LRCX - Lam Research Corporation (LRCX) Presents at Morgan Stanley Technology Media and Telecom Conference (Transcript)
2023-03-07 19:44:04 ET
Lam Research Corporation (LRCX)
Morgan Stanley Technology, Media and Telecom Conference Call
March 7, 2023 01:25 PM ET
Company Participants
Doug Bettinger - Executive Vice President, Chief Financial Officer
Conference Call Participants
Joe Moore - Morgan Stanley
Presentation
Joe Moore
I think we're on. All right. Hi, everybody, welcome back. I'm Joe Moore from the Morgan Stanley semiconductor analyst. Very happy to have with us today the CFO of Lam Research, Doug Bettinger. I think -- I'm supposed to read this first for important disclosures, please see the Morgan Stanley disclosure website at morganstanley.com/research disclosures. If you have questions, please reach out to your Morgan Stanley sales rep. And then, Doug, I think you have a safe harbor as well.
Doug Bettinger
And I should remind you that I may make some forward-looking statements, please take a look at our disclosure here up on the screen and have a look at it on our website. And with that, Joe, why don't we jump into it?
Joe Moore
Sure. That would be great. So Doug, maybe you could just talk about the general environment. I think you guys have talked about a mid $70 billion WFE year. That was your view back in October, which I think you were the.
Doug Bettinger
Still the same view, Joe. Still the same view.
Joe Moore
Yes, it still seems like -- and others have sort of come around to that kind of general view. So can you just talk about what the environment looks like? Obviously, your exposure to memory is probably a little different than others, but can you just talk generally to the environment?
Doug Bettinger
Yes, I mean we finished last year with wafer fab equipment spending in the mid-90s. And as we look into the current year, we think it's kind of mid-70s. Within that, we're in the middle of what I would describe as a classic memory cycle. And the investment in memory is meaningfully down this year, call it, 40-plus percent -- and within that, NAND down even more. That happens to be the strong spot for Lam Research. And so we're dealing with that. Foundry-logic is holding in a little bit better, Joe.
And so overall, when we look at it, it's down 20-ish percent to mid-70s, and that is unchanged. And I think it's largely what you're hearing from everybody in the industry.
Joe Moore
Yes. Okay. Well, there's a lot to unpack there. I guess if we could talk a little bit to the supply chain challenges that you guys have been seeing, a big theme for the last couple of quarters has cost you revenue. Now you're sort of seeing some of the benefits of drawing down that deferred.
Doug Bettinger
Yes, we're going to caught back up. Supply Chain is getting caught back up.
Joe Moore
Yes. I mean, I guess you talked about June being kind of more of an equilibrium quarter relative to that, at least, can you talk to that view?
Doug Bettinger
Yes. I mean when you look at what's happened in the industry and with us specifically over the last 12 months, call it, there have been meaningfully constrained supply chain situation we were dealing with. And basically, how that showed up for us is we were shipping equipment that wasn't complete at times, and then we would complete the tool in the customer's fab actually, refer to that as backorder shipments.
And so for the last 12 months, you've been seeing our deferred revenue balance grow as a result of that, such that we finished the December quarter at roughly $5.3 billion in revenue with a decent-sized deferred revenue balance. That is coming down as we go through the March quarter. And as we exit the March quarter, Joe, I think we will be largely caught up whatever normal is from a supply chain standpoint. And so that deferred revenue balance will kind of be at a normalized level exiting March.
Joe Moore
Okay. And having been through this now, how do you think differently about the supply chain? I mean is it -- are you going to go back and hold more inventory? How are you going to -- what are you going to do to ensure that you don't run into this kind of thing again?
Doug Bettinger
Yes. No, it's a great question, Joe. One of the things we needed to do and we're able to do as we went through this environment, I guess -- we second sourced a lot of stuff. We qualified other sources of supply simply because we needed to. And the customer was motivated like qualify that along with us. So -- to a certain extent, that will benefit us as we go forward, in that we will have a more resilient supply chain because of what we've needed to do, Joe, as we came through that.
I think also, and we're having these debates internally at Lam Research is what does it mean relative to the inventory strategy of the company as well? I think it's likely you'll see us hold a little bit more inventory in the future than historically, we have held so that in the event that we get back into a situation like we just came through, where we've got more resiliency. So you'll see a little bit of both of those things from us, Joe.
Joe Moore
Okay. Great. Maybe we could shift to talk about China and the export controls and the impacts there. You had initially talked about a $2 billion to $2.5 billion impact this year to revenue. But it seems like the stuff that's remaining has been pretty robust within China in terms of foundry-logic spending isn't encompassed by those controls. Can you talk generally to how that's been going and what you see out of China?
