Good morning, everyone, and good afternoon if you're in Asia. Welcome to this press conference of ASML. We are going to talk to you about the fourth quarter and full year 2022 financial results. We're here at the auditorium in Veldhoven at our headquarters, and I'm actually pretty happy to have a real audience here after two years of COVID. It's nice to have a few journalists that we've known for so long here in the room. There are also people watching and following this press conference online. They are also allowed to ask questions. There is a form on the website. Just fill that out, state your name and your publication name, and then send it in and we hope to come to your question. My name is Monique Mols.
I always forget to introduce myself. My name is Monique Mols. I am responsible for media relations at ASML. With me, I have Peter Wennink, CEO of ASML, and Roger Dassen, CFO of ASML, and they will talk you through our financial results, but also developments around our industry that I'm sure also has your interest. Let me invite Peter Wennink to the floor, and we'll kick this off.
Thank you, Monique. Welcome everybody, yes, you're right. I mean, it's good to have people live here, but also people on the webcast. I wanna just talk about a few things, key messages, I think more the industry, the state of the industry right now, then Roger will talk about the financials and the outlook, then we'll take the Q&A. Key messages. Well, in fact, it's what we said and what we told you at the Capital Markets Day. I think despite the fact that what we're currently seeing is, of course, a lot of turbulence, you know, is questions on are we going to get the recession, if so, how deep will it be and how long will it be? There are issues on geopolitical developments and export controls.
There's a lot of uncertainties right now. Our company and our industry really looks beyond the short term. I think we really look beyond the next 12 months, the next 2 years. We look 5 to 10 years ahead. Why is that? Because we're convinced that semiconductors are currently playing a critical role in the economy and in day-to-day life. It will only increase. It will only need more semiconductor compute and storage power going forward. This is why we strongly believe that we need to invest, and our customers feel the same way. They need to invest for the longer term. We have to remind you that building a semiconductor fab or a factory, we call it a fab, you know, it will take 3 years.
If you have to start planning it from day one, probably 5, yeah. This is all about the longer term, and when we look forward and look at the growth rates of, you know, semiconductor sales, you'll see it later, I'll talk about it later, that we see a very significant upside for our customers and also for ASML. This is why we need to invest. This is also why we believe there is an opportunity, depending on different market scenarios, a higher or lower market scenario, that by 2025, we'll be anywhere between EUR 30 billion and EUR 40 billion in sales, and we have between 54% and 56% gross margin, which will grow to 2030, depending on the scenario, between EUR 44 billion and EUR 60 billion, between 56% and 60% gross margin.
I think it's not only us, it's also throughout the supply chain, we're significantly investing together with our partners. Which also means that the company is going to be profitable. We are profitable. We're very profitable. Of course, because we will grow, that profitability will grow, and also for our shareholders, there's going to be a significant return. Now, when we look at the industry, three major drivers. I think it's still, we all know this, it's connectivity. It's connectivity whether it's a cloud or the edge computing, it's all connected. These things are not separate. It's not that there's no centralized computing in the cloud without the decentralized computing at the edge. It's all connected.
That's needed to drive artificial intelligence and the hyperconnectivity now being supported by 5G, but by the end of the decade, it will be 6G. It's going to be faster, lower latency, more bandwidth. There's another driver, it's called the climate change and the resource scarcity. I mean, semiconductors will, and I'm going to show you a couple of examples, going to be very significant building block of making sure that we can manage climate change, whether it's electrification, smart mobility, the energy transition, and of course, how are we going to use our resources smarter. Our last big driver is the social and the socioeconomic shifts.
I mean, yes, with the pandemic, have actually taught us to, you know, taught us to use different ways of working from home, but also, that, this remote connectivity is extremely important also in areas of medical technology, for instance. We'll talk about it later, technological sovereignty is a big driver for this socioeconomic shift. Okay. Just a couple of examples, just the energy transition. We believe it will be one of the main drivers the coming decades. It's about energy generation, it's about storage, about distribution, and it's about consumption of this electrical energy. If you know an example, if you want 1 megawatt of wind power, it needs about EUR 3,000 of semiconductors. 1 megawatt of solar power want EUR 4,000.
If you think about the gigantic increase in demand for electricity. The renewable energy coming from wind and from solar, just from wind and solar, just from generation, you can think about the enormous impact this will have on semiconductor consumption. I don't talk about the grid, about the smart grid that needs a lot of sensors and actuators and compute power. Yeah. And how do we use it? I mean, it's used in the electrification of our mobility. The, you know, the electrical vehicle will at least use twice the number of semiconductors as compared to a combustion engine. Yeah. I think the electrical vehicles is not our numbers.
It's just an, you know, estimate of one of our customers who is very, very clearly involved in the electrification of the automotive industry, said at least it's over $1,500 per vehicle, and some even say over $2,000. That's just one example. It's also people talk about advanced semiconductors and the importance of it and mature semiconductors as if these were two different worlds, but they aren't. This is a good example which comes from NXP. This is Lars Reger. He's the CTO of NXP Semiconductors. He actually showed this slide, and it shows that if you take a car, it consists of mature semiconductors. Here starts on left-hand side with 40 nm. It's just the actuators picking up the sensor data.
