Lam Research Corporation (LRCX) Earnings Call Transcript & Summary

Great. Thank you, everyone. Good morning. I'm Vivek Arya. I cover semiconductor, semi cap equipment at BofA Securities. I'm really honored to have Doug Bettinger, the CFO of Lam Research, join us this morning. I know we had a little bit of a delay, but we are all sorted. Really warm welcome, Doug.

And maybe let me just kick it to you. Just to kind of give us a quick kind of intro, a state of the union. What are the top 2 or 3 things you're finding the most interesting and exciting about the industry right now? And then just some incremental headwinds that you are facing.

Sure, Vivek. Maybe let me start, remind everybody of our safe harbor language. I may make forward-looking statements today. Please have a look at our forward-looking statement. I don't know if the slide has popped up, Vivek. But anyway, I'll refer to that. Listen, Vivek, I've been in the industry a long time. And quite honestly, I've never been more excited for what's going on in the industry. When you look at the last 12 months, COVID has really accelerated the digital transformation of society in many, many ways. Maybe a great example is how we're interacting right now. The demand for semiconductors has never been stronger and is very broad-based. So that's the backdrop when I think about what is going on with Lam Research and our business is this overarching demand for semiconductors, one. Two, when I look at kind of what's going on with our own business, when you layer on rising capital intensity on top of the very, very strong demand for semiconductors, you just have a great backdrop for where our business is trending. And in fact, we came into the year expecting the industry to invest, I don't know, call it, roughly $70 billion in wafer fab equipment. Fast forward only 90 days, and we suggested it's now trending above $75 billion. And it's a broad-based set of investment this year. We came into the year thinking, it looked like a little bit of a first half-weighted year. And in fact, now we believe it's a second half-weighted year. And it's strength in every one of our end markets: NAND flash, DRAM, foundry and logic. On top of that, when you look at the fact that, as we look at each of our customers' road maps, we see a growth in our addressable market at every single process node. And as an example, as you walk from 14 to 10 to 7 to 5, 3 and beyond in foundry, our opportunity gets bigger. That's what gets us excited. And when you layer on top of, what I really believe, and I'll thank the employees at Lam for delivering this, we're executing extraordinarily well in spite of the fact that COVID is still with us. It may be getting better in the United States. It's still a challenge in the Asian regions. We've learned how to operate quite efficiently, quite effectively in this environment. And we're delivering to customers. We're taking care of our customers first and foremost. So maybe broad brush, Vivek. That's what has us and me particularly excited about the opportunity I see for Lam Research right now.

Very good. So Doug, maybe just picking up on that. You mentioned when the year started, the WFE was supposed to be about $70 billion. Now it's over $75 billion. I know we have heard numbers as high as $80 billion, right? Next year seems to be even higher. So just looking at calendar '21, how are you positioned from a supply perspective to deliver against those growth commitments? Like does supply put somewhat of a cap on how much WFE can be this year? Or even next year for that matter. What are you doing to improve supply?

Yes. All of that, really good questions, yes. As things strengthened, we have worked to figure out how to deliver more to customers. I will acknowledge our lead times are somewhat stretched out right now. The strength in demand is a little bit more than we expected coming into the year, obviously. And so what are we doing about that? First, every location at Lam that we manufacture, we're doing our best to actually extend our output capability. I'll talk through that a little bit. We were fortunate in that we had plans this year to launch our new factory in Malaysia, which I would just mention, over time, will likely be the largest factory in our network over the next several years. We're doing our best to accelerate that road map so that we can deliver output sooner and at a greater volume. We're having some level of success with that. Second, I mentioned in our earnings call not too long ago the fact that we bought out our joint venture partnership in Taiwan where we manufacture our Reliant product line, our refurbished product lines, so that we can more directly control our ability to extend output there. We've got manufacturing capability in Korea. We're extending our capability there. Similarly, our largest factories today are in California and Oregon. We're incrementally increasing our capability there. And I would also just mention, along with all of this, we are bringing our supply chain partners with us, to also increase their capability. It's a complicated set of things that we're working on. But what I would say is we're incrementally delivering more output as we speak. So yes, it's a challenging environment, but one we know how to execute in. We know how to do this stuff, and we're working to increase our capability.

Got it. Given this tight environment, Doug, how do you think about the visibility for next year? Like when customers look at this tight environment, are they working closer with you than usual so that you might have a better idea of the following year versus what you might have had historically? Do you think that situation is no different, right, than what you have seen historically?

