Welcome, everybody, to the press conference of ASML. We are going to talk you through the results of Q4 2023, and the full year 2023. As usual, I'm not doing this all by myself. I have got, I will have the company of Roger Dassen, our CFO, and Peter Wennink, our CEO. Let me talk you quickly through the program. You've seen our press release, and by the way, my name is Monique Mols. I'm head of media relations. Peter will start by explaining industry developments and our progress, and then Roger will take over and talk you through our financial results of this year, and then we have plenty of opportunity for Q&A. We also have people online. Thank you for joining, and they will have the opportunity also to ask questions.
There's a little screen where you can submit your question. Please also state very clearly your name and the publication you work for. The Q&A session is only for media, and of course, we've got people in the room who, of course, are allowed to ask questions. Thank you for coming all the way to Veldhoven. For some, there was a little bit more of a journey than for others. But let's start, let's start with our presentation, and we'll come back to you later with the Q&A. Peter, if I can invite you on stage?
Absolutely.
Oh, you have this.
Good morning, and good afternoon, or good evening, or wherever you are. First of all, I'd like to start off with some key messages, but, you know, it's not going to be a big surprise. I think this industry, and ASML specifically, is driven by megatrends, and I said it on several occasions. The major societal transitions that we're witnessing these days, whether it's the climate change, the energy transition, the digital transition, the world population that's getting older and needs more care, these are major transitions that don't happen once in a lifetime. They happen once in several lifetimes. And I think we need to realize that. And those megatrends, they need solutions, and I think it...
No matter how you look at it, every part of those solutions will be at the core. There's this semiconductor. There's this either compute power or data, or the artificial intelligence that we need to actually solve those problems. I mean, it, it will happen, and we need it. And lithography has been always, [audio distortion] by the way, semiconductors are driven by Moore's Law. It's a law of scaling. It's an empirical law of economics, and the scaling engines are one of the biggest and probably the most important scaling engine has been lithography. Basically, they're making the structure smaller. So lithography intensity will go up, when we need to innovate to actually address the challenges that come out of these big societal changes, and we have the product portfolio to do it.
I mean, I'll show you later on, we've always had a product portfolio to do it, but the complexity of our challenges, and thereby the complexity of the machines, the lithography machines, is also increasing. But we have that complete portfolio. It's not only lithography, it's also metrology and inspection systems. It's software. It's the integration of our capabilities, our holistic portfolio, that will enable our customers to deal with those challenges. And, and, and based on that, we do believe, strongly believe, that what we've said a few years ago, that we believe that by next year, our sales will be anywhere between EUR 30 billion and EUR 40 billion, with a gross margin of anywhere between 54%-56%.
But by the end of the decade, it's going to be anywhere, depending on, I would say, macroeconomic growth profiles, anywhere between EUR 44 billion and EUR 60 billion, with margins between 56% and 60%. And it's not only us, it's our task, because we're a system integrator, we're a system architect, that our supply chain and our partners will follow, and they, and they will do that. But everything in the context of our responsibility towards society, our ESG targets have been very clearly explained last year, and we'll publish our annual report, and you can see our progress going forward. So of course, in the end, we're a business. We'll make money, and we'll return cash to our shareholders, and we've always done, and Roger will talk about that a bit later. Now, I won't dwell on this. Market expectations remain unchanged.
Market research firms all about the end of the decade, about $1 trillion, which is just under a doubling of the business for our customers as we see it today. Now, and it's these trends that are really driving this. You know, it's the connected world that is a significant driver for our business and the business of our customers, and of course, it's the infrastructure, it's the cloud infrastructure, but increasingly, it's AI. And you're only seeing that this is not just a promise for the future, it's happening today. You know, we've seen this in 2023, and NVIDIA was very clear on what their capabilities are and showing those capabilities. We've all seen the rise of, you know, generative AI. It's gonna happen, but what it needs is more compute power.
