Applied Materials: Sanding Atoms - [Business Breakdowns, EP.118]

23:13Back in 17, 18 timeframe, they bought a lot of stock back. Back in the 2008 bubble, they bought a lot of stock back at those timeframes. And the other beautiful thing is there hasn't been a time period, at least since the dot -com bubble, where for a two -year period, they didn't grow. You say, okay, year on year was negative this year because of the cycle. But if you take a two -year period, right? They've got this cyclicality sort of managed really well in terms of being able to take advantage of it, but also having the foresight to understand that cycle and know that it's going to be better soon enough, as soon as the cycle turns, but take advantage of it by buying back stock.

23:46In preparing for this conversation, I spent time in the 10K and couldn't help but notice that customer concentration here is more significant than you see in most companies. 20 or 30 % of their revenue is exposed to a handful of companies. A, can you just explain the relationship they have with those businesses with Samsung and TSMC and Intel and what that means for how their business is forecasted and how they think about their growth? Yeah. So the relationship between the equipment companies and the biggest fabs in the world is such a close one, right? TSMC spent $30 billion plus last year on CapEx, right?

24:24And of that 80 % of it was equipment related to manufacturing semiconductors. The rest would be actually building the fab, the shell that you put all this equipment in and piping around the gas infrastructure, all this sort of stuff. And Intel, same line of reasoning and Samsung, same line of reasoning. So you've got more or less three companies who are spending $30 billion a year on fabs to manufacture chips. The difficulty there is there's concentration on both ends, right? So if I want to buy a specific kind of equipment in general, there's one, two, maybe two, and very rarely three companies that can provide that kind of equipment.

24:56And even if there are one, two, or three types of companies, the process is so complex. In broader strokes, you could say, hey, there's three companies that do this. But when you narrow down to exactly the chemical reaction you're trying to induce at what nanometer precision and all this sort of stuff, you actually end up with, ah, there's two, maybe one, right? Companies in this field. And so when TSMC goes from five nanometer to three nanometer, they're introducing various technologies that help them do this, right? They're improving the lithography, of course, which they're working with ASML on, but they're also doing with every piece of equipment, they need to optimize certain things.

25:29And so for years, years ahead of before they start manufacturing a three nanometer chip, they're working with the equipment companies on, hey, well, we need to do this to do this to do this. And when you go up the chain, it's, hey, we actually need you to make this part of your tool better in this way, or hey, help us provide suggestions. How do you think we can do this? And so they work intimately together in what is called their research fabs. They'll have these massive fabs and you hear about Intel a lot in Arizona and in Ohio trying to build these fabs and TSMC in Arizona, right? And Samsung in Texas, at least the Americentric view.

26:03But in reality, all of these companies, Intel, Samsung, and TSMC have one research fab, right? Which is where they conduct all of their research and development. And in that research and development fab, applied materials and ASML and all the other competitors are deeply embedded with their equipment, their latest versions of the equipment, helping prove out and design and figure out how to manufacture the next generation of chips at high yields. As we mentioned again, the lifeblood of this industry is yields. Fine, I can make one or two chips of that kind, right? I mean, this is the whole story of Intel failing.

26:36Yes, they were able to make some chips with 10 nanometer, but they were never able to yield it, right? Whereas TSMC did 10 and 14, 10 and seven in the timeframe it took Intel to figure out 10 nanometers. And that's because TSMC worked really well on yields. They actually worked really closely with the equipment companies in a way Intel didn't. Intel got a bunch of hubris and thought they didn't need the equipment companies help as much. You ended up with this situation where the equipment companies really helped TSMC figure out how to do things and work closely with them, co -developed the tools together, right?

27:07Co -engineered them. So yeah, you have customer concentration, but that's also where the strength comes from, right? Because there's only a few customers for a big chunk of the equipment, and there's only a few suppliers, and each of the suppliers is specialized in certain ways. So you end up with this sort of business model where the companies accept that they make decent margins. And TSMC makes about a 50 % to 60 % gross margin. And Applied Materials makes about a 50 % gross margin. And ASML makes about a 50 % gross margin. And while there is always hopes of making higher margins, there's also a lot of mutual respect where it's like, okay, this is an acceptable, fair price for what you're doing and all the services you're providing.