We're actually picking up the sensor data, a semiconductor technology that is even older than 40 nm, a lot older. Yeah. Then it needs to actually go to the central core of the car where it needs to be processed very fast. Yeah. High processing capacity because the input of that first analog and then digital data into the car is massive. The 5 nm chip doesn't work without the input. If the input cannot find a placeholder being in the central computing engine, it doesn't work either. It's all connected. Mature and advanced semiconductors are part of a systemic approach, yeah, to a problem which you will see with more increase in complexity that the approaches will be more systemic. We don't only need advanced semiconductors, but very much we also need more and more mature semiconductors.
Now, customers realize this. Just take an example. Our three major customers, Intel, TSMC, and Samsung, they have announced fab projects, investment projects. Well, I do the math for you. It's a $326 billion announced. $326 billion announced in new fab expansions. It's not only leading edge, it's also mature. That's needed. That's needed because of what I just talked about.
If we then look at the total semiconductor market, when you take all those developments and semiconductors are going to be pivotal in our day-to-day life, and we look at the growth rates that we're currently seeing and the new areas of growth in connectivity, in the energy transition, in the let's say, as a result of the socioeconomic shifts, it's not a surprise that when you do the math, it's not only us, but at different sources come to a semiconductor industry, the business of our customers, there will be anywhere between $1 trillion and $1.3 trillion, which is a doubling of where we are today. The semiconductor industry took 40 years, close to 50 years, to grow to $600 billion sales. It needs another 8 to 10 years to double it again. Yeah.
This is the challenge that we have. What do we need to do? We need to build capacity because we need to scale. Yeah. We need to find talent around the world. We're investing in our production capacity and in our EUV factory here, our Princeton factory here, our Deep UV capabilities here, but also in Berlin. We invest in Taiwan. We invest in Korea. We invest around the world to make sure that we can deal with this increase in capacity needs. Yeah. Our current build capacity, 375 Deep UV systems. Well, we want to have 600 by 2025, 2026. We have last year just, you know, a build capacity of around 50 units. This year, 60 units for EUV. We want to get to 90.
That's all needed to drive the growth of this industry, it also means we need to hire people. Now we just this morning, I told there's a news agency that over the last 12 months, we received about 300,000 applications for a job at ASML, which we hired close to 10,000 people, we still have a choice. We now employ close to 40,000 people worldwide, of which about half here in the Netherlands. That is going to be a constant challenge for us because we need to grow. We need to hire people. That's one. I need to onboard and make them effective. Big management challenge. What do we do then on ESG? It's very important. We launched our ESG strategy last year.
What do we do this year? For instance, we have a good example. We connected all the buildings here in Veldhoven in what we call a One Energy Grid, which basically means we find places where we can transfer waste heat, which is in one of the buildings. You could imagine this is waste heat in the factories, and we use it to heat the buildings. Ultimately, we will save 70% gas, in this site, which by the way you've seen is a big site.
70% gas reduction. That also means that we're significantly in the, in the social domain, significantly scaling up our investments in the community, in the infrastructure, to basically make sure that all people in the communities that we work in have as much as possible access to personal growth and can also benefit from the value that's being created. I think also from a governance point of view, we're one of the founding members of this Semiconductor Climate Consortium, which was actually launched a couple of months ago and was opened by Al Gore. I mean, we take our responsibility also as an industry. Now, geopolitics. I'm probably not going to get any questions on this, so that's why I'm going to preempt something. Basically, it's a clear issue.
It's driven by the fact that we all realize that semiconductors are now a pivotal element in creating a sustainable society. It's perhaps a bit mind-boggling that only now governments seem to realize that we have created, you could argue, some single-sided dependencies on certain countries on the planet because they basically have 80% of the world's manufacturing capacity. That means everybody's now rethinking this and says, "What do we need to do? We need to stay relevant. We don't need to do everything ourselves, but we need to stay relevant, so we need to push the development of semiconductor manufacturing in our region," whether the region is India or the United States or Europe or Japan. All governments are rethinking their semiconductor strategy. What are the consequences?
Over the last 40, 50 years, the semiconductor industry has consolidated itself into a global collaboration network, an ecosystem without borders, without boundaries. Which has great advantages because, you know, customers could talk with us and throughout the ecosystem very clearly about the roadmaps, about what needed to be done to drive innovation together. We have worked together with peers, yeah, with peers to just make sure that we created processes that customers could use and actually created a semiconductor industry that innovated faster, very and almost unlimited coordination with academic partners and of course, with suppliers in an open structure and open collaboration model to drive innovation. That's what we created and also by sharing risks and rewards.
Because of the, you could say, the bifurcation of the world into new socioeconomic blocks with a significant impact that has on the way they look at semiconductors, you see, of course, a bifurcation happening. One element of that is export controls, but that's not only it. I mean, it's the CHIPS Act around the world. The CHIPS Act that come with consequences, that comes with liabilities. If you want to get the money from the government, you have to give something in return. Which basically means that this world is indeed bifurcating. What does that mean if that happens? You see that the chip availability could be reduced.