Yes. We always work very closely with our customers. We need to know what their road map is, where they're going. If there's a new location coming online, we have to be able to hire and train a workforce there to support them. So I would suggest to you, we always have pretty good visibility into where things are going. You never know exactly to what magnitude perhaps the equipment purchases are going to be. So I don't know that, like all of a sudden, it's a whole lot stronger. It's always been quite strong. Or at least over the last several years, it's been quite strong. Certainly, the communication from customers about what they want in the near and midterm is pretty robust right now, as you might imagine. I would suggest to you, we're probably talking to customers about this kind of stuff every single day. But we always have pretty good visibility, and we have pretty good visibility as we speak.

Got it. So just given your lead time, do you think that you're comfortable thinking about 2022 being a growth year for the industry?

Vivek, at this point, we haven't quantified 2022. It's too early. I mean, it's early June, right? I mean, there's a long time between here and next year. Maybe I'd give you a little bit about how I'm thinking about it, perhaps. I'm going to decline to quantify things simply because it's too early for us. We do that later in the year. It does feel to me like -- first, let me step back. The demand for semiconductors is strong, and that is -- that's here, right? That's not going to go away. I mean, that, again, I mentioned it earlier, that's the backdrop for everything that's going on. Layer on top of that, you've got capital intensity ticking up pretty much everywhere you look. That's an important backdrop to think about as you think about a multiyear -- and in fact, the most important thing to me isn't what's '21, what's '22? It's what does the next 3 to 5 years look like in this industry? And it looks good, it looks strong. So that's how we think about running the company, is what does the next several years look like? But when I think about technology road maps in '21 and going into '22, in foundry and logic, you've got one of our large customers talking about a 3-year investment profile that suggests they're going to grow their investment, right? I think that's pretty well telegraphed. You've got road maps in DRAM that are moving from 1x to 1y to 1z to 1alpha and beyond. You've got a demand backdrop that's strong for bit demand there as well. We've got a similar thing in NAND flash as we go to 128-layer devices to 196 and beyond. So when you look at technology road maps and rising capital intensity, that's a pretty good backdrop. When you think about the demand for semiconductors, that's a pretty good backdrop. That's a strong backdrop. So as I think about going into next year, feels like it's going to be a good year. I'm not quite ready to give you a number, but it feels like the demand trends are strong. And we're set up well. So I'll leave it at that. We'll see how this year unfolds. Feels good this year. It feels good going into next year.

Got it. Got it. So let's talk through the different parts of the business, Doug. So on the NAND side, every week, right, there is always some concern about, oh is NAND peaking? Because in the past, you had set a market size of $14 billion. I think this year, that's probably annualizing to a much higher number than that, right, maybe $17 billion, $18 billion or so. How much of that is just kind of one-off spending in China, as an example? How much of that do you think resets the base for NAND? And if -- without quantifying, do you think from this point on, can NAND actually -- what will help NAND grow in the second half and then next year?

The one thing, again, if I think about the long-term backdrop, demand for semiconductors is strong, right? I mentioned that. I think that's my third time. And the industry doesn't consume logic devices without low-latency DRAM and NAND flash storage. So this stuff all goes together. If you're excited about one, you're excited about all of it. And I am excited about all of it. The timing of investment in each of those segments may be slightly different, but the important thing, I think, to keep in mind is demand trends are good. And so when I think about that and then tie back to essentially what you're asking about, yes, 3 or 4 years ago, we gave the $70 billion over 5 years to generate high 30% bit growth, which is where we believe market supply and demand kind of clears itself. Honestly, in hindsight, maybe that oversimplification has confused people a little bit. Here are some things to think about. You alluded to the fact that when a customer segment is early in the ramp of a new capability, a new technology, there's probably a little bit of an inefficient spend. That may be the case in China a little bit right now. So that's one thing to think about. Also, I think it's easier to understand that, as the NAND stack gets taller, investment actually goes up over time. So when we gave the metric that you referred to, it was averages over the next several years, and we're now 3-plus years beyond that. So it's probably somewhat higher than what we suggested back then. So anyway, it was an attempt to help people understand things that may now be confusing things a little bit. But be that as it may, like I said to us, the most important thing is we see technology road maps that are multiyear. We see strong demand for semiconductors. We see supply and demand kind of getting in balance in the NAND space, to directly answer your question. And we're excited about our enablement of that. We'll see how things play out over the next several years. But like I said, my excitement about where things are going is really very strong. And I think next year sets up to be, like I said, a good year.