What it needs is more data storage and the link between the data storage and the compute power. And we're only seeing that in our order book. Our order book for Q4 is evidence of this. I mean, significant EUV order book was also driven by the fact that memory customers came back. Why? Because there's a bottleneck now. We cannot access the full power of the compute power in the, in the logic chip, for the simple reason that the bottleneck of of performance memory, performance DRAM, is there. So how do you solve it? Yeah, you know basically create high bandwidth memory. How do you do that? EUV. It's that simple. So all these-- and what you will see is that AI, currently, we see it in the data centers, will move to the edge. You will see AI coming back into your phone.
Now, the applications might not be there, but that's always how it goes. You first create the technology, and if the technology is there, the applications will follow. Now, then, when we look at the other blocks, climate change and resource scarcity, I mean, the energy transition is not going to happen without a further electrification of society, which needs semiconductors. And smarter use of those limited resources, the smart grid, is all semiconductors. And also in the social and economic domain, when you think about life sciences, there's an elderly population. I mean, we are. You know, the demographics are against us in that sense, so we need more technology. It's all semiconductors. Now, and if you then look at those three blocks, you could roughly say that the advanced stuff, advanced semiconductors, more on the side of the connected world.
But the climate change, resource scarcity, and the social and economic shifts that need to be supported with semiconductors is more mature semiconductors. So when you look at this, and everybody talks about AI and about advanced semiconductors, about sub-2nm driven by High-NA EUV, yes, we will—that will happen. But the vast majority, the square centimeters, the square inches of silicon, will be driven by the other part, yeah? Which is where the massive application and just in unit terms is. So this is, it's an important slide, not only just to explain you the trends, which we intuitively all know, but also has a translation directly into the semiconductor world.
It's our world of advanced semiconductor manufacturing and mature semiconductor manufacturing, which I think is gonna be a significant driver of our business, both of them, advanced and mature. Now, clearly, we're experiencing a cycle. We're used to deal with cycles. I mean, that's what the semiconductor industry is all about. I mean. If you look at these big cycles driven by these macroeconomic shocks, we can go back more than 20 years ago, where we came out of the Y2K era, and we had the dot-com implosion and created a macroeconomic shock. And these downturns last two, two and a half, perhaps three years, 20 years ago. Now, we had another shock, which was the financial crisis, less than two years.
Now, we have coming out of COVID, after a decade of quantitative easing, where money was free, you know, but of course, driving inflation, interest rates going up and creating a macroeconomic situation of today, which is also a macroeconomic shock, 2-3 years, max. Well, the downturn started mid-2022, 2023 was clearly a downturn year, first half of 2024, we're probably in that same coming out of that, two years, two and a half years. So what about then 2025? 2025, in my opinion, just look at the, look at the cycles. It's going to be a, it's a full upswing year. It's gonna be a good year. So we need to prepare 2025 by doing the right things in 2024, and Roger will also talk about that. Now, on top of that, 2025 being an upcycle year, we'll have this: Our customers are investing.
Across the globe, more than 20 new fab, fabs, factories, fab projects, starting in 2024, taking tools in 2024 towards in the course of this year, but definitely in 2025, where fabs will be opened, but the fabs were opened in 2024, will take machines in 2025. It's all driven by geopolitical movements. You know, the world started to realize that dependence on semiconductors is, to a certain extent, good, but not too much dependent. So you see CHIPS Act popping up around the world, you know, in, ranging from the U.S., Europe, India, you know, China, Korea, everywhere. And that creates a situation whereby new fabs will be built and will be opened, and those fabs need machines.
That's the second layer of why we believe 2025 is gonna be a very strong year, and we need to prepare 2024 to be able to meet that demand in 2025. And can some of that be pulled into 2024? Potentially. It depends on the slope of the recovery, but we'll see. Currently, it's too early to say that, so that's why we stay relatively conservative. But this is how we look at the world. We look at 2024, 2025 as a transition period into new areas of growth. For growth, we need people. Now, we hired people, 3,000 people last year, 144 nationalities, and for the first time, we can say that we're in the 20% range for our female population, for the first time.