Could be as a result of export controls that go too far and cut off certain parts of the chip manufacturing capacity, which will have an impact on significant industries like the car industry, like the energy transition, like the medical tech. Yeah. Supply chain might get disrupted because, you know, countries say, "Yes, you can get my money from the government, but these are the conditions." You know? You only need to spend the money where we think it is wise. Means that could put a lid on the collaboration with peers, with academia, because there will be borders. This borderless ecosystem, yeah, now is a danger of creating hurdles. It will also have impact on the supply chain. That was, I would say, seamless. Yeah. It will not be as seamless going forward as it is today. Yeah.
It also means that countries will double down on investing in their own industry, whether it's in the United States or in China or in Korea or in Europe. It also means that we will have a less efficient infrastructure. Costs will very likely go up. That's the risk that we are facing, but that's the result of, let's say, the realization that this ecosystem that we have built was almost seamless. It's now indeed bifurcating in different socio-economic blocks, each with their own, economic and political interest. That's where we are today. I'm happy to take some questions later on. Roger, why don't you do the following and talk about something boring like financials? Thank you.
I have one. This was a former CFO talking. How about that? It's a small step from U.S. politics to U.S. GAAP, as I said yesterday. That's why we're here. Small step for a man, a giant leap for mankind. Here we are, financials. Q4 results of 2022, strong quarter all in all, if you look at the numbers. The net sales came in, you know, at the midpoint approximately of our guidance at EUR 6.4 billion, as you see here. You know, based on EUR 4.7 billion of system sales and EUR 1.7 billion of installed base management sales. That was pretty strong. We saw that there were, you know, quite some customers that actually do want upgrades to their system in Q4.
Actually, a number of upgrades that were planned for Q1 have actually been pulled into Q4. That really beefed up the installed base number for Q4. Gross margin pretty strong at 51.5%. We guided 49%. There was some one-offs in there. Again, a major driver here is that same installed base management sales. There were quite some software upgrades in there. On software upgrades, the gross margin is very high. If you see installed base management go up the way it went in this quarter, that also beefs up the gross margin. That's what you see there.
Operating margin, 33%, and then goes all the way down to numbers that you see here as net income, earnings, et cetera. Important to note, the net bookings, you know, in spite of, you know, in spite of the climate, in spite of a number of our customers saying that they are looking at ways to reduce CapEx, it's very clear that we're still adding to our order book. If you look at the EUR 6.3 billion net bookings number that you have here, and you compare that to the first bullet where we talk about net system sales of EUR 4.7 billion, that tells you that we're still adding to the order book.
You know, if we look at the order book today, at over EUR 40 billion, you're in fact looking at, you know, more than twice the expected system sales for this year in the order book, which is very, very strong, I would say. Included in the EUR 6.3 as we have it here, EUR 3.4 billion for EUV. If we look at the full year, a number of highlights that we would like to emphasize and share with you. Of course, first off, the sales growth. Sales growth of 14%, for, you know, over 2021. Strong, you know, really strong demand, reflecting strong demand in all of our products. To EUR 21.2 billion, at 50.5% gross margin.
Pretty strong net income of 5.6 billion at an EPS of 14, over 14 EUR. Lithography, very strong once again, grew 12%, 7 billion EUR. Important to recognize what we have here. We recognize revenue for 40 tools, for 40 EUV tools. We actually shipped 54 tools. You know, I won't bother you with accounting rules all that much, but the essence is because of the high demand for EUV tools that we had in 2022, and, you know, some of the supply chain limitations that we had, we had to optimize the way we ship tools to our customers. That meant that, you know, some testing in the factory didn't happen.
As a result of that, we could only recognize revenue for 40 tools, but in essence, we shipped 54 tools. That delta of 14 tools is therefore not recognized in the growth number of 14% that you see for the full year, but that will come in the course of this year. Also important to note is that all five EUV customers actually have now placed High-NA orders, and I think that really, you know, emphasizes the confidence that our customers have in the High-NA technology that really is going to get us, you know, into the second half of this decade. There you will see that High-NA will play a pivotal role both in high-end logic and also in the memory business.
The DUV business continues to be very, very strong. On the back of some of the comments that Peter made in the mature business, but also in the advanced business, grew 13% to EUR 7.7 billion, as you see here. This really reflects the continued capacity growth that Peter also showed you in the earlier slides. Two new models were being introduced, the new ArF immersion system, the NXT 2100, and the advanced dry system, the NXT 870. In the application business, strong growth, 28% to EUR 660 million, really firing on all cylinders there. There you see that we have quite a few new products that we were able to launch.
Installed base business, we talked about that as a major driver for the growth in Q4. Very strong all in all at EUR 5.7 billion. You know, both based on service, but also on the demand from customers on upgrades to increase the productivity of the installed base that they already have. Finally, in terms of capital return, you know, everyone is sharing in our success, but also the shareholders are sharing in the success. In the aggregate with a combination of what we paid in terms of dividend and share buybacks, we were able to return to shareholders an amount of EUR 7.2 billion. Very quickly, in terms of the sales breakdown, you see technology-wise, not a real big difference.
You see that, you know, that the EUV, the, the wet business, RFI, et cetera, you see that it's more or less in line composition-wise with last year. You see that in terms of end use, actually memory is picking up, moving from 30% to 35%, which might be a surprise to some because, you know, some of our customers, are, you know, clearly indicating the memory market in 2022 was down, also prices were down. This really tells you that, you know, the investments that our customers make in the memory business are strategic. They're really looking, you know, not at the demand in Q4 of 2022 or even the demand in Q1 of 2023.