27:45No one's really trying to do a race to the bottom because how complex this is, right? Especially when you drill down into this, right? Thousands of process steps, each of them are going to be different tuned. In any individual process step, there's only 50 or a hundred engineers in the whole world who know such a level. Maybe 15 of them are Intel, 15 at Samsung, 15 at TSMC, and then the rest at the equipment companies or something like that. So when you drill down very, very deeply, there is a huge amount of sort of, let's not say monopolism, but concentration in every regard of customers and suppliers.

28:17But that's also the strength of the business. And so you kind of loosely touched upon it, but if I were Intel or i170 and I wanted to switch away from applied materials equipment, is that even feasible? In certain segments, right? And this is why I think Applied Materials is maybe less appreciated than a company like ASML, because ASML is a very, very clear, this is 90 % of our business, which is lithography. It's one kind of equipment. We have a few versions of it, and each version is better than the last. And in the last generation stuff, there's two competitors, but in the new generation stuff, we're the only one.

28:51But it's a very clean story for people to look at and understand. With Applied Materials, you've got arguably six or seven franchises, right? Lithography is like 20 % of a fab spend of the equipment in a fab. But these other pieces that applied materials are sort of the monopoly in or oligopoly in, many of them are three, 5 % right here and there. So I mentioned CMP earlier, it's like 3 % of fab spend, but applied materials does 90 % of the business there. And so the question is, can I replace applied materials? Well, no, because if I actually look at what's called chemical mechanical planarization, you end up with this problem of like, hey, who's the other company?

29:26there's a company in Japan for this specific vertical. And in that vertical, where do they compete? Well, actually this company, their tools are really selling into this type of chip, right? And also this type of manufacturing. When I look at a CMP tool for five nanometer, actually applied materials is the only one really selling into it. So then the question is, if I wanted to rip and replace applied materials for this Japanese company, could I? Maybe, but let's think about the economics of that, right? So this is 3 % of my fab spend and I'm TSMC, I have 40 ,000 engineers of which maybe 10 ,000 are working on the next generation process step.

30:00But when I'm talking about just individual process steps, hey, I'm the master at this, I'm the master at that. You're actually looking at, hey, I only have like 20 or so people that are good at it. Now the question is, so for that 20 or so people, here's my area where I can use them, right? I could either spend them on A, figuring out how to use the inferior equipment from this other second tier supplier and replace applied materials, or B, I could figure out, Well, instead of that, maybe I want to spend them spending their time on improving the yields. Every chip has billions of transistors, right?

30:30An iPhone smartphone chip has tens of billions of transistors. And there's hundreds of chips of those on each wafer that I manufacture. Do I want to spend the time making sure I can replace this equipment or do I want to spend the time making it so Apple's really happy with me because I was able to get 5 % more chips per yielding, or I was able to improve the performance by 5 % because I improved the electrical characteristics. takes. So I'm going to spend those 20 engineers who are super focused on CMP and that's all they do. And that's all they know. I'm going to spend them on actually designing the next process node or improving the output or improving the performance.

31:02I'm not going to spend them on switching out equipment for second tier supplier who perhaps their equipment can't even work here, or I need to walk them through all of these years of engineering where at 14 nanometer, we use the equipment in this way. And for 10 nanometer, we modified it and we did this. And at seven nanometer, we modified it and we did this. And at five, we upgraded it and did this, right? So on and so forth. Do I want to lead them through all of these changes, handhold them? Because they weren't there. They weren't helping me. They weren't co -engineering. They weren't spending their 20 guys who are experts at this working on this problem.

31:33And funnily enough, right? When you look at even the areas where the equipment is more competitive, not necessarily CMP, but some of these other areas around etch and deposition, where it is a bit more competitive, it's actually still ends up being, well, Well, LAM Research, which is another competitor, their etch tool is better at this kind of etch with this kinds of chemistry. And then Applied Materials is better at this one. And so you have different teams working with them on improving that specific one, right? Like we're going to use it for this step, this tool from Applied Materials. And step B, we're going to use LAM Research.

