So what we did also say on our earnings call and still absolutely what I'm saying today is we expect 2025 to be a growth year for WFE and I'll unpack that a little bit. I'm not giving a number yet because it's too soon for us to do that. We'll do that after the December call or on the December call. But from where we are this year, and maybe let me unpack this year briefly, right? We think WFE in 2024 is roughly $95 billion, call it mid-90s. That's how we've been describing it. And so as we look across kind of what happened this year and what's happening next year, we see some level of growth, and we'll quantify that soon. But if I unpack it, though, I'll give you a little bit of color. We're coming up 2 consecutive years of very little investment in the NAND segment of the business. I believe next year is going to be largely characterized by NAND investments in conversions, meaning upgrading the installed base that's already in the fabs, and it's got to be more than it's been this year and the year before because it's been, frankly, I think, of it below maintenance level. So that's one thing to think about. Second, I believe leading-edge foundry and logic is going to continue to be pretty strong, pretty good, right? And I think if you listen to some of my customers talk, they're describing that. So they get all around node looks pretty compelling. By the way, I would also point out relative to that, that's incremental addressable market for us. And I should mention, when you think about NAND upgrade, we get a very good share of wallet when the upgrades happen because we are the constraint tools in a NAND fab. So that's important to think about. So living as foundry and logic gate all around node, we should have good performance there as well. DRAM will continue to, I think, be strong, right, largely characterized by investment to enable DDR5 and more importantly, high bandwidth memory. Those are bigger buy-in and high bandwidth memory, you need more equipment to do what I call the drill and fill. That's the TSV that we enable. So you've got that there. Trailing edge foundry, logic, probably is a soft year next year characterized by likely a softening investment in China, all right? So you've got a lot of spending there in addition to the fact that the global industrial, automotive, analog guys are still moving their way through an inventory cycle, so they're not spending a ton on equipment right now. And so those are the puts and takes to think about going into next year. When we pack it all together, it looks like a growth year in WFE for us -- or for the industry, frankly.

And one of your competitors or peers, we'll say, said, "Hey, the NAND WFE now granted off of a low base could go up 2x next year off of a low base." And I wrote a big piece on NAND upgrades, and I concluded that it's $40 billion over the next 3 years. If you look at how much stranded old capacity there is. So you can get some pretty decently sized numbers off of where we are today and the NAND WFE market, would you generically agree with that?

I don't know if 2x is the right number, but growth. Could it be 2x? It could be. Could it be less, could it be more? Maybe. But what I know is going to happen next year is largely going to be characterized by spending upgrades. And to your point, our share of that spend is really good when that happens. And for the simple reason that we own pretty much all of that mold stack deposition, that long, etched down through what's called the channel hole. We pretty much own all of that, do own all of that. And then the metallization tungsten, we pretty much own all of that. So when you think about having to upgrade the installed base, that's largely what needs to be upgraded. In addition to the fact that we see the beginning implementation next year of molybdenum beginning to replace tungsten and that also will be good for us. So when you put all that together, I feel pretty good about our ability to outperform certainly in NAND, for sure.

And I did want to ask you about molybdenum because all of the installed base needs to upgrade.

Eventually. Now, that's not all going to happen next year.

Yes. But it seems like it's going to start being middle of next year and maybe SSD first and maybe consumer later.

I think that's right.

And you're the incumbent. So you're going to get that business.

We're well positioned for this. We feel really good about it. So yes, this is going to be a good transition, and it's going to layer on top of just the normal upgrades because it's a new piece of equipment to deposit moly instead of comptungstens.

I know you love to talk about CSBG, so I wanted to make sure I got to talk -- I asked...

Thank you for asking earlier. It's usually the very last question.

That's right. I get to it early. So maybe you can talk about CSBG. You've done a really great job. Yes, the Reliant component has been a bit of a drag this year because it's more of a wafer fab equipment.

But it's been okay.

Business.