They really look at potential pickup in the second half of this year. They want to continue to make the investments. They really see lithography investment as critical on their strategic roadmap. That is reflected in, you know, in the pickup of the memory business. In terms of regional breakdown, it's more or less in line again with last year. In terms of unit numbers, you see it pick up from 309 units in total in 2021 to 345 in this year.
If we then quickly look at the composition in terms of end use, again, as you see here, memory, strong growth, logic continued to continued strong growth, and then you see this pickup of the installed base business that we already talked about. It's interesting to see the five years, how particularly the logic market, it has really, really picked up on the back of, you know, the end use developments that Peter has been talking about. If we then look at the boring numbers here, it's once again good to take the five-year perspective, and then you see that in fact, in this period, the company doubled in many different ways. We doubled in terms of in terms of sales.
Gross profit is picking up. You see a slight decline in the gross margin percentage for 2022 in comparison to 2021. There's a number of reasons, but the two primary reasons: first off, there is inflation. Inflation has been hitting us in 2022 as it has been hitting everyone. You know, not all of the inflation burden could be passed on within the ecosystem, so that's one driver. Secondly, and very importantly, we are investing in growth.
As you see if you look at the campus here, if you look at some of the slides that Peter was presenting, we're very much gearing up for further growth. It also means that in the gross margin, there is, you know, quite some money that is being spent on being ready for increased capacity in 2024, in 2025, and 2026. That means that, you know, that creates a bit of a burden on the gross margin. Nonetheless, if you take the longer-term perspective, very healthy development on that, on that front.
R&D also more than doubled, which I think is still a good sign because it really tells you that the company is still, you know, has still an abundance of ideas and is driving a very aggressive roadmap towards, you know, the targets that you've seen that we articulated at the Capital Markets Day. As the SG&A grows in line with the growth of the company and more or less doubled in comparison to 2018. The rest, you know, is just a translation of that into the bottom line. The last one, just to emphasize, is the very, very strong net bookings for the year at over EUR 30 billion already on the back of very, very strong order intake in 2021.
As I mentioned, that led to an order book of over EUR 40 billion. I mentioned the return to the shareholders. You know, we had very strong share buybacks last year. In terms of dividends, once again, we present to the AGM an increase in the dividend, an increase of 5.5% towards EUR 5.80 per ordinary share dividend over 2022, and that is the proposal. For the, you know, for in terms of share buyback, we in November introduced our new plan for EUR 12 billion of share buybacks over a 3-year period.
In terms of outlook, for the quarter, EUR 6.1 billion-EUR 6.5 billion in terms of net sales, which is more or less in line with what we had in Q4. Installed base a little bit lower because, as I mentioned to you, there was quite some pull-in from customers into Q4. That has a slightly negative effect on the installed base sales that we expect for Q1. That also has a slightly negative impact on the gross margin, although for the full year, we expect the gross margin to be slightly up in comparison to the gross margin that we had for the full year 2022.
R&D, once again, a step up, to EUR 965 million, and SG&A commensurate with the growth of the company to EUR 285. Those are the key parameters for Q1. If we look at the full year, you know, we are looking for 2023 at an over 25% increase in terms of sales. That really tells you that, you know, in spite of some of the, some of the signs that you get, it is very clear that our customers are really looking at investments into lithography as strategic investments.
While it's clearly, you can clearly recognize that there are some end markets that are under pressure, big end markets like PCs, smartphones, et cetera, you see that growth continues in other markets, like in the auto market, and I think the example that Peter just gave you both on electrification but also ADAS systems I think is a good example there. Clearly, there is some, you know, some pressure, I would say, on some of our customers and some of the growth of our customers, particularly in the first half of the year. Most of our customers, if you listen to their commentary, they expect the market to, you know, to regain strength in the second half of this year. The logic of our customers is pretty clear.
They say, "Okay, if the market, you know, within a 1-year timing is going to pick up, if the lead time that ASML has for ordering new tools is over that," as I just mentioned, it's on average about 2 years' time, "we better take our tools. If we now let go of our tools, that means that we're at the end of the queue, we have to wait 2 years before we get the tool, if indeed the market picks up within 1 year, we will be too late to cater to the pickup of the market." That's the reason why, you know, in spite of some of the pressure that some of our customers feel, we still see a very strong demand for ASML.
I mentioned the over EUR 40 billion in terms of backlog, very, very strong, a record number, there. Yes, we are, based on the increase in capacity that we're funding, we are looking at plans to ship 60 EUV tools this year and 375 DPV tools. That gets you significantly north of the 345 lithography tools in total EUV and DPV combined that we had for 2022. EUV business year-on-year growth, 40% is expected, 30% for the non-EUV business, and the installed base business, we expect a moderate 5%, and that is because we think most of our customers have done a lot of upgrades last year. We expect them to slow down a little bit on that in 2023.
With that, dear friends, I've come to the end of my presentation, and I hope Peter and Monique can join us for Q&A.
Great timing, Roger. I see we have some issues with our website, the connectivity is not too good. I think this is related to the general IT disturbances that are going on worldwide. Hoping to get an update in my ear, let's just start here. We're hoping to fix it as soon as we can.