32:01And then you kind of still have this hyper -specialization. And so if I kind of assess what the future of Applied Materials looks like, what are two of the three most important things you think that will contribute to their continued growth from here in where they have a leading position? One of them is these sort of industries, verticals where they're the monopoly, right? And so this is going to be areas like epitaxial growth, CMP, ion implant, RTP annealing, PVD. So these are some of the areas where applied materials is going to remain 70 % plus share. Then there's other areas where applied materials is really gearing up to fight their competitors, right?

32:38So one of them is they released a new tool, They call it pattern engineering. And the whole point is they're trying to fight with lithography in some types of process steps. Not all of them, right? But some areas they want to fight with lithography. And if that's successful, then they can take billions of revenue from ASML. That's a huge, huge opportunity. There's other areas that are related to, hey, as we continue shrinking, which is, hey, at three nanometer and at two nanometer and at 1 .4 nanometer, so on and so forth, there's going to be changes to the chip. We don't just shrink the transistors and go about our day.

33:09Actually, there's meaningful changes on changing the shape of the transistor, right? It's called gate all around. It's called nanowire, that sort of stuff. So there's a lot of tremendous opportunity and, hey, there's brand new process steps, right? Can we compete to insert our tool into that brand new process step? Can we co -engineer with the partners to do this? That's a sort of another really key thing to hope for our watch in applied materials. But then the third is just continuing to execute, right? Because you can try and go super deep into what exactly each of these is and how it's going, but you can also zoom out and say they've just consistently grown, right?

33:43Because the industry consistently grows and semiconductor industry is consistently growing above GDP. The equipment industry is consistently growing above semiconductor and then applied materials is consistently growing above that. The dark side or potential negative is they don't, but most likely they do, especially with the increasing politicization of the industry. The industry is so politically important. right? And so you see this in the form of subsidies, right? Yes, there's the stick that the governments are waving around, but there's also the carrot, right? Which is, hey, the US has $52 billion in the CHIPS Act plus a tax credit, which is estimated to be something like $30 billion.

34:20And that tax credit is specifically on equipment purchases, right? Japan is spending something like $30 to $40 billion. Europe is spending something like $40 billion. China, it's harder to sum up their tax breaks because a lot of it's in subsidized loans. Estimates I've seen are something like $200 to $250 billion on the semiconductor industry. You're looking at hundreds of billions of dollars that are flowing in to the semiconductor industry. And all of this eventually flows to the picks and shovels, which is applied materials. That's the viewpoint you could take with a lot of this geopolitical tension.

34:53It's actually tremendously positive for applied materials and their business. And while the industry is going through a bit of a downturn and there's going to be some cyclicality, these subsidies sort of help make it less cyclical from the perspective of applied materials, right? Because now instead of cutting back because orders are weak, it's like we're going to cut back because orders are weak, but oh, hey, we have all these subsidy dollars that we're going to spend anyways. And so if I kind of look across the semi -cap industry, A, how many billions of dollars are we talking about? And then at a high level, what do kind of the market share nuances look like?

35:26or how do you think about market share when you're assessing whether someone is gaining or losing share? In the semiconductor fabrication industry, the fabs such as TSMC, Intel, Samsung, Micron, SK Hynix, the list goes on and on, spent about $100 billion, a little bit less last year. They're going to spend a little bit less than $100 billion again this year. The projections are they're going to spend less than $100 billion, but a little bit less than $100 billion next year, right? So somewhere in that $90 billion range, just because there's some macroeconomic issues, but in the past, they used to have been growing from 50 to 70 to 90.

36:00So they're going to spend that amount of equipment. And then you can sort of divvy it up, right? You can either say, A, hey, Applied Materials sells 20 % and ASML sells 20 % and Lamb Research sells 15%. You can view it like that. You could take the other view, which is actually ASML and Applied Materials do not compete with each other at all. There's a few small niches, but for the most part, 90 % plus of the business does not compete with each other, even though they're the two biggest companies in the space. Why? Because they don't make equipment that's the same. They make completely different equipment.

