Applied Materials, Inc. (AMAT) Earnings Call Transcript & Summary

All right. Thanks for joining. I'm Blayne Curtis, semiconductor and semi cap equipment analyst at Barclays. Very happy to have Applied Materials as our next presenter. From the company, we have Dan Durn, CFO. Dan, welcome.

So I always like to start off with a very high-level question. And I think we were just in the breakout room joking about how everybody is kind of waiting for cyclicality. And I think it's inherent in our nature of being in these markets to kind of wait for that turn down. Things so far for the conference have been very good across the board. I wanted to ask you a long-term question on the TAM. I think this year, everybody was waiting for that downturn. It can't possibly be that strong. Clearly, that hasn't come. We'll get into next year. But I'm kind of just curious how you think about things. If you look at the wafer front end TAM, it was flat for '07 and if you take out the financial crisis, maybe even longer, then you saw the spike up in '17, '18. This is why people think it needs to kind of turn down. So maybe you can you just kind of reflect on what people missed there? What changed? Why was it flat forever? And then as you look forward, I think people have talked about $100 million -- $100 billion TAM, right? Assuming semis keep growing, why should the semi cap TAM grow as well?

Yes. So I appreciate the opportunity to join you this morning, Blayne. So thank you for that. And thanks to everybody on joining us virtually to hear the Applied story. So it's great to be here with you. I do think we're at a special point in our industry where the opportunity has never looked better. And if we take a look at the fundamental drivers of the industry, I do think there's something pretty important, pretty profound happening as it relates to end-market demand. You're seeing digitization take hold. You're seeing an interdependence of global economic growth and technology enablement. And we think those trends only accelerate going forward, that they don't slow down. And when we think about the drivers of end-market demand for semiconductors in the handoff from handset cell phones as the predominant driver, a consumer-oriented device to something that's nonconsumer discretionary spend, largest companies exposed to the largest industries, where investment as a key enabler of their business' success, it's existential. And when you think about every industry on the planet being transformed by technology, you're talking about an economic pool that's orders of magnitude greater than a consumer-oriented device. And if you take a look at third-party research, whether it's VLSI or an industry organization like SEMI, they're talking about the semiconductor industry doubling in size in the next decade. So close to $1 trillion by 2030. And if you look at it, it took the semiconductor industry about 20 years to double, 2000 to 2020. And now they're calling for it to double again in the next 10 years. So growth is accelerating in the industry, not slowing down. Our industry, I do think that the long-term average of 12% size of our industry to overall semiconductor revenue is something that does stand the test of time. It's a 4-decade average. And we do expect to grow in line with overall semiconductor industry. In fact, you could argue that there's an upward bias to that capital intensity number. And if that plays out, then you could see us growing at a rate greater than overall semiconductor industry. So based on everything we see, we're incredibly bullish about where we sit today. A multi-decade-long secular trend that's playing out, that's going to benefit the overall semi industry in a material way, but it's going to be an incredible opportunity for us at Applied and how we're positioned in the industry. So we're incredibly excited by what we see.

I do want to drill down into that capital intensity, that's 12% long-term average. Obviously, I don't think wafer sizes are changing anytime soon. Maybe with EUV, things get a little easier on the logic/foundry side. But obviously, devices are getting more complex. So how do you think about that? And you said it's biased higher, what would be biasing it higher?

So I think there's a couple of things, right? Complexity is our friend in this industry. We know the road map is becoming more complex. And when you think about the demand for compute power never being greater than it is today, we think those trends today are stronger than they were a year ago. And we think we do see a fundamental acceleration in the adoption of technology in those industries. So demand for compute power has never been greater, but traditional Moore's Law scaling in the 2 dimension space, that's hitting a wall. That's not working anymore to deliver the simultaneous benefit of power performance area to cost at an attractive time for market. That playbook the industry has been operating under is beginning to slow down. It's beginning to hit a wall. And Gary's talked about this now for several years. The industry is gravitating towards a new playbook to deliver the power performance road map. And it's going to have multiple elements to it. It's new types of architectures, new types of structures on the chip going vertical, new types of materials, new way to connect chips together, new ways to shrink and it's not only us that is sharing that perspective. You see the industry gravitating towards that perspective. If you look at leading foundry customers and leading logic customers that talk about their road maps going forward, a leading logic company will have 6 elements to that playbook. We have 5, a leading foundry customer has 5. The industry is -- has a consistency in their messaging around what's going to drive the power performance road map. We, as a company; we, as an industry, have a very material role to play in enabling each vector of the new playbook. And so complexity is our friend. New playbook is taking hold. And I think that's going to provide a nice tailwind to us as an industry and upward bias in capital intensity over time.