It shows how important semiconductors are. We all rely on it every day, and when it doesn't.
Yeah, don't blame it on the semiconductors.
Oh, my God. Yeah.
We have a microphone here in the room. Kelsey Zegers will come to you if you indicated you have a question. Sander, my colleague, Sander Hofman, is checking the website to see if there's any questions coming in through the website. Let's start off with the first one. State your name and your publication number, Jeroen, we'd be happy to answer your question so people can hear it.
Thank you very much. It's Klaus Max Smolka of Frankfurter Allgemeine Zeitung in Germany.
Mm-hmm.
Could you elaborate a little on the communication between ASML and politics? I wonder, in political talks, in what way is ASML getting involved specifically when the Dutch prime minister has talks in the U.S., for example? How many ASML representatives will be part of the delegation, and in what way will they communicate with politicians? Thank you.
Well, it can be very short on the content of the delegation. I mean, this is geopolitical. It's not geo-business. I think there are no delegation members of ASML there because it's political. Again, you know, yes, we cannot deny that we are at the heart of a lot of political discussions because of the importance of semiconductors. I just explained it in my introductory comments. Our role is predominantly the role of making sure that people understand how this industry works. Now, until 2, 3 years ago, you could have asked the average politician and said, "Tell me how the semiconductor industry works." He wouldn't have no clue 'cause semiconductors were always there. It's like 1973 revisited. Oil was always there until it wasn't, and then it became a strategic commodity. Fast-forward, 2021, semiconductor's always been there.
Nobody needed to worry until they weren't, we figured out there was a worldwide dependence on only a few countries. This it's only very recent. Our role is education. It's basically providing information, data. How does this industry work? How is this global ecosystem connected? Where are the interfaces between those pieces, yeah? How does that work? If you think about a bifurcation, but also of the world and the world creating different sets of CHIPS Acts and think about export controls, our role is really to provide the information and the insight into what it means. What are the consequences of certain scenarios? That's our role, yeah?
Ultimately, this whole concept of national security and economic interests is something that is of a higher political nature that, of course, governments, each with their own views and desires, they need to discuss this amongst themselves what they want to do. Yeah? That's them. That's not us. This is our role. In your question, we're not part of the delegation, but we're definitely part of the discussion, yeah? Our role is to provide that information and that insight. Good question.
Okay. Thank you. I'm just getting an update that there's quite an issue with our website. We won't be able to get the questions in from people online. Sorry about that. We're available after the press conference, if you have any questions, want to reach out to us, please do to any of my team members. Next question here.
Hi. It's Pieter Haeck from POLITICO. I also had a question on export controls. We've seen last week there were discussions between Rutte and Biden. To what extent do you think that this should be a debate more on an EU level? Because I can imagine that there, if there are bans in place, that this would have a ripple effect to certain of your suppliers. What are your thoughts on that?
Yes, I think that's a good comment. I think our prime minister was pretty clear on that. He said, "Listen, this is a complex situation because it involves several countries." He didn't say it involves only the Netherlands and, you know, the United States. He referred to several countries with several companies in a complex supply chain, which, you know, we need to look at. The consequences of doing anything could be quite significant. He actually mentioned that. It's, it's sensitive. I don't remember he came out of it and said, "Oh, yeah, no, we struck a deal." I don't think he said that. I think we're still in discussion because of the complexity of that matter. You're absolutely right. I mean, it's what I tried to say.
This is an ecosystem that has been built over 40 years. Yeah, if it's seamless and you put locks into that system, you need to understand what the consequences will be. You know, you just ask the automotive customers, you know, how they think about the, you know, chip shortage. Try to still order a car, you'll find out that your waiting time is more than 12 months, perhaps even longer. I don't know for sure because of the chip shortage. It's all connected. This is why these are decisions that can have far-reaching consequences for the worldwide ecosystem and the capacity in that ecosystem and the flow of semiconductors in that ecosystem. This is why it's not between one or two countries.
Yeah, I think this is what the prime minister said, and I think he has a point.
Okay. Next question.Guys, you have to do this also for the people online. if you know anyone who's watching, open your chat and.
Joost Even from Volkskrant. You have dwelled on the many drawbacks of the geopolitical situation. To what extent do you sympathize or understand the concerns that the Americans have about China getting more ASML machines?
Well, I think that concern it is already there. As you know, our extreme ultraviolet machines, EUV, are under export control. That is a fact. Now, I've of course listened to the arguments, the arguments are very, you know, these chips will be then used into military applications. You have to realize that every time when I hear that, I think about the export control or proposed export control, it's about our advanced technology and being able to make advanced semiconductors, yeah? The military applications use rugged, very mature semiconductors. Yeah, the semiconductor is sometimes 10, 15, 20 years old, yeah? Which are widely available across the globe, also in China.
To fuel a significant amount of chips into a military complex, yeah, a military, you know, machinery, there's a significant amount of systems out there that are not advanced. They're mature, yeah? It's already there. Having said that, of course, there is artificial intelligence, there's smart algorithms. They need high power compute. Well, those chips are available. They're available in the world market, and they're made on ASML machines, on EUV machines. Not in China. They're made in Korea, in Taiwan, in the United States. Those chips are available for every end user and manufacturer on the planet to buy. It's also true for China. China buys, is the biggest importer of chips on the planet. I mean, American chip companies, I don't know, it's public data, just look at their financial statements.