Yes, let me unpack CSBG. So for us, that's the customer support business group is the acronym if you haven't followed the company for a while. And there's 4 components that go into that. Spare parts, all of our equipment needs spares to be replaced with some frequency. Service, right? All the equipment needs to be serviced. It needs to be taken down, cleaned and so forth. The equipment can be upgraded frequently, both technology upgrades as well as productivity upgrades. That's part of CSBG and then what we call the Reliant product line, which is the mature tools. Sometimes it's stuff that we refresh, but increasingly, it's just older new equipment. So when you pull that together, actually, the last several quarters, that's been north of 40% of the company's revenue. Now the thing to understand is our equipment runs for decades. You'll sell it, install it, it will need to be serviced. Spares will need to be changed. At some point, it can be upgraded to the next tool capability. Sometimes the customer will then move it to a different application, want to move it around in the fab. Sometimes, we'll assist with that. That's part of the service profile. And then it used to be the case, although it's not been for quite a long time, that sometimes we would buy used equipment back, refurbish it and resell it. When you think about that cycle, on average, equipment that we sell generates more revenue over the life of the tool than when we actually sell the tool itself. This industry and investors are always so focused on what's WFE, how big is the WFE. But frankly, this is the generation of profit after WFE occurs, except in Reliant, in the case of Reliant. So anyway -- so we give a number at the end of every calendar year, which is the of chambers in the installed base. It grows every single year. So the opportunity to continue to generate growth in business from CSBG continues to grow at least partly or maybe largely because chambers grow every single year. So the ability to sell more spares, more service and upgrade things gets bigger every year.

And any color you can provide or remind us of what you said about the relative size of each of the pieces within CSGB?

Yes. What the company said is the biggest individual component is spare parts. And Etch tool actually is pretty spare parts intensive. So spares is a big piece. And then more recently, I've been describing Reliant as now the second biggest piece, at least this year, it is all above the other two.

Got it. Can you talk about margins, Doug, and just how to think about the puts and takes on gross margin?

Sure. Yes. Let me see where to start. We've had some, what I would describe, favorable customer mix that has benefited margin over the last 3 or 4 quarters, call it. Last quarter, the reported number was north of 48%. We guided the December quarter to 47%. We've been benefiting from a favorable customer mix, read that smaller customers as a greater percentage of the total revenue of the company. As we move into next year, I think that dynamic likely mitigates a little bit in '25 and so it becomes a little bit of a margin headwind as we go into next year. Offsetting that, though, is, we've actually pivoted the company's manufacturing footprint to have a greater Asia manufacturing profile. And so as business grows, if we're right about WFE growing next year, and I think we are, and we're outperforming next year, the incremental volume will show up manufacturing-wise in that factory -- those factories in Asia. It will benefit from a better cost structure. The supply chain has got better cost and freight and logistics because most of the fabs in the world are in Asia because you're applying these large pieces of capital equipment a shorter distance. You save dollars on frame logistics spending as well. So that will show up as a tailwind as we go into next year, assuming we're right about the growth next year.

So do the Malaysia tailwinds, which Malaysia now is net neutral to gross margin, becomes a tailwind next year. Do the Malaysia tailwinds offset the headwinds from China and mix?

I'm not ready to put numbers on it yet. Clearly, I'll do that when we guide the next quarter. But it will depend on the relative -- how much goes each direction, right? So I'm not ready to guide you there quite yet. But the 2 will offset. I want to make sure people are thinking about that aspect.

Great. I wanted to ask you about deferred revenue, which I -- you probably didn't think you were going to get asked about that.

I didn't know you're going to ask that one. But that's okay. Go ahead.

So one thing that was interesting last quarter was that deferred was up.

It was.