So Doug, if I had to ask the question a different -- let's say, if you had to come up with a NAND number for the next 5 years, right? I imagine it would be higher than the $70 billion. Would it be $90 billion, $100 billion? Like -- or do you think we are just temporarily in this elevated situation, that we will get back to the $70 billion? Like what is -- what can conceptually be the number for the next 5 years, right, if you had to make [ a number ]?

Yes, Vivek, I don't have a new number for you, except it trends up over time. Capital intensity rises process node by process node. That's the right way to be thinking about it. And the demand for bits is continuing to be very strong.

Understood. And Doug, do you see a difference in views between some of your peers have guided to second half being stronger in DRAM, more muted in NAND. I think your view has been kind of more NAND-weighted. So first, is that the right -- am I describing it accurately? And if I am, has this happened before also, where there have been different views about which part of the memory market, right, could be stronger or weaker, depending on the kinds of tools you're supplying?

Vivek, we haven't said one thing is stronger or weaker than another [ segment ]. What we said, just to clarify it, we see growth in every segment of our business in the second half as compared to the first. That's true in NAND, it's true in DRAM, it's true in foundry and logic. I haven't said one is stronger, one is first, second or third. It's just everything is growing in the second half. I would suggest to you, we have good visibility in what's going on. When we talk about this stuff, it's based on plans we're hearing from customers. And I would also suggest to you, Lam Research probably has the best visibility into NAND flash because our enablement of all of those critical applications of anyone in the industry. So I feel good about what I'm telling you. I don't know what other people are saying, right? I can't get into the 4 walls of their company, nor should I try to speculate about what they're seeing. All I can do is tell you what we're seeing, and that's what we're seeing.

Understood, and fair. Now Doug, one thing you announced at the last Analyst Day was that Sense.i, right, your new etch platform. So talk to us about what kind of -- where you are in that ramp. What kind of feedback have you seen from customers? And what I find the most intriguing is the replacement potential, right, of all the tools that you have out there. So talk to us about this new platform. And when you will -- when do you think it will start showing up in the financials?

Yes. Let me talk about, first, look, what is Sense.i? Sense.i is the first bottoms-up redesign of the etch platform from Lam Research in over 20 years, right? The last time we did this was when the industry converted the 300-millimeter wafer size. So that's a big deal. That's a strong statement of the leader in etch technology redesigning the platform to develop capability that we see the industry needing for the next decade or more. So what's special, what's unique about Sense.i? I really think about 3 things. First is technology, right? We're designing the chamber to enable what we see as some of these very high aspect ratio etch needs of the industry for the next decade. So that's important. We haven't done that for quite a long time, one. Two, it's the most intelligent platform we've ever delivered to the industry in terms of equipment intelligence. I like to almost think of a tool as being self-aware of what is going to be going on in the configuration itself such that it can adjust things. It's able to do some level of self-maintenance, which is really an important value proposition for our customers. And then third, it's the most efficient platform that we've ever delivered. Most efficient platform in the industry in terms of output per square meter of fab space. So when you put all that together, we think the value proposition here for the industry is going to be extraordinary. And I would tell you, we've chosen to first deploy this into the NAND space. Eventually, it will go everywhere. And the customer pull that we're seeing is very strong. We're working collaboratively to fine-tune what different capabilities of the tool are and will be, such that we'll be ready to begin higher-volume shipments as we get into the latter half of the year and, more importantly, into next year. We announced the Vantex chamber design, which is our dielectric chamber, very recently. So that's an understanding of the road map continues to evolve and will continue to evolve. But I would tell you, Vivek, when I see customer pull for something as strong as I'm seeing for this, you know you got something special here. And we're excited about where this is going to go.

Got it. Yes, that's what I find very intriguing, Doug, that is there a way to think about what are the additional economics this can provide as you go through that? Like is there a way to construct a simple replacement model, right? How many tools will it be able to replace and over what time frame? And then what's the additional content you could get per tool over that time frame?

Yes. I'm not ready to give you hard numbers on this yet. I mean, it's a key component. If you remember back at the Investor Day in March of last year, we talked about an objective to getting 4 to 8 points of market share in etch. This is a key enabler of that. It will be a very competitive product in terms of us going after more of the semi-critical space, as an example, because of the cost-efficient output capability of the tool. So that's as much as I'm ready to share with you right now. We'll have more and more on this as time unfolds. But like I said, this is a key aspect of the market share gain objectives that we have for the company.