Just add the five biggest chip manufacturers on the planet in the United States. They sell $80 billion worth of chips to China. It just shows you how complex this is, yeah? It's and those chips, those advanced chips, they go into anything, yeah? I don't know where, but they're being sold, so they probably end up somewhere, yeah? I think this is the complexity, and this is also where the reference to the complexity was made. It's highly sensitive, yeah? It's a much bigger and more complex situation than just saying, "Well, that chip goes into a military application." That's, it's far more complex than that.
Meanwhile, I have really good news.
Mm.
If those who are watching refresh the page, then it should work again. I have one question in through WhatsApp, but I saw someone wanting to ask a question there. Kelsey, can you give the microphone here?
Mike here.
Sure, yeah.
Could you say what the influence of a possible ban on just the DUVs would be? Oh, that's to sell DUVs to China. Yeah. I think that fully depends on what the deal is, you know? I think if they strike a deal, we can. Actually, we laid out on different scenarios what the potential consequences are. It's up to them to basically decide on the basis of that information. I cannot even give you that answer because I don't know what the outcome is, yeah? That fully depends, yeah? I'm not even willing to speculate on this because this is so sensitive. If I start speculating here, you know, I know what's going to happen. I know what you guys are going to write, yeah?
I'm not going to do that, yeah? It's too sensitive.
I just want to remind everyone that this is annual report, annual results. You are allowed to ask questions, so I don't want to answer.
I will answer those questions, yeah. You know, you're perfectly happy to answer anything.
Absolutely.
He was so absolutely clear about what it was.
Yes.
I can imagine that the question is going to a different direction.
Yeah, I have one here through the, through the web, so very bravely.
You should guide them, you know?
Yeah, yeah. So Eva Schouten from NU.nl bravely broke through the barriers, and she sent us a message. She says, "Peter noted that too much export control will lead to fewer chips. Could you define too much? Where do you draw the line?
That's exactly what I tried to say. It depends on how deep those export controls. I think export controls are a legitimate instrument, I think. That's clear, yeah? It's how do you want to use them.
Mm-hmm.
If you overuse them, it will have an effect on availability of mature chips, which are being used not into, in a military complex, but are used in a much larger, way larger industrial complex, yeah? Which is basically what we buy every day, you know, from cars to washing machines, to PCs, to smartphones, to infrastructure. We talk about the energy transition. The energy transition needs massive amounts of semiconductors. Massive, you know? I think we also need to be careful-
Yeah, we shouldn't block those.
that we don't-
Yeah
...that, you know, instruments that are well-meant, yeah, don't have a consequence that we all feel sorry about, huh?
Yeah. Yeah. That's also the precision that Marco had referred to.
Yeah.
I saw Toby. Yeah.
Toby.
Hi. Toby Sterling from Reuters. Not purely my question here, but It's financial, there you go.
Hey.
Just I think the general idea is, there's some skepticism about the, you know, the weakness in the chip markets not having much of an effect on ASML. I know you've explained it a bit.
Yeah.
Maybe you could say, I mean, there's such a big building program happening right now. Is there a chance that that could be slowed down by some of your major customers? Yeah. Could you break down your projections for DUV and EUV sales, not by systems, but by the number in total?
I think he probably put it on the slide, so we can refer to that for this year. You need to remember, we are not insulated. We're not from any macroeconomic events, and especially not when it's such a big inflation coming at us and, you know, and there's a recessionary fears, or some people even say some parts of the world are in a recession. Of course, that impacts our customers and will impact us. You have to remember that we come out of a period like 2022, where we could hardly ship 55% of what our customers wanted.
They want this because they know that these investments are long-term, so they, the 1.5 to 2 years of lead time, and then they need to build factories, and it's gonna take time. The base yet, we have this demand, which was significantly over our build capacity, which has come down. True. This is a reflection of what we see in the, you know, the market, but it's still quite a lot above our build capacity. That means our customers are actually saying, "Well, what I think the duration of this recession is significantly shorter than the lead time of your tools, and that also means that the recession will turn up into an upturn by the time I need your machines. Why would I cancel or push back those machines?
If I do that, and I'm wrong about the duration, it's shorter than I think, I'll be in the back of the line because these guys don't have enough capacity. This is where we are.
Yeah.
Maybe a naive follow-up, is there, when somebody falls out of line, does it just simply go to the very next customer?
Absolutely. You know, there is a shortage, so if you said, "I don't need your machine," then your neighbor says, "Thank you, thank you, thank you very much." That's gone.
Where's the next slot? Toby, in terms of numbers, for EUV, we're looking at 60 for this year and 375 units for DPV. Those are the numbers that we're envisaging for this year. I think it's important to note, is that the capacity that we are building, we're not building that capacity for the short term. I mean, the capacity for the 90 EUV and the 600 DPV, and that's the capacity that we're driving towards, of course, we drive that for the medium and for the longer term. Any cyclicality that might be there and demand might be there, but one thing is for sure, we will need that capacity.
Maybe we don't need it in full for 2025, but then we will need it in 2026 or in 2027.