Now, we have actually hired, of all the people that we hired, 27% women in 2023, which is good. We're getting more and more, we're becoming more and more attractive, you know, for the female part of the world population to work at a high-tech company and bring their knowledge and their experience in science, technology, engineering, and math studies, bring that to ASML. Also in the Netherlands, you know, we have, of the 42,000 people, 22,000 are in the Netherlands and 123 nationalities. Now, we have a discussion in the Netherlands on the international student population, whether we need all those international students. Trust me, we need them. Now, there's more than that. It's also ESG. You know, ESG, I talked about it, is what we did in the environment.
While we are building and expanding our production capacity, where every year we're reducing our CO2 footprint in absolute terms. We actually, the energy efficiency of EUV as compared to 2019 improved with close to 40% in 2023. On the social part, we're investing in the community. You know, we can talk about the need for science, technology, engineering, and math-educated students, but you have to do something about it, and that's what we do. We put more money than ever now, but also going forward into the social domain and into our, you know, communities. And do our people appreciate that? Yes, our engagement score keep going up every year while we're adding thousands of people.
That's not easy because all those people need to land, they need to feel at home, but engagement scores going up is a very good sign that we're doing the right thing. Our attrition rate drops worldwide to 4% in 2023, which is low by any standard. And of course, I was already proud to, you know, say that now we're hiring 27% of our intake is female. And on governance, transparency, and has been recognized by many institutions. Now, High NA, that is the shipped, that's the tool that we shipped in at the end of the year. Very proud, shipped it to one of our major customers, Intel. It arrived in the United States, and it's going to be a workhorse.
It's going to be a workhorse the second half of the decade into the, well, into the next decade, yeah, and to make sure that we can actually execute on that promise of artificial intelligence, but also the promise of solving some of these big societal problems with the help of semiconductors. You know, something I would like to just share with you, I mean, April first this year, we celebrate our 40th anniversary, and we started in 1984 with a system, a PAS system, as you can see. Just look at the size of the, of the human next to the machine, and then you look at the size of the humans or size of the machine next to the humans today. I mean, it's a bit strange.
Over 40 years' time, you know, we've increased the size of the machines while we reduced the size of the semiconductors. So the smaller the feature size, the bigger the machine. That's what you see as a clear evidence. But also, we've always been able to provide the market with what they need, yeah? And we will keep doing that, also going forward, and the High NA tool is just the next step in that evolution, and we're very proud to do that. Now, very quickly, is there a new machine? And this is for people who love technology. You know, we enable 1.7 times smaller features, yeah, so which means almost a three times increase in transistor density on the chips. And how do we do that? We have a new lens.
We have new stages that work so much faster, 2-4 times faster. We have repositioned the EUV source so that we have better imaging, we have more transmission, we have more light on the wafer, new frames, improved metrology. It's a continuous battle for the next level of innovation, and we need, for the next level of innovation, the help of our customers, with the help of our people, the help of our suppliers, the help of our shareholders, and help of the community. That should enable us to create an environment where we can be super successful. And with that, I have 10 seconds left, so it's up to you, Roger. Thank you.
Thank you, Peter. I have the clicker.
You have it?
Thank you, Peter. Good morning, everyone, and I will use those 10 seconds wisely, obviously, right? So, let me, let me take you through the financial accomplishments and also the outlook that we have on the financial side against the backdrop of all the developments that Peter just described. So this is the quarter. The quarter, well, there's a lot we can say about the quarter. We ended a little bit better than we guided, so we always give a guidance with a certain bandwidth, and we were able to arrive a little bit better than the, than the higher level of the bandwidth. So that's always good. Primarily driven by the fact that actually the installed base business was better than we anticipated.