Doug Bettinger
Yes. Let me step that maybe and describe because -- well, it's a big room today. I didn't realize there was so much interest in Lam Research. It's really nice to see it, by the way. But let me step back and describe for you if you're not familiar with my company, what's happened?
And so when you look at the focus from the Department of Commerce in the United States, they looked at restricting the capability of the industry in China to invest in certain technology nodes. And that's impacted our business. We have a very robust business in China, less robust today than it was before these restrictions came down. But what the Department of Commerce does draw technology lines in the sand in NAND flash devices that were 128 layers or higher, now require a license for those.
In DRAM, similarly, 18 and below need a license and in foundry at 16-nanometer and below. And so that had a fairly immediate impact on our business. Those regulations came out on October 7, and you have it exactly right, Joe. When we look into what we think would have been here this year and what we now see this year, we lost $2.5 billion of revenue from equipment sales, spare parts, service upgrades and so forth and really showed up impacting three customers for us.
So went away, we're potentially working on licenses that may bring some of it back, but I'm not sure that happens, Joe, but we'll see. So that was a headwind for our business, obviously, and we're dealing with that. Having said that, though, the investment in China still remains pretty robust. And I see it showing up more in trailing edge foundry for the most part, right, lagging specialty node, devices that go into power devices, RF, image sensors and things like that.
Joe Moore
Okay. Great. And then the other part of your business in China is the multinational fabs in China who obviously are -- I mean obviously, if you're here for Twitter. But to the extent that we've all been focused on this, there's a situation where the multinational fabs need a license to get equipment for their fabs in China. They immediately got a license, and they're getting assurance that, that license will be renewed at least one more time. But there's still an issue of in two years, I need somebody to approve that I'm going to invest in my facility. Do you see that changing the geographic mix of your business at all?
Doug Bettinger
It might, Joe. I think everybody that has a big footprint in China is trying to think about, what is the future going to look like? Specifically, you're right. So relative to those technology lines that I just described, initially, we all thought, hey, that applied to the multinational customers as well as the Chinese customers. What ended up happening is pretty quickly licenses were granted to continue to ship to the multinational customers pretty much what you just said, Joe.
I think everybody though, will look at, okay, am I going to keep growing my business in China or will I redirect the growth maybe to go somewhere else? And I think that's likely what you're going to see from our customer base. That growth that might have been contemplated in a Chinese fab will show up in Korea or in Taiwan or somewhere else, I think, is what's going to happen.
And so what ends up happening to that equipment that is sitting in China, I think, remains to be seen. I think it's probably unlikely that it will leave China, but it will likely need to get replicated somewhere else. And so to a certain extent, even though we're in the middle of a pretty big memory down cycle, that might be upside at some point as and when growth comes back and the inventory gets depleted in memory, that equipment likely needs to get replicated somewhere else at some point, Joe.
Joe Moore
Yes. And I think that having been a subsidized environment, I think it's very difficult to put the tools out.
Doug Bettinger
I think that's right. I think it's unlikely that equipment shows up somewhere.
Joe Moore
Okay. So maybe we could talk a little bit more about memory. Micron has described this as the worst supply demand imbalance in over a decade. Obviously, the margins of your customers continue to erode. But do we still get to a subsistence level spending where this is the amount that we need to spend to remain competitive? And do we get there in the next couple of quarters?
Doug Bettinger
I think that's largely what the year is going to look like, right? I think investment in wafer fab equipment on the memory side is going to be at a really low level this year. And classically, what happens, we're in the middle of a memory cycle, inventory built up for one reason or another. Maybe if I reflect back on the last 12 months, part of it was -- okay, the supply chain was constrained, and everybody was trying to get whatever parts they could get and memory was available to a certain extent.
At the same time, that demand actually softened in smartphone units and PCs. I mean, things just got softer during the year. And so what ended up happening is inventory built up. Classically, when we go through these memory or inventory cycles is that's what happens, and then the inventory needs to get depleted. Pricing adjusts such that supply comes down, and that is exactly what's happening this year. And so there's a fairly low level of investment in equipment this year.
As the inventory gets consumed, eventually pricing will stabilize from the consumption of inventory and profitability will improve and -- what I can tell you having been in the industry for quite a long time is I never know exactly when growth is going to come back, but I know it's going to come back because it always does. The demand for semiconductors is very robust independent of the fact that there's a little bit of a cycle here. And so really, what we're doing with Lam is getting ready for when growth does come back, we're going to have the company positioned to be a better, stronger, more profitable company than when we went into the cycle.