Got it. Similarly, the other important product that you also announced was dry resist for EUV applications, right? Talk to us about where you are in the collaboration with the key stakeholders. When do you -- when will you start to see that start to appear in the financials?

Yes. Thanks for asking about that, too. Yes, we're very excited about this as well. And the commonality between this and Sense.i is the customer pull that we're seeing, right? At this juncture, every one of our customers that uses EUV or plans to use EUV -- excuse me, either has a unit installed in their R&D facility or very, very soon will. So the collaboration is very strong. We are partnering exclusively on dry resist with ASML at imec, and very excited about what's going on there. So work continues. My confidence in where we're going with this is stronger than when we announced it. Certainly much stronger than when we announced it a year ago. Our view of this is, and it hasn't changed, is we see an incremental $1.5 billion in revenue over the next 5 years as this ramps. And we're working collaboratively with the industry right now to help fine-tune this as well, help get this where it needs to be to be production-ready. And stay tuned. We'll have more to talk about on this as time unfolds as well.

Got it. Is this incremental, Doug, this $1.5 billion opportunity over the next 5 years? So number one, is it incremental? Is it replacing something else? Then importantly, do you think, if the -- this will be linear over time? Or will it be more front half-loaded or back half-loaded in those 5 years out?

It will be a little bit more skewed to the back half of the 5-year time frame. And I would say the 5-year time frame started when we first announced this a while back. So have that in mind. But it will ramp, and it will more strongly in the latter years as process nodes progress forward.

Got it. The third technology question, Doug, I wanted to ask you is on the DRAM side. As the industry moves to more vertical structure, right, to more 3D structures, Lam did very well when the NAND industry made that transition. Historically, you're known more for NAND than for DRAM. So what is your position when the DRAM industry moves to 3D structure?

We're very excited about this transition. Work is underway today to figure out what is this going to look like. Too soon for me to put numbers on it for you, Vivek. But listen, Lam does well when structures and devices are 3D. And I think by definition, 3D DRAM is 3D, right? And so when you look at that, it plays to our strength. We see a variety of different challenges in the etching that's going to be required. We're working collaboratively right now, again, with the industry on this. There's likely some selective etches in there that are maybe somewhat new, and we're pretty excited about where this is going. But it's too soon for me to quantify, Vivek, it because the process flows haven't been ironed out yet. But I would suggest to you it's coming. It's going to be a good transition for Lam. And again, stay tuned in terms of us giving you the numbers on it.

Got it. So when you put it all together, that -- doesn't it suggest a higher WFE intensity for the industry moving forward? So we have been kind of in this kind of low mid-teens range over the last few years. Do you think, from an industry perspective, we should be prepared for higher WFE intensity going forward, right? And if -- on top of that, if you even add the initial investments that different countries want to make in terms of reshoring capacity, is it possible that we could be headed towards higher WFE intensity, at least for the next several years?

Yes, no, Vivek, I think it is possible. I've been reminding people, we -- over the years, we look at capital intensity as -- WFE as a percent of industry revenue. I would suggest to you, that's actually the wrong metric to be looking at. I believe people look at this metric to try to understand an affordability question. To understand an affordability question, you need to look at capital investment as a percent of profits, not revenue. And actually, when you look at what's been going on in the semiconductor industry, pricing is actually quite good for everybody in the industry. And so profitability has actually gotten better for everybody in the industry over time. So the right metric, in my opinion, to be considering in this -- to answer this question, is WFE as a percent of operating income or as a percent of EBITDA. And actually, when you look at that metric, it -- it's been pretty steady. It's been pretty constant. And that defines the affordability question, I think, people have on their mind. And so I'd encourage people to begin, if they're not looking at that metric, to consider that metric because that one does help define the affordability question. But then stepping back, again, I go back to -- I think about demand for semiconductor is very strong, rising capital intensity. And when I talk about rising capital intensity, it's -- I'm talking about CapEx per wafer as opposed to WFE as a percent of revenue. People use these 2 descriptions differently at times. So that's what I'm talking about. And when you think about those things, it does suggest WFE is stronger today than it has been in the past, for sure.