On the customer capacity, you are right. I mean, we see a demand reduction, again, above our build capacity. Yes, I think customers are reshuffling, retiming some of their capacity investment. Sure. Yeah, I mean, that's what they do. I mean, the, in the memory industry, some of our customers have actually set out, said, "We're going to cut CapEx for this year." Yeah. Others have said, "No, we're not going to do that." In terms of this very strategic long-term, long lead time item called lithography, they're very cautious. If for whatever reason, the duration of the recession will extend to years, of course, we will, you know, see the effects. Yeah. It's not what they currently think. At least that's our conclusion.
This question was really simple. I mean, I could look at the average sell, sale prices, but selling prices for each type of machine. The question is, in dollar terms, is it about 50/50 the way it was in 2022 for 2023? Do you see what I'm saying? I can't multiply 375 times X.
You mean DPV versus versus EUV?
Yeah.
Yeah. I mean, we said that EUV is gonna grow a bit harder, faster.
Faster
than the non-EUV business. EUV, we expect to grow by 40% and the non-EUV business by 30%.
Yeah.
That gives you.
Yeah.
The shift.
Yeah
... a little bit more towards EUV, which makes sense. You know, it's most expensive, and that's where the capacity, a lot of the capacity expansions that I showed you, the $326 billion is about EUV-driven fabs. Yeah.
Thanks for that, Toby. I have a question in from your colleague, John. Chris Bryant from Bloomberg Opinion, is the drive for technological sovereignty, including export controls, more positive than negative for an equipment supplier like you?
Well, there are two sides to it. I think the positive side is it will create more, capital inefficiency.
Mm-hmm.
Capital inefficiency means high capital allocation, so it's good for equipment makers. That's good. The bad thing is that if you have more locks or hurdles into the system, you have more friction. Friction, the flips of friction is cost.
Mm-hmm.
Yes, you have a higher level of business, but it's not as efficient. It's also, that also is true for us. Yeah. It means that you will create higher cost. I think on balance, equipment companies will sell more machines.
Mm-hmm.
That's it. On balance, you could say it's positive.
Mm-hmm.
It's a bit selfish.
Yeah. I think so. Yeah. Hey, that was the question. Anyone here? Yeah, sure. Yeah.
If nobody else is, then I'll.
No, go ahead. Go ahead.
On the talks in the U.S., when do we expect a decision which would help you assess the commercial ramifications?
Yeah.
Thank you.
No, I think it's in the papers, and they're right about that. The politicians want to do this rather sooner than later, probably also driven by the fact that there's a lot of uncertainty created. The quicker you know where you are, the better it probably is. It's probably under high pressure, and I mean, that's also what I read. We'll see, you know. Given the complexity of this, I mean, you have to distinguish between some kind of general agreement and the details. How do you work out the details of these regulations and these laws?
Mm.
There's a lot of details involved, and working out the details will probably take significant amount of time. You don't talk days or weeks, I think you talk months, yeah, to really talk about the details of something like that, to really understand what does it mean, what are the ramifications of something like this, you know?
Mm.
That will always take, you know, time. I think we can safely say that the thing is under time pressure. Yeah.
Maybe as an example, the October export controls, I think everyone in the industry is still trying to figure out what the details are and how to deal with that.
Yeah. That's absolutely true. I think, until the end of this month, there is a, let's say, a, you know, feedback opportunity from...
Mm
... the industry and industry participants on what it actually means. I think, you know, final wording is still under, you know, review. This is how these things go, you know. To really say we fully understand all the ramifications, and we can translate that into business plans, I think it's going to take some time.
Okay. Question online. Computable: Do you expect any very long-term effect of China getting a position in manufacturing its own semiconductor machine equipment? How do you assess the probabilities of Huawei developing competing chip building machinery or probably any other party in...
Yeah
... anywhere in the world?
You cannot be naive on this, you know. When semiconductors are so pivotal and vital to a responsible economy and a responsible society, and you can't get hold of those semiconductors, you're going to do it yourself.
Mm-hmm.
That's clear, you know. I think this will happen. That's one of the consequences of, you know, not having this global fluid, seamless ecosystem. It's going to develop themselves. Having said that, I cannot speak for our peers in the semiconductor equipment industry, but for ASML in general, I think there's a lot of equipment in the industry that is focusing on a very complex process that is inside the machine. The machine itself is not that complex, yeah. For lithography, we make a relatively simple process. We just blow light through a lens, yeah. We do this in a very complex machine.
Mm-hmm.
Yeah. That's a bit different. This is why basically, rebuilding a lithography industry, which as it concerns to ASML, consists of hundreds of suppliers-
Mm
but each of them are world-class in what they do. I just need to name TRUMPF and ZEISS and VDL, and they're world-class in what they do. That's just three out of a couple of hundred, yeah. Now, you need to basically it's not patents, it's knowhow, it's people, it's the brains, yeah.
Mm.
You need to just take that and then redo all of that, yeah, to come to the same conclusion. It took us 40 years.
Mm.
Yeah. I'm not going to say that laws of physics are different in China or in the U.S. or in Korea or in Russia for that matter. No, laws of physics are all the same. The accumulated knowhow with hundreds of companies where we as a system architect, as a system integrator.
Mm
... bring it all together, that's a bit of a challenge. I think, yes, of course, the efforts will be there.
Mm-hmm.