Joe Moore
Okay. And then if you talk about NAND, which you said is down more than memory overall -- can you talk about the benefit that Lam sees a little bit and we've noticed historically in an environment where you're spending just on technology that Lam tends to outperform the overall spending level because you are spending on vertical scaling where Lam's got higher portion of the tools? Can you talk to that effect? And how are you thinking about that?
Doug Bettinger
Yes. Again, for those of you that are new to the Lam Research story, we are the equipment company that is critically enabling to NAND flash, right? In NAND flash, the architecture inflected in a 3D structure, I don't know, 2014, 2015, from a planar, call it, a 2D to a 3D structure. And the way the architecture pivoted or changed is the storage array got built up like a skyscraper with floors in the skyscraper.
We are the company from an equipment standpoint that deposits nearly all of that stack. We are also the company that does the etch that is critically enabling down through that stack, and then we deposit the conductive material that makes the whole thing work. And so when you look at that, we do extraordinarily well when the NAND segment is investing. In a year when things cycle down a little bit, what happens is the industry will add less raw capacity in terms of wafers, but they will convert the equipment that's already in place.
And so I think what you're intimating, Joe, is when that happens, the industry will spend less money. But because they're converting kind of the constrained tools or the enabling tools which are largely ours, the percent of spend on what we supply to the industry goes up. They spend less, but they spend more on etch and deposition at Lam Research. And so that's very much what you're going to see this year. Investment will be down a good deal, but the proportion of it, what we supply to the industry will actually go up.
Joe Moore
Okay. And then conditions in NAND are clearly a little bit worse than DRAM. I'm still surprised spending is that much different than you -- at this point in the year, you call it slower just given that...
Doug Bettinger
They're both on...
Joe Moore
I mean last year, week here already for the -- NAND was not for you, I think, but for the industry. You've got a situation where the technology investments are still pretty critical there. Is it possible that those numbers end up being kind of more similar to one another?
Doug Bettinger
I don't think so, but things can change. Things do change. What always happens though is almost always, when we're in the middle of one of these cycles is the technology node transitions still happen, maybe at a little bit slower pace, but raw capacity additions stop or meaningfully slow down. And that's what I see happening this year, Joe. But both are still moving forward to a certain extent, maybe a little bit slower like I said, utilization is down more than I've seen it for a very long time.
And like I said, when we're in one of these cycles, what ends up needing to happen is the inventory needs to get consumed and it will, and it does, and it is happening today. And once that happens, then spending growth will reemerge.
Joe Moore
Okay. Great. And then in terms of the foundry-logic piece of your business, I think to feel like nowadays we have to split that into multiple parts because you have this trailing edge dynamic that I talked about, okay? And first on the cutting edge, how do you think Lam is positioned? And I think as you look at next-generation structures like gate all around? Is Lam -- it seems like these etch steps are becoming more critical? Is Lam positioned to see the same type of inflection that you once saw in NAND that you could see something maybe not the same magnitude, but something similar in logic?
Doug Bettinger
Yes, it's more evolutionary in terms of incremental growth is what I see from the node to node to node different than. There was a big step-up in NAND that you've been asking me about already. I see incremental growth occurring in foundry and logic, especially at the leading edge. And maybe a great example is there's a new structure called gate-all-around coming out at the N2 node very likely, that requires a fundamentally different approach to etching the structure.
And if you look at it, again, lots of people in the room that don't know the story, the transistor architecture has a bunch of sheets deposited down. And to etch that, it requires a whole different approach. It's a selective etch approach, which we have done extremely well with in terms of winning production tool of record decisions there, Joe.
And so when I look at the evolution of leading-edge foundry and as it moves to the gate-all-around structure, we will benefit because we are the leader in etch technology in the industry. Again, for those of you that are new to the Lam story. And that is a tailwind for our business at the leading edge as it goes forward.
Joe Moore
And you talked about a bigger presence with some big customers there, that would seem to imply that you're going to be bigger when you get to these new structures?
Doug Bettinger
Yes. One of the things specifically that we have going on is a meaningful share gain on the conductor etch side of the business with one of the large -- larger customers, you can guess who that might be. But we've been scribing this share gain for several years, knowing we have won the positions, Joe. And you've heard us talking about this for quite a long time.
My CEO described on the last earnings call, a doubling of conductor etch share node over node as that large customer moves forward. It's very much what we're seeing. In fact, last quarter was an all-time record for us in our logic segment of the business. So we're really excited about what's going on there. Again, independent of the memory cycle, we're extremely strong in memory. We've got some very good things going on in on the foundry and logic side of things.