You see it here at 1.6. We expected around 1.3, so that was a major driver why the quarter came in better than we, than we anticipated. Also, it boosted the gross margin. Gross margin also a little bit better than we guided, so, you know, all good. But I have to say, the one number that everyone was talking about this morning when we had our initial conversations with the analysts, what was nothing of the first ones, it was the one at the bottom, which is the net bookings. So the fact that we were able to report net bookings, so new orders coming in for the quarter at EUR 9.2 billion, which really is a record number, that was the one that really attracted the attention.
And that, you know, in combination with some of the more positive color and context that also came, for instance, from the earnings release from the TSMC last week. I think it does make people realize that, you know, indeed, the whole industry is working its way through the, through the cycle, and, you know, these are positive first signs. And that's really, I think, also the way this was appreciated, also earlier today, by the analysts. If we look at the full year, in fact, what you see is that the company grew 30%. And just reflect a little bit on that, as we just said, you know, the industry is actually, was actually going through a down cycle.
2023 was not a great year for the, for the industry. If you look at most of our customers, they would actually report significant year-on-year declines in 2023. So the fact that we were able to grow 30% is actually a pretty good achievement, and I think it also, you know, tells you a significant story about, you know, the significance of what we do in the entire ecosystem. So good achievement, and I think kudos to the entire organization and all the 42,000 people that have been working extremely hard to actually be able to make this happen. So we grew to a net sales level of EUR 27.6 billion, as you see here.
A good growth margin, which improved from 2020- 2022, and also a very solid net income of nearly EUR 8 billion. Growth primarily on the, on the lithography systems, as you see here. So EUV grew 30%, and you just saw the first modules of the High NA leaving, leaving our factory and going at, going at Intel. So I think that is a major major achievement and a major landmark, I think, for ASML to see that happening. On DUV, growth even more, 60%, very significant part of that.
Also, immersion, as you see here, 32% of the unit numbers were immersion, so that is a pretty significant growth just there. And also there are some good product innovation with the 1980Fi immersion system being shipped to customers, and also a brand-new i-line system that found its way into customers. So a lot of innovation, not just on EUV, but also a lot of innovation going on on DUV. On the application business, actually a decline, and you might wonder, you know, how is that at all possible? If you grow so strong on the lithography business, how can it be that metrology and inspection is lower? And the answer is very simple.
The lead time, so the order lead time on metrology and inspection, is very short. So, you know, if customers are going through a downturn, are going through a down cycle, and they have to somehow save money, you know, because they have to manage their, their CapEx well, they're gonna look at the stuff, you know, "What is the stuff that I can wait and wait and wait, you know, a couple of months, because I know if I order it, I get it quite rapidly?" If you wait and wait and wait with ordering an EUV machine, you're in trouble, right? Because, because their order lead times could be, like, 1.5 years.
But for metrology inspection, you're talking about weeks or a few months, so there, actually, customers can afford to wait ordering and receiving the stuff. So that's the reason why application business is down. In spite of that, you know, we continue to innovate there, and you also see there the introduction of our new YieldStar 500 system. Installed base business, a little bit the same, right? So, if you're running at fairly low utilization levels, as many of our customers were last year, then you don't really need a lot of upgrades.
You don't really need a lot of productivity upgrades, and as a result of that, there too, you see customers take the foot off the accelerator a little bit and being a bit more frugal and economical with ordering upgrades. So that's the reason why installed base business this year a little bit down in comparison to last year, off what I would still call a fairly healthy base. Capital return, we returned back to shareholders EUR 3.3 billion. That is actually quite a bit lower than we did in previous years. Particularly share buyback at around EUR 1 billion cash this year was quite a bit lower than in subsequent years. Main reason, the same.
I mean, in a climate that we have right now, you know, with, or that at least we had in 2023, with customers going through, through declines in their sales and their, and their margin levels, we're actually helping customers. So we were, helping customers by giving them extended payment terms, et cetera, such that they could pay us, pay us later, and could push the payments out of, 2023 into 2024. So as a result of that, we used, our cash position, we used our liquidity to really help the ecosystem, first and foremost, our customers, but also the supply chain, to make sure that everyone could continue in spite of the fact that the industry as a whole, that was going down a bit.