36:31And these different pieces of equipment are used for different types of processes. So in reality, you can actually dive down and look at market share in a completely different way. This $100 billion pie, more or less, is split into how much are we spending on lithography? How much are we spending on deposition? How much are we spending on etch? of that deposition spend? How much are we spending on physical vapor deposition? How much are we spending on sputtering? How much are we spending on chemical vapor deposition? How much are we spending on atomic layer deposition? Then you could go even deeper like, okay, well, just for atomic layer deposition, is it thermal or is it plasma?

37:05So you can spend a lot of time going deeper and deeper and deeper and breaking out the market to where you're finally at. Okay. This is the segment where two companies actually do compete and it's a billion dollar segment, right? And this company has 800 million and this company has 200 million. You can go that deep and really splice up the industry, which is the sort of stuff we do. But in general, how to assess market shares, unless you're going that deep, it's easier to just say, hey, look at the revenue of the company, remove out the services business, because the services business, while it is 25, 30 % of the company's revenue, it's not the CapEx, not the wafer fabrication equipment, it's services.

37:39And then look at where that is. How is that growing or shrinking across the industry? And in reality, it ends up being quite stable. Every once in a while, you have a technology inflection, like with ASML and EUV lithography, where it goes from 18 % to 22%. So pretty big shift of total spend. And obviously the rest of the industry does lose a little bit. But at the end of the day, that spend change is still minuscule compared to long -term applied materials has been 20 % and they remain about 20%. They've grown a little bit as consolidation has happened in the industry, but that's sort of a better way to look at market share.

38:14And as I was explaining earlier, it's so complex. These process steps are so different that you don't actually end up with so much competition and so much market share. You've done a great job outlining what this business does well and why it's well positioned for the future. But if I were to speak to an applied material skeptic or a bear, what would they be cautious about in regard to the company's future. Where are the significant risks in this story? The biggest risk that faces applied materials is definitely the China risk. And now the China risk is multifold. One is just Chinese companies stop buying applied materials tools, which is something that even the US government is assisting in, in some regards.

38:56But the more concerning thing is that the Chinese equipment market is developing and it's developing rapidly. There's still nowhere close to what applied materials can do in many areas, but there are some tools that the Chinese equipment market is starting to be able to produce. There's numerous Chinese equipment companies, Nara and ACMR, and the list goes on and on of companies attempting to make equipment. Some are not successful at all yet, especially in certain places like Litho and Dryatch, but there's other places where they are getting significant market share, things like cleans. And so these segments, it's all about just tackling sub -segment, being good, being much cheaper, and having the companies switch over to that tool and you lose out some market share.

39:42And then a lot of the new fabs, 30 % or 25 % are being built in China anyways. So there's a lot of switching that could happen if these companies get up to snuff and actually can compete with applied materials and land research and so on and so forth. And so this is the biggest risk for applied materials is that there is a switching, that true competitors do get built up by all the subsidies that China is pushing and all the domestic nationalization that China is pushing. And they haven't seen a new competitor in decades. It's been an oligopoly of sorts where everyone's kept pricing power and no one's chased margins down.

40:16But what happens if there's a competitor who sort of is, let's say, irrational by the Western sense of competition, completely rational to the Chinese with subsidies chasing the market down and trying to cut costs and be much cheaper to get market share before they even try and make a profit? And so we've somehow gotten this deep into the conversation without mentioning artificial intelligence. So I commend both of us for that. But question being, this shift towards the importance of semiconductors in relation to the compute demands of AI, are there implications for applied materials that are important to consider?

40:54Yeah. So I think one interesting thing, just to note about the equipment companies, is they see things earlier than the rest of the industry, in fact. If you look back at old investor days from applied materials, they were talking about the PC era pretty early because guess what? Intel needed to place orders for the PC era. They saw that firsthand. For the mobile era, same way. They saw that they were saying this sort of stuff. And then for the era of AI, they were talking about AI. Obviously, everybody was talking about AI, but they were saying this is like literally the next era of the semiconductor industry, right?