And because of down payments from what seems like new customers in China. So my question is really about like the pace of the new customers in China because this is not a new phenomenon that these new customers come to you and you say, well, I don't know who you are. So I need to take a bigger down payment, and that's what's been driving a lot of your deferred. Now it seems to me when I was looking through your results that deferred went up, but that it's a larger percentage on a smaller number of customers. So like, is it fair to say that the pace of the new -- the new customers in China is slowing?

Slowing... I don't really think it is necessarily. The level of spending might be though, frankly, right? As we think through -- okay, I think China WFE is down next year. I don't -- not too many customers are no longer there. They're all still there. They're just probably spending a little bit less, which would also drive the same dynamic you're asking about.

But it's not -- yes, I guess my question is like, there's been this tailwind of new entity -- let's say, 15 new entities per year, spending a couple of hundred million dollars each -- small number, but a big incremental number of...

There's a lot of customers in China for sure.

Is the number of incremental customers that are coming out of the woodwork every year, is that slowing?

Yes, maybe a little bit, yes.

Do you think that that's -- and where I'm going with this is, do you think that there's a process of consolidation in China that's begun, that either by government forces or by natural forces that we saw this expansion in the customer base and now it's beginning to contract where the government or, like I said, natural forces of economics drive some consolidation of this explosion of these?

No, I don't think that's happening. To my knowledge it's not, anyway.

It's not happening yet? Okay.

Not to my knowledge.

Okay. And then I wanted to ask you just about sort of the competitive dynamic. One of your competitors talks about Cryo a lot.

They've been talking about it for years.

For years, yes. And can you just talk about sort of your position there?

Extremely good, right? We recently announced -- we call it Cryo 3.0, incremental performance beyond Cryo 2.0, which was incremental beyond Cryo 1.0. We are the only company in the industry that has Cryo in production, right, almost 1,500 chambers today. And yes, 1 of my competitors, peers, whatever you want to call them, has been talking about, "Hey, we're going to win business and so forth and so on." And so far, they have not. And they keep talking about winning business. And when they talk about it showing up, it continues to be later and later and later. So -- we're in a great position, the best ROI for the industry and for our customers is to upgrade what's already there. And that is Lam Research. So we're good at what we do. We're defending things extraordinarily well, and we intend to continue to do that.

I want to ask you also something from last quarter. So you guided December, you guided China down $250 million...

I guided it to approximately 30% of revenue, yes. So doing what math you're doing, yes.

So it's down, yet you're guiding revenue up $150 million. So something else is making up the difference. Something else is an incremental $400 million. That's roughly the math. So it's -- and then you talked about Taiwan being a little stronger. I know that advanced packaging has been stronger recently. Is that sort of like a harbinger of what you're talking about, that leading-edge foundry logic is going to really be strong next year, and you're seeing it maybe like under the surface you're seeing it already in the December?

Listen, really, if you step back, the global part of the industry outside of China is growing, offsetting China going down. I mean at a simplistic level, that's what showing up. Yes, leading-edge foundry continues to be strong. DRAM continues to be strong. That's a large part of what you have going on to.

Can you talk about just fan-out panel-level packaging, which is going to ultimately replace CoAs? How are you -- you acquired SEMSYSCO, I think?

SEMSYSCO is the name of the company we acquired, I think, 1.5 years ago, yes.

And that gets you into that.

Yes, it does. It helps accelerate into that. Listen, almost everybody in the industry is looking at whether it makes sense at some juncture to begin to package devices in a panel form factor. It can be bigger. Theoretically, it can be done at a lower cost. And so the industry is working on this. When we acquired SEMSYSCO, for us, it played to the fact that, one, they already had a tool out there. It also leveraged the strength of something we're really good at doing at Lam, which is electroplating. So bringing those 2 things together, we looked at it and we saw technical synergies, which is almost always why we do kind of an acquisition. And then today, the whole industry, like I said, is trying to figure out what is this going to look like? What is the form factor going to be? Each customer is kind of thinking about it a little bit differently today. But I think eventually, this will show up as business in the future. It's not there yet.