Joe Moore
Great. And then your competitor last week announced the tool, Applied Materials announcing an angle ribbon beam etch system and kind of the headline of reducing the need for double patterning on these EUV layers. I don't want to put you in a position of having to sort of talk about. But can you just talk to -- is there a discontinuity there, that's something different than what you would expect? And they've talked about working with customers on this for several years. Is there any kind of discontinuity that you see change in the multiple patterning?
Doug Bettinger
No, there's not. And this is something we looked at several years ago and chose not to invest in. It is a relatively small application. We just didn't see -- we saw bigger return areas for us to deploy our R&D spending than in this application, Joe, frankly. It's a pretty small application.
And having said that, multiple patterning and foundry and logic is still extremely prevalent in the process flow. This isn't going to change that in any meaningful way.
Joe Moore
Great. So then talking about the trailing edge portion of the logic business. We've heard today from some of the customers that they continue to be tight on some of the…
Doug Bettinger
Trailing edge…
Joe Moore
Five nanometer nodes, yes. You're seeing the investment in China that you alluded to. You're seeing investment in areas like silicon carbide. Like I've been kind of cynical about this idea that these markets would stay strong, but it looks like...
Doug Bettinger
But they are staying strong. They are.
Joe Moore
So once again, I'm wrong, it happens. But it seems like I guess, what's your visibility into those growth drivers?
Doug Bettinger
Reasonably good. I mean when you look at kind of the -- everybody is focused on leading edge investment and technology nodes going to N2 and so forth. But there's a robust investment needed in areas that serve end markets like industrial, automotive, power. In some ways, when you look at semis broadly right now, I see more resilience in some of those segments than in maybe smartphones as an example.
And so there's a decent amount of investment occurring at nodes, 28-nanometer and above that aren't necessarily brand new tools, but tools that we've been selling for many years where they're needing to add capacity to service those segments to the market. And I see secular growth there, frankly, right, in automotive, autonomous vehicles, electrification and so forth, driving a real content story. And I think that's going to continue for the foreseeable future, frankly.
Joe Moore
Great. That's helpful. So -- and then one of the elements of your story that I think is really compelling heading into a more challenging tools environment is services and your...
Doug Bettinger
I love this part of the business. Thanks for asking about it.
Joe Moore
I mean your service revenue was 35% of trailing 12 months. It seems likely to be quite a bit higher.
Doug Bettinger
It's likely it's going to be higher than...
Joe Moore
What's happening in the tools business. So can you just talk to that business? And how is it that you can kind of drive steady growth?
Doug Bettinger
Yes. Let me spend a little bit of time. Again, lots of people that might be new to the story in the room here, but there's an amazing part of our business. So people know us for selling new equipment that enables the manufacturing of leading-edge semiconductors.
An awesome part of our business model actually derives from the fact that our tools actually will run for decades. We have tools that have been in the field for decades. And that provides a real profit stream over the years in areas like service, spare parts, our equipment can be upgraded. And at some point, can be refurbished or we will sell more of that legacy equipment to different segments of the business.
Again, for those of you that are new to the story, sometimes I will compare this to an automotive dealership. A lot of the money in an auto dealership is not made selling the new car, it's made in the service department. And the same is true at Lam Research. It is an awesome part of the business model, in that, like I said, runs for decades, grows almost every single year.
Now having said that, it's down a little bit this year because utilization in the industry is down. But it is super cash generative. Most of the R&D needed to do what we do is deployed when we develop the new tool. And then we call it the Customer Support Business Group, or CSBG, will generate a profit stream over the life of the tool for decades. It's an awesome part of the business model. It's almost the razor/razorblade model, super cash generative and a key part of how we go-to-market and continue to drive the profitability of the business.
Joe Moore
And can you talk about, I mean you're growing that business considerably faster than the installed base, everyone is actually the top five equipment companies are?
Doug Bettinger
Maybe us, more than anybody, Joe.
Joe Moore
Yes. No, I think you guys have. But I think can you just sort of talk to that? I mean, what does that look like from the customer's perspective? How is that money getting spent such that people are spending more money over time on servicing these tools?
Doug Bettinger
I guess a couple of things I would point to. One is what we call advanced services. So historically, the service part of the Customer Support Business Group has kind of been an engineer on-site, show up, do a task, do a preventive maintenance step. So the tool runs in the most productive fashion. It's do a task.
What we've done over the last several years is begin to change the way we deliver service to be more about delivering a result, using machine learning, AI-type techniques so that we understand what's going on in the tool and can go into the customer's facility, observe what's going on with the fleet of tools and deploy service offerings that guarantee a result, better uptime, quicker mean time between cleaning the tool. And that changes the conversation with the customer from, okay, this is a cost plus business, meaning you're doing a task to what is the value of the result that we're able to deliver to the customer.