That's the way we do it. We first use the money in the business, you know, where the key stakeholders need it, and then whatever is left, that is being distributed to, you know, to our shareholders. Breakdown of systems in the different ways. You see that technology on the technology side, you see the immersion technology being, you know, really very popular this year at 41%, so a big increase in there, but you also see EUV still strong. So you see the leading technology really being the technology that was in very high demand, and that, again, is in line with the way I just commented it. These are...
The leading tools are the tools where you have the longest lead time, the longest order time, so that's the stuff that customers will continue to order even if the industry is going through a bit of a down, down cycle. So very strong on the, on the leading technology. In terms of end use, you see that logic was extremely strong, this, this year, but as Peter said, we do see memory coming back right now. But memory was really going through a very significant down cycle in 2020- 2023. And then finally, on a regional basis, you see the spread there, and obviously you also see that China was, was quite big in comparison to, to last year.
And you know, we've talked about this before, the orders that or the shipments that we had into China were primarily executing on orders that, to a very large extent, were already there by the end of 2022. So China had, you know, had already put in quite some orders in previous years, but the order fill rate for China was low because we had such demand also outside of China, that we could only allocate less than 50%, around 40% to China. So, you know, of the orders that they put in, we could only deliver somewhere between 40% and 50% of those orders for a couple of years.
This year, with you know the slowdown in demand that we had for a few other customers, as a result of that, we were able to ship more to China than in previous years, and I think that is clearly represented in these, in these numbers. I think here you see the same picture. You see that, in spite of the memory down cycle, you see that memory is still you know quite a bit higher than it was in previous years. But you see the formidable strength of the, of the logic business. Interesting, I will not get you through every single number here, but I just wanna highlight a few of them.
So, it's interesting to know that, actually, in three years' time, we doubled our sales, right? So if you look at 2020, it took us three years to double the sales from 2020. So I think that is a pretty big achievement and a pretty big testament, I think, to the strength that we enjoy within the industry. Also, take a look at R&D. R&D, it took us four years to double the R&D amount. EUR 4 billion, EUR 4 billion. That's an interesting data point. EUR 4 billion is the R&D expense that ASML has, and as a matter of fact, if you want to look at spend, you can actually, you should actually add, I would say, at least EUR 1 billion of CapEx that we also do on the R&D front.
So our total spend on R&D, I think, gets close to EUR 5 billion, and that is just our spend. In addition to that, of course, there is a lot of spend that happens in our ecosystem. There is a lot of R&D spend that, you know, our suppliers do on our behalf. That's, you know, the research institutes that we work with, that they do. And then if you look at the total R&D spend for the Netherlands, according to the central agency for statistics in the Netherlands, that's around EUR 20 billion. So you can see that the total contribution of the, of the ecosystem that we, that we operate in to the total share that they have in the, in the R&D spend in Netherlands is quite significant.
Of course, not all of the spend is in the Netherlands, but the lion's share of the spend really is. And I think that really tells you, you know, about how strong this ecosystem is and how much it contributes, I think, to our, our Dutch economy. And then you see that the growth that we experience also, you know, works its way into a net income, a very strong net income, as I mentioned, nearly EUR 8 billion, and obviously also into earnings per share and what have you. In terms of return to shareholders, as I mentioned, in the aggregate, the return to shareholders in 2023 was a bit lower than it was in 2022. On the dividend side, we continue to grow dividend, and that is our policy.
We try to have a dividend that grows over year, over the years. So, the dividend that pertains to 2023, we propose to the, to the AGM a dividend of EUR 6.10, and that is a 5.2% increase over the dividend over 2022. So I think that's, that's healthy growth. But as you can also see, and as I mentioned to you, the share buyback was quite a bit lower than it was in previous years because we decided to help our customers and the rest of the ecosystem and put our money there rather than, distributing it back to, to shareholders in 2023.
In terms of outlook, outlook for the quarter, it's a