Yes. I mean TSM is talking about panel level packaging '27, '28.

Yes. So you got to invest today to figure that out because this is a fundamentally new way of doing things.

Doug, can you talk just a bit about DRAM? There was -- I think your view is a little different than maybe what -- ASML presented a slide that showed that 3D DRAM may never get to the cost scaling of some other things without new material. So how do you think about some of these technology vectors in DRAM that if DRAM begins to go more vertical that really works in your favor?

Well, listen, when everybody in the industry is working on a problem and working on a similar thing like 3D DRAM, I would tell you, everybody is working on this. It always happens at some point. It's a question of when it happens. In the meantime, you're going to see something called 6F squared go into 4F square, the intensity of the etch in that goes up. So that will be a good thing. But that's not really a 3D structure. When you think about things going 3D, it tends to be a lot more etch and deposition intensive because you're depositing material, you're removing material. We really benefited when NAND did that, went to 3D NAND. 3D DRAM is going to do that. It's a question of at what point. To us right now, it seems like late in the decade, maybe 2030, but we can debate what year I'm highly confident is going to happen.

And then just in the vein of DRAM, I wanted to ask you about HBM. And you said that advanced packaging is over $1 billion in revenue today and HBM is a portion of that?

Yes. It's a portion. We said HBM this year for us actually is tripling, 3x from last year.

So can you talk about your -- I think of your position in the advanced packaging part of HBM as being #1. There's the front-end part of HBM where maybe your share is not quite as high as your peers because they just have higher share in DRAM generally than you do. But in the HBM, in the advanced packaging part of the HBM, you have the top share, is that correct?

I believe we do, right. Through silicon via, which are the most enabled applications, we pretty much own all of it, right? I call it the drill and fill, the silicon etching. It's a tool we call Syndion, extremely strong position. And then the conductive material, you need to put a copper metallization down. It's done with an electroplating process, which is the tool we call SABR 3D. Those are the most enabling applications in HBM in my opinion. It's through silicon via, and we pretty much have the critical steps in there. Same as colos, by the way, right? And so when you think about all of this together, the TSV is some of the most enabling applications that show up in both of those spaces.

I want to go back. I'm just getting a question here from the audience that I want to try to get to the bottom of. So the China commentary, you sort of made some generic commentary about China next year, just the mix. Does that take into account? I mean you had a view on export control...

And it's largely what showed up yesterday and so it's consistent. Okay.

And then can you talk about NAND? Are you seeing -- there have been some other companies that are saying, "Hey, NAND actually is showing signs of really -- like it's game on now in NAND spending."

I don't know if I'd call it game on. I mean, I observed our customers still being cautious, prudent, use whatever word you want, right? It's not back to peak investment levels by any means next year. But it's been such a low level of investment now for 2 years that we look at it, you'll be through the inventory cycle midyear -- early midyear, and the demand is out there, profitability has improved in such a fashion that we think we're going to move into the phase of upgrading what's in the installed base.

And how do you think, Doug, I want to ask you about the balance sheet and capital return. You have some debt coming due next year. I would think you'll just pay that off with cash rather than refi it, but...

I'm waiting to see, frankly. Yes, we certainly have the flexibility to do that. There's a $500 million maturity April, May of next year. So yes, I'm trying to figure out what to do with that. There's a maturity of the year after that as well. So it's possible we'll just pay it off and wait and see what happens with interest rates. It's good to have the option to do that. And frankly, we may exercise that option that way.

And then in terms of capital return, how are you thinking about maybe how that could evolve over time?

Listen, I don't think it's going to change much from what it's always -- well, it's been since 2017-ish, which is we plan at the company to return 75% to 100% of free cash flow. The Board recently authorized a $10 billion buyback, which was above the previous authorization, which was $5 billion. We'd be pretty much all the way that. We intend to grow the dividend on an annual basis. I know that's what shareholders want to see happen, and that's what we plan to do. And when you put all that together, we end up doing a reasonable amount of cash allocated to buying the stock back. So -- and last quarter, we bought back $1 billion of stock.