It changes the whole conversation around what service is and moves it more to a value-based conversation as opposed to what does it cost to do the task. That's been a key part of what we've been able to do, Joe, from driving dollars more than just the number of chambers in the field. So that's one.
Second, I would point to the most advanced etch equipment tends to be somewhat more spare parts intensive, because we are pulsing a higher level of power inside the chamber. An etch tool is -- it's a destructive process. You're removing something from a silicon wafer. Again, to use the automotive analogy, it's like brake pads, things just wear out.
And so the intensity of spares replacement is higher and etch probably than any other tool in the fab. And so as we're bringing more and more of the most advanced etch capability out, the spares intensity actually goes up from that as well. So those two things together, Joe, are largely what has enabled us to deliver the growth as strong as we've delivered on the service side.
Joe Moore
Great. I'm going to pause and see if there's questions from the audience. It's a big crowd, but we have some in the front here.
Doug Bettinger
I see some [indiscernible] over here. Don't ask me any Twitter questions, please.
Question-and-Answer Session
Q - Unidentified Analyst
Doug, maybe can you talk about what's your early view on 2024 WFE? And how are your discussions with your customers? What are they telling you, especially given the long lead times, you're probably having those kind of specialty?
Doug Bettinger
Yes. Too far -- too early for me to get out too far over my skis in terms of what the actual dollar amount of WFE is going to be next year. Hard to say, frankly. And part of it will be dictated by what is the economy going to look like and what is the demand for semiconductor is going to look like and so forth.
Having said all of that, I would point to you, though, that the biggest segment of our business in memory is at a 25 year low. They have a hard time envisioning that's not going to improve somewhat as we get into next year. I think time will tell, but I'm not ready to give you a number quite yet. Yes, please.
Unidentified Analyst
So you all have made some announcements on the EUV side of your business, extreme ultraviolet lithography, if I'm not mistaken, at some point.
Doug Bettinger
In dry photoresist.
Unidentified Analyst
Dry photoresist, yes. Does that mean your material sales are going to increase at some point? Does that -- is there any correlation with that? And what's the forecast in the future? When do you think that will materialize actually?
Doug Bettinger
Yes. So we don't sell materials. We sell equipment. And so what the dry photoresist product offering is new equipment that enables EUV in a more significant way when the industry moves to high NA probably at N2. I would point to the fact that this is a segment of the business that we don't sell anything into today. So this is all incremental to our business.
I would tell you that every one of our customers that uses EUV has our hardware in the lab looking at what's able to do? How is it evolving? What is it going to do as we go into the future? I'm extremely excited about kind of the interest in this. Like I said, everybody is using it, everybody is looking at it. And so stay tuned. Technology is progressing quite well.
Joe Moore
Maybe we could just close with the way you think about your priorities going through this downturn. Are you protecting this year's earnings? Are you thinking about the next upturn? And what are the uses of cash as you go through this? I mean, one of the really nice things about this business compared to some of your customers, you're cash generative even as bad as this can get, you're still going to be in that position of generating incremental free cash. How do you think about those priorities going forward?
Doug Bettinger
I guess what we're trying to do as a leadership team at Lam is position the company to be better when growth resumes than when we went into this current downturn. That's very much what we're doing. We're pivoting more towards our Asia footprint where we are closer to the customer. We will be able to turn information with the customer more rapidly. We will need to fly equipment shorter distances, which will make us more efficient, less cost and so forth.
So Joe, then relative to uses of cash. The first thing I would point to is, this will be -- even though our business is down, this will be a very strong cash generating year for us, because working capital will come down in a significant way. Inventory will reduce. So that will show back up as cash on the balance sheet. Accounts receivable will come down. That will show up as cash on the balance sheet. And so I'm pretty encouraged by what cash is going to look like. This year it's going to be a very strong cash generating year.
And the intention of the company has been and will continue to be -- we return 75% to 100% of free cash flow to equity holders through an annually growing dividend. We've got a very compelling dividend at the company as well as a pretty significant stock buyback. I think I've got about a $5 billion stock buyback authorization still remaining as we ended the last quarter. So you'll see us continue to do what we have always done, which is return cash to the equity holders over and above what we need relative to investing in the business.
Joe Moore
Great, Doug. Thank you very much. Appreciate it. Thanks everyone for your time.
Doug Bettinger
Awesome. Thank you guys.
For further details see:
Lam Research Corporation (LRCX) Presents at Morgan Stanley Technology, Media and Telecom Conference (Transcript)