Do want to bring it back to more immediate term here as I know you haven't officially guided for 2021, but you do have some thoughts on it. Maybe you can walk us through, and like I said, I'm hearing -- I would have thought maybe some markets will be down next year, they may not be. So moving target, so I'm not going to hold you to it, but kind of just give us your thoughts by your end markets for next year?

Yes. So here's what I would share, if we take a look at how 2020 is playing out, we still have a couple months -- well, actually 1 month left to go in calendar 2020. Here's our best read of the market as we sit today. Overall, wafer fab equipment market, we said up 10% to 15%, but we're gravitating towards the higher end of the range. We think it's a 15%-ish grower for the year given the strength that we're ending the year with. As we disaggregate that across device types: foundry/logic, a strong year, greater than 55% of the spend. Memory, outgrowing foundry/logic, NAND is the strong grower in the market. So if the market is growing 15%, NAND is probably growing 2x the market this year, 30%-ish growth. DRAM is probably in line, a little bit above-market growth. Foundry/logic is growing below the overall industry average against that backdrop of 15% overall growth with strong NAND growth. We, as a company, we're going to grow our systems business at the midpoint of our guide over 25%. So strong performance against a memory-driven dynamic this year. We feel really good about that positioning and performance. As we look into next year, we would expect foundry/logic to continue to be strong, also greater than 55% of the overall spend. Within memory, though, where 2020 was a strong NAND year, we think the spend levels in NAND are going to be roughly flat year-over-year. The market is exiting 2020 with a more balanced supply-demand dynamic. DRAM is going to be your strong grower next year. We still think it's undersupplied exiting the year. They're working through inventory, and we do think that's going to start flashing as a strong grower from a spend perspective next year. So that gives you the rough contours of how we're viewing the markets into 2021 off of where we think we're going to exit 2020.

Thanks. Obviously, I think everybody's aligned that DRAM should have a good year, and it's kind of coming off of kind of like NAND was of a little underinvestment, it's undersupplied. In fact, the end markets probably are better with servers probably up next year. Handsets are quite strong. All the consumers of DRAM seem quite good. So I want to ask about the other markets, right? I mean, I guess, with that same backdrop, with these strong end-market trends, why would you see foundry/logic, kind of -- you said strong, but I'm assuming that's going to remain at strong levels, right? Not actually grow dramatically next year. So how do you think about the puts and takes of foundry/logic?

Yes. So and I'll come back to the secular trends where we started. We do think the demand for compute power is greater today than it's ever been. We see strong momentum at 7 nanometers and uptake by end market customers, same strength we see at 5 nanometers, and we're going to see follow-through in strength again at 3 nanometers based on conversations with customers and how they're talking about their business. You're not going to see reuse amongst those nodes. Each of them are going to be large substantive nodes that the industry is going to gravitate towards because they've got very attractive power performance characteristics. Against that backdrop, our opportunity as a company at 5 nanometers is greater than 7, and our opportunity at 3 is greater than 5. So node-over-node momentum for us as a company, we're really encouraged by. The other observation I'd make is what keeps foundry/logic strong. And again, we think this is a long-term secular play that we are in the very early innings of, and we think this is a multi-decade trend. So we're in the very early innings of it, but it's just beginning to play out. You'll also see a diversification of spend. You're seeing lots of investments from a trailing, no geometry standpoint. As part of the Internet of Things is part of chip-enabling physical infrastructure of the world, you're seeing a lot of intelligence being pushed to the edge. You're seeing a lot of sensors at the edge, which can generate data, which companies are using to gain insight to make their companies more efficient and service their customers even better tomorrow than they do today. And so we see that diversification and spend trailing those geometries, leading-edge geometries, multiple customers, multiple nodes and that diversification and breadth of spend takes volatility out of the system as the industry grows and matures. So again, we continue to be really encouraged and excited about what we see. We think the opportunity has never looked this good in the industry. And saying that for a while, every quarter that goes by, we see it playing out. We're just really excited by what we see and the trend around this industry.

You mentioned the trailing edge, you mentioned it on your earnings call as well. I think since then, we've seen actually companies talk about shortages. And so I'm kind of just curious what you've seen, is that accelerating now that we're seeing companies not able to ship? That -- can you just talk about that trailing edge capacity and whether you've seen a pickup there?