One thing that has struck me is how much WFE share ASML has gained over the past few years, maybe gain to...

Finally showed up.

UVs here. So there's been a lot of litho capacity added, but there seems to be a mismatch. If the industry requires that much litho capacity then it doesn't seem sustainable that we're buying this little etch and depo.

Well, here's the thing to understand, and you know this, but I'll remind everybody in the room. You can't buy one thing without the other, right? If you're going to put a pattern down on the wafer, you have to deposit the material to create the features and then you need to remove the material to shape things. That's etch and deposition. Now always, you buy the litho tool first because lead times are a good deal longer than they are with what we provide to the industry, but it all has to go together. And you buy litho first. Etch and deposition follows later on because our lead times are much shorter.

So you would say that -- I mean--

This will ebb and flow year by year.

Two years, yes.

It just ebbs and flows year back. And it also depends kind of which end market is buying, right? If it's a foundry year, that's more litho intensive. If it's NAND year, it's not. And so that mix thing, I talked about the fact that NAND has been depressed for 2 years. That's also skewed a little bit of that from what it may be in the future.

And Doug, when you look at like where your gross margin could go to, I know that one of your peers have been talking about increasing price. But at the same time now, there's probably going to be some used tools in the market because Intel is going to be selling some used tools...

We'll see.

And Microchip might be selling used tools as well, who knows. But how do you think about like -- why can't you push gross margins higher? Like what's the customer discussion? Do the customers push back so hard already...

Listen, the fact that we just printed a 48.3% last quarter, that's pretty good, right? That's the committed financial model of the company from back in 2020. You're always trying to make sure you're getting paid fairly for everything you do, right, do the best you can with profitability cost structure. It's why we kind of pivoted a little bit to the Malaysia footprint. Price is always a negotiation, right? It happens every single year with every single customer, right? You're always trying to do your best there. And listen, I feel pretty good about where profitability is right now at 47% in the current quarter, and it was north of [ 48% ] last quarter. So -- and now we're going to do an Investor Day in February, and come see us, and we'll talk about what that new financial model might be, but I'm not ready to give you something new right now.

I know you, Doug, you -- the last Analyst Day you had, which I think was before COVID.

I think it was like a week before COVID. It was a very -- yes, strange time.

And you gave us some metric that people have always tried to hold you to, but you've since said that doesn't hold any more about an incremental. So for a given amount of bit growth that requires X amount of spending for a given amount of bit growth that.

Even if the information is not current...

It's old. But I mean, are you going to -- at the Analyst Day, is the thought to sort of like refresh all that because...

I don't know. I don't know that we'll do that or not. We're still debating what we're going to do. That was good data for a while and then it wasn't. And so when it was good data, it was helpful to everybody. And then when it wasn't, I spent every meeting talking about why maybe it's not the right information to be thinking about anymore. So I saw a benefit for a while, but then it was actually a negative. So I don't know if we're going to go back and kind of give you a new set of information that's similar to that or not.

And maybe last question, Doug. You meet with investors all the time. Is there something about Lam's story that you feel like people don't understand. That is sort of miss...

Listen, I think there's a couple of things that I always feel like people under-appreciate it, and thank you for asking about CSBG because often in a meeting, will be not asked and I remember at the very end to say, "Hey, don't forget about 40% of the business you haven't even asked me about. So I think the quality of that business is under-appreciated. I think people also historically have perceived that, hey, we're the memory equipment company. And frankly, if you look at the success we have had in foundry and logic over the last several quarters, frankly, we're doing great in foundry and logic, and we'll continue to. So I think people have long memories in this business, in this industry, and sometimes you have to come out to events like this and keep reminding, hey, things aren't the same as they always have been.

Absolutely. Okay. Doug, thank you.

Thanks for having me, Tim.