It's going to be massive challenge going forward because it's not about ASML, it's about the ecosystem.
I believe the best defense line that we have there obviously is to keep on innovating, right?
Yeah.
I think that's critical as well.
Yeah
... sometimes you get the question, you know, with the market share that you have, you know, why don't you lay back and be a little slow? Well, we're not.
Mm
as you could see, we're investing very, very significantly.
Good point.
... on our R&D roadmap. We truly believe that pushing down the accelerator there is the best potential defense into anyone, you know, trying to get on par with our technology.
Yeah.
You know, it's good to be challenged sometimes, huh?
That's absolutely true. I know I had a discussion with the founder of TSMC on, you know, what do you do when you have to think about competition coming from different areas? He says, "Well, there's one solution. You just have to go faster.
That's it. Yeah.
That's it. This is also why, you know, you will see that, we will invest, to your point. It's not only in capacity, which will of course have a temporary-
Yeah
... effect on your gross margin.
Technology also. Yeah.
In R&D, we will spend.
Yeah
... more on R&D because we spend the R&D not for next year or for this year, we spend the R&D for the second half of the decade and the first half of the next decade. This is where the R&D goes, yeah.
John. John Koch from Bloomberg.
Yes. Jonathan Stearns from Bloomberg. What do you see as the biggest challenge and risk in the U.S. implementation of CHIPS and Science Act?
I think the biggest opportunity, let's talk about the opportunity first, yeah, is that you will build a part of the semiconductor ecosystem capacity inside the United States. It means that the dependence, the geopolitical strategic dependence will of course go down, yeah? That's why they do it. I think the risk is that both in the U.S. and in Europe for that matter, the experience and the knowledge that semiconductor makers have in bringing very advanced semiconductor manufacturing onto own soil is highly complex. It's an art, yeah? And that art is mastered over the last 10, 20 years by the TSMCs and the Samsungs of this world, and also Intel.
This is why you see the invitation almost, well, you know, it's an invitation with a big carrot, yeah? Which is called subsidy and investment support, yeah, of governments to bring those companies with that particular competence, yeah, within their borders. This is exactly what's happening. This is, you know, why we had the Arizona fab opening of TSMC. You know, ultimately 2 fabs, $40 billion of investments, that's Taiwanese knowledge and, you know-how. Bringing that into your country and then ramping this up in a way that you haven't been doing for the last 20 years also means that talent is a massive risk. You need to get the talent to actually do that. You cannot just, you know, draw everybody from Taiwan or from Korea to do that.
No, you need to create your talent pool, and that's a big challenge, yeah? I think that is a generic risk, not only to the U.S. chip sector, but also to the European chip sector, yeah? People.
That risk is also the single biggest opportunity because if you're able to do it, and if you're able to build an ecosystem around that, if you're able to make that, the knowledge and the expertise in that ecosystem, if you're able to really create that on soil again in Europe, in the United States, I think it's a major boost to your, to your competitiveness. It's the biggest risk. Are you able to attract that talent? Also, by far, the biggest opportunity. It's not just about, you know, creating capacity. It's creating that knowledge infrastructure that will help you know, through the next and next-
The ecosystem.
Yeah.
That's what That's the secret word, you know? You need to build those ecosystems 'cause if you look at where the biggest centers are of chip manufacturing, it's in Taiwan around Hsinchu and Tainan, and it's not only those factories, it's the suppliers, it's the universities, the knowledge institutions, the customers. It's the partners, the technology partners, it's the peers, yeah? That's all there. That's the ecosystem, yeah? You see, if you find the same in Seoul and in Hwaseong, yeah? if you find the same in Silicon Valley, but not manufacturing, but the rest is there, you know? You have the design companies there.
It's not just the money, throw the money at it, find a deserted air place, you know, Air Force base, and just build a factory. No, you have to build an ecosystem, yeah?
Okay. We have time for one last question. I have one online, but if anyone's here, up for it? Nope. We go to Switzerland. Neue Zürcher Zeitung. Oh, again, not a financial question.
The C word in it again.
Sorry. I tried. How do the U.S. restrictions on the involvement of U.S. persons in the Chinese semiconductor industry affect ASML's daily business, if at all?
Currently not.
No.
'Cause I mean, we have quite a significant workforce in China which is, you know, non-U.S.
Mm.
That's both Chinese nationals and people from outside the U.S., so that's not, that's not the case. Now, we are, we're a European company, yeah?
Mm-hmm.
I mean, so we have European know-how in, with people with European or with non-U.S. nationality, so that's less of an issue for us.
Yeah. One minute. Last call. Yeah.
One question.
Wouter.
Yeah, follow up on that. Are you still able to supply your Chinese customers with parts that are more American-made than European maybe? Because that's, I remember that.
Yeah
well, it,
That's not always possible, so that would have an effect on the operability of the installed base. Again, we're largely, you know, predominantly a European company with European technology. The parts are predominantly European. Here and there, yes, there are issues that we need to deal with, you know, which actually means that some of the tools will then stop working. It's minimal right now, but that would be the consequence.
Thanks.
Okay. Well, thank you everyone for being here. I'd like to conclude this call. Thank you everyone online for for being with us, and hopefully we'll see you next year. Thank you.
Thank you.
Thank you.
very much for your time.