Yes. So we do see the spend environment continuing to be robust. I think you're going to see a recovery over time in automotive and industrial as a result of the pandemic and the initial supply chain disruptions and consumer behavior. We do think those markets were initially disrupted probably more than others. But we see recovery in those markets over time. Again, the environment continues to be robust, and we do think that we're going to continue to see that segment of the market probably outgrow the overall WFE market over time. Because, again, there's so much capability that sits at the edge from an intelligence standpoint that companies can use to gain insights to make their operations more efficient, serve their customers better. So the customers' customers' ability to monetize those investments, I think, only gets greater over time. And again, I think we see this playing out, and we're pretty excited about it.

What I see on the NAND market, obviously coming off a very strong year, I can understand why you maybe would think you wouldn't grow. If you look at some of the plans from companies that are in the NAND market, it may be up and particularly strong in the first half. So kind of just thinking about the puts and takes in that market gaining perspective first half or second half next year, maybe just waiting for this oversupply. And pricing has been down, it hasn't been all that bad. So maybe there's -- this fear, once again, isn't as bad as people think. But just any kind of thoughts on all of those items in the end market?

Sure. And so this is where I really do think the behavior of the customer, there's an evolution of the customer behavior, where we're seeing very disciplined capacity adds to really dial in bit supply growth in line with end-market demand. And the industry won't be perfect. But when we entered the downturn 18 months ago or whatever it is now, a couple of years ago, whatever it is, when we entered the downturn, what we saw is our customers dialing back capacity deployment even before their margins were peaking. And I think that's the first time we saw in the industry that type of behavior from the customers in a very disciplined way. We see similar behavior coming out of the downturn now as customers are dialing in their bit supply growth to match end-market demand. And in NAND, we saw the market fire from a spend standpoint and a growth standpoint before DRAM. And we think we're exiting 2020 with that market more in balance bit supply, bit demand. And so we think that as we look into 2021, it's about technology migrations, we're not going to see capacity adds. We think wafer starts in the industry remain pretty consistent, and our customers are just going to drive their technology road maps. It's their cost structure, cost per bit. It allows them to make money and be competitive. And so we think that dynamic is alive and well, and it's going to lead to a consistent level of spend. Long term, we think the supply-demand balance growth each year is mid-30s, plus or minus, depending on which customer you talk to. And I think that just bodes really well for the industry. The last observation I'd make on this is, I think the bit supply growth and the capital to drive it is far more efficient today than it was 2, 3 years ago. You're investing in planar technologies and 3D technologies. So concurrent investment in multiple generations of technology. And you saw some pretty strong and significant spend 2, 3 years ago. But I would say that, that was relatively inefficient from a bit production standpoint. Industry is far more efficient today. So I'm encouraged by the discipline of the customers, the way they're viewing the market, and resisting the temptation to overshoot true end-market demand.

I do want to ask you by region. Obviously, China has been very strong, outgrowing the overall TAM, I think people generally say around $10 billion. You've removed most of SMIC out of your numbers. I think everybody's treating it a little bit different. I'm kind of curious, I think a lot are still seeing strength in that region even without SMIC. So just perspective on the strength of China, even without SMIC in your numbers and the sustainability of that region as you look to the next year.

Sure. And so as you pointed out, 2 months into our most recent quarter, there was a licensing requirement imposed on SMIC. Our Q4 revenues, Q1 guide would have been greater absent those licensing requirements. Where, for the purpose of the guide, we're assuming that the licenses do not come through. So you can see how we have derisked our view into Q1. If we're surprised, licenses come through, then you'll see upside to the numbers that we've put out there. But what I would say is that there's multiple customers and multiple end markets in China. If you think about the $9.5 billion of capital deployed, it goes to 200 and 300-millimeter geometries. Within 300, it's NAND, it's DRAM and it's foundry/logic. So SMIC is one of a number of local domestic China providers in the ecosystem that they're building. 2020 happened to be a strong spend year for them. Customers will go through spurts of spend and then digest. 2020 happened to be a big year for the one foundry customer in China. And I think our expectations around 2021 is they were not going to be a meaningful part spender in the market. So 2020 was their big investment year. And we were able to get through most of that investment cycle before the licensing requirement came up. So we see little impact to the overall market in 2020 as a result of that. And then as we window into 2021, because of their investment cycle, we didn't see them being a meaningful investor in 2021. And you see other companies showing signs of investment in 2021. That's not a change in our outlook or plans. It's what we'd expect 4 quarters ago. It's what we expected 2 quarters ago and sort of playing out in line with how we view it. Is 2021 an up year? Is it flat? Is it down a little bit? I think it's too early to tell. It will be another good year, but you won't see the type of growth that you saw in 2020 in domestic [ Chinese market ].