41:23You cannot build an NVIDIA AI chip without applied materials equipment. They have all this equipment and it's required. You can't make a chip. And with AI, 7 nanometer, 5 nanometer, 3 nanometer, these transitions are increasingly important. For a bit of times, there was a bit of a bearish view maybe a year or two ago, where if you were talking about like, hey, mobile kind of stopped. Mobile doesn't grow anymore. And Apple doesn't have the ability to continue to increase pricing on their end customers. And obviously Android just follows the route and PC stopped growing. So what do we need 3 nanometer for?

41:54What do we need 2 nanometer for? But then it's like, oh, all of a sudden everybody's like, yeah, we need 3 nanometer and 2 nanometer because these models are so complex to run and so hard to run that going and shrinking is going to provide so much benefit. And because of that, it's like 30 % more process steps when you go from five to three nanometers, right? And that 30 % more process steps also involves some modification of the equipment as well. So you're looking at generative AI being tremendously positive to companies like Applied Materials because, hey, we need a lot more semiconductors, right?

42:22We need more memory, we need more processors, we need more all of this, and it's all manufactured with Applied Materials as well as their peers' equipment. And so our concluding question is typically lessons that can be borrowed from the applied material stories and applied to other companies from both an investor perspective and then an operator perspective. What are kind of those key lessons for you when you look at this business across everyone else in your very, very wide range of coverage across semiconductors? Yes. I think one of the beautiful things about applied materials is, hey, they are in a cyclical industry and one of the most cyclical industries, right?

43:02We're talking about 35 % drawdown in revenue. Some of the times historically, they didn't let that shake them, right? They always took the long -term view. They always continued to invest in R &D. Sometimes it was 15 % of R &D, but in these drawdowns, hey, it shot up to 20 % plus. They took that long -term view. They said, hey, I need to work closely with my customers because right now it sucks for both of us, but let's continue to work very deeply on the R &D. And when it comes back, we'll be entrenched more solidly, right? They'll need to use us more and more and they'll understand that we're a long -term partner.

43:35We are not just a opportunistic company. And so that long -term partner is incredibly important, right? As a company in a space that's concentrated, even if you aren't concentrated, this being a long -term partner helps them with the rest of their business, which is, hey, we're not building the most advanced fab. We're not selling to Samsung. We're not selling to Intel. We're not selling to TSMC. Actually, we're selling to this other company, right? Like Texas Instruments, or we're selling to NXP on Semiconductor and all these other fabs out there that are not making the most advanced chip. Well, all of that fed back in.

44:05Now we can do better and better there as well. We can help them in so many ways because we have access to this. So that's the other flip side of the coin is if you have concentrated customers constantly investing in R &D, even when business looks tough, knowing that there is a light on the end of the tunnel and maybe that's a bit of a risk, but also using that to feed back to the less concentrated customers. Hey, being able to use that R &D to help them and also the services. There's two ways to look at the services business, which is I used to sell you equipment and I hate that and I want a subscription.

44:35Or two, I can look at it as, hey, I want to actually make it so you're getting the most output out of your equipment, which is really what Applied Materials is doing. They're saying, hey, this is our equipment, buy it. But also let's make sure you get every single chip you can out of that. Let's make sure you can utilize it as much as possible because that's going to make you more profitable. And that's just going to make you want to invest more in our equipment. That's the other flip side of the coin. And especially with the smaller customers who don't have the capability of servicing nearly as much, this is incredibly important, right?

45:04TSMC will have some more of their own technicians on site that are experts on specific kinds of tool. But what about the smaller customers? They don't have that capability, but because you're offering it to them, you're helping them be a better business. So I think that's also important is this commitment to your customer, your end customer is really important. Well, this is a fascinating business that I know you can speak for hours about. I think you did a great job summarizing the basis of the business and its competitive advantages going forward. Thanks for joining us. Thank you for having me.

45:34I had a blast. To find more episodes of Breakdowns ranging from Costco to Visa to Moderna or to sign up for our weekly summary check out JoinColossus .com That's J -O -I -N -C -O -L -O -S -S -U -S dot com