Got it. And then Doug, the final topic I wanted to touch upon was what, I believe, is one of the most attractive aspects of the semi cap industry, right, your Customer Support Business Group, CSBG, right? Because that is sort of this recurring part of the business. Where in other parts of technology, people are willing to pay 10, 15, 20x sales then, right? But here, somehow that this part kind of gets lost in the shuffle. But when I look at your CSBG business, right, last number of quarters it's been hitting record highs. But it's also growing at a much higher level than historical trends. So what is causing that? Does it get back to trend line? What's your level of confidence it can continue to grow from here?

Yes, no. I feel great about how that business has performed. And I sometimes jokingly will say, "This is my favorite part of the business model when you look at it." It's a very recurring aspect of what we do. Let me describe it briefly for people, perhaps, that are new to the story here. CSBG is Customer Support Business Group, is the acronym. There's 4 things in this business unit in terms of how we manage the company. First, it's spare parts. If you think about what's going on in the industry this year and why that is so strong, industry utilization is high, consumption of spares is high as a result of that. Those 2 things go together. So that's one, spares. Second, service. Our tools need periodic maintenance. And when utilization is high, you need maintenance more frequently to keep maximum output being generated. So that's the second cylinder of our 4-cylinder engine. The third, and we've alluded to this briefly, but our Reliant product line, which is focused on lagging edge process nodes, right, read that to be a 28-nanometer and above, where we go to market oftentimes with used equipment or refurbished equipment. That's very strong. Several -- many consecutive quarters of record business there. And then the final component is the upgrade business, productivity upgrades, technology upgrades. When we sell our tools, oftentimes, we're able to go into the fab and upgrade certain subsystems, if you will, on the tool itself to make it more capable, more productive, the next technology node. So when you -- and that business is always a very high return on investment for our customers. It's part of the value proposition of buying a tool from Lam. So when you think about the backdrop of the industry right now, this business is hitting on all cylinders. So that's the way to think about this. And yes, it certainly is performing above what I would have expected a year ago. A tailwind for this business as you think about the next several years is actually the number of chambers in the field. It's a very strong WFE year this year. The number of chambers we're shipping is going to be quite high. And so that defines the opportunity for this business to continue to grow. Because the more chambers in the field, the more spares, the more service and so forth that are able to be delivered to the customer. And so we always give chamber count as a metric at the end of every calendar year, and we'll do that this year. But I expect it to be a very strong year in terms of number of chambers we're shipping. And I believe that defines a lot of the growth opportunity we see for the Customer Support Business Group. So I feel great about where we're going with this business. It's a very recurring part of the business model, very cash-generative, which is important when I think about our capital allocation model. This helps support a lot of how I think about capital allocation.

Absolutely. And I was just going towards that final question, which is kind of return of cash to shareholders, right? We have seen very strong returns over the last 5 years. And I think you -- you mentioned CSBG kind of supports dividend and dividend growth. So how do you think about capital returns going forward? Because I know M&A in this industry is a tougher proposition, given how consolidated it is. So how do you think about kind of dividend growth versus buybacks? Because the stocks, right, are still kind of trading below market multiples, which I find astounding. So what's the right balance...

Yes, no. I do, too, think it's...

Yes. Sorry, we are on the same page. So what's that right balance between dividends versus stock buybacks?

Yes, let me tell you how I think about -- how we at Lam think about capital return. First, we've said publicly, our objective, our plans are to return 75% to 100% of free cash flow back to shareholders. So that's the backdrop. Again, actually, we've done more than that over the last several years, but that's how we think about it. Like I said, the business is generating a lot of free cash flow because of the strength of the business, because of the growth in CSBG. And listen, we're committed to continual growth of the dividend. People like to see that growing on an annual basis, and that's what we intend to do, grow the dividend every year. We're coming up on the 1-year mark not too far down the road, so I'll have some more to say about that not too far from now. And then we supplement it with the share buyback. I continue to believe deployment of cash to buying our shares back is a good use of cash because I'm very excited about the future opportunity for us to grow the company. So that really does define how we think about capital allocation. And your suggestion about M&A, yes, I think large-scale M&A in this industry is largely behind us. So I don't really view that as something that's going to consume a significant amount of cash. It's -- M&A is pretty much done in this industry.

Great. On that positive note, thank you so much, Doug, for joining us this morning, taking time to share your insights about the industry and Lam Research. We really appreciate it. Thank you so much.

Yes. Thanks, Vivek. Thanks for having us, Vivek. Good luck with the rest of the conference.

Thanks very much. All the best.