Do want to turn it to -- from a product perspective, we talked a lot about the overall TAM. Maybe turning to AMAT specific, are there areas, product areas, where you feel you have opportunity to gain share? Are there new product areas where you can open up additional TAM expansion areas? Maybe take a longer view on that as well.

So if we look at 2020, I think you see significant outperformance from a DRAM standpoint. I think you see strong outperformance from a foundry perspective, and you see in line performance from a NAND standpoint. If we disaggregate that by product, CVD, PVD, epi up 8 points of share year-over-year. Our metals business is up 42% year-over-year. So super strong performance despite the high share that we have in that position. Our inspection business is up 46% year-over-year. Our conductor edge business, up 30% year-over-year. Our CVD business, up high 20s year-over-year. Strong performance across the board in multiple product areas and multiple end markets, and we're really encouraged by that setup. As we look forward to strong drivers of growth, I would expect inspection, etch and CVD to be your strong growers. Those are areas where we don't have quite as high shares as we do in maybe some of our other businesses, which means you've got more headroom to grow, and we've got a great product story going on. CVD and etch patterning, it's a great story for us. We're co-optimizing patterning films with our etch technology. You're seeing strong areas of success in DRAM, strong areas of success in foundry/logic, and I would expect to see that momentum continuing going forward. Inspection, also a great product-driven cycle and story, optical pattern wafer inspection, our e-beam technology, best in the industry. And then you optimize across those 2 sets of capabilities and technologies to really dial in the technology ramp of customers but also in line, high-volume manufacturing. We're showing some significant signs of momentum at the very leading edge of technology. So we're really encouraged by what we see. But if you -- if I were to point 2 to 3 areas from a product standpoint, inspection, CVD and etch, I think, are your stories going forward. But we'll also do well in the other parts of the portfolio.

Great. You've done a great job answering all these questions. I'll give you a CFO question. The kind of -- obviously, the pandemic has impacted people's gross margins, shipping logistics. So maybe just talk about the near-term headwinds in [ ours ]. And then I asked you this when I came over from covering semis and returned to semi cap and said, these gross margins are fairly low. And I guess as you think about you going forward, I mean, we'll see if China can make a fab without your equipment. I think you'll feel the answer is no. Certain areas, they can't go around you. And I think maybe just talk about that puts and takes. Obviously, you might see some consolidation further on the foundry side. We'll see what memory does long term. Obviously, customer concentration I get. At the same time, you've seen in semis, margins rise by 10 points or more as particularly more mature industries like analog recognize the value. Now you saw that mixed signal, why can't semi cap by the next one where you actually get a better gross margin story? Obviously, all margins have been strong.

Yes. I think the long-term trends, I think you'll see that play out over time. I do think there's going to be an upward bias. The reason I get there is the technology that we are delivering to the market, and we, us and our customers, are delivering to the market are underpinning huge secular trends globally, the interdependence between technology and global economic growth, it's greater today than it's ever been, and it's going to get greater -- it's going to be greater going forward. So I do think that that's going to be a nice positive aspect for the collective ecosystem where more economic value accrues to the semiconductor value chain. So I think that's a good setup for the industry. The technology we deliver is highly enabling. I think what you see is a consolidated customer base, which facilitates very deep levels of discussion and collaboration, more so today than 5, 6 years ago with our customers around their road maps and the new playbook of how we can deliver that key enabling technology. So the depth of those relationships are getting greater as a result of the collective challenges and the opportunity statement for the industry being as large as it is, that's developing a nice collaborative environment to drive these road maps even faster than we have historically. The other thing I would say is we specifically, as a company, year-over-year, 110 basis point increase in our gross margins. So I think the company is performing well. If you were to take the most recent quarter, 45.7% gross margin, and you were to normalize it for the COVID environment and headwinds we see in this environment and map it to the long-term target gross margin, 45%, plus or minus 2 points, depending on where we are in the cycle, we'd certainly be at the high end of that range in the current environment. So I think that model holds. Long-term trends, though, we think are going to be very favorable as semiconductors are on a critical path. And certainly, we are on the critical path of enabling the semi industry. As these trends play out over time, more economic value increase in the electronic supply chain, and they do expect us to benefit over time. Is it going to be a next year dynamic? No, this is going to play out over multiple years. It's not a one-quarter journey, but I would expect there to be an upward bias in this dynamic over time based on the value of the innovation and enabling technology we and our customers are delivering to the global market.

Dan, we've run out of time. I really appreciate you joining. Thanks for supporting the conference, and take care.

Always good seeing you. Happy holidays.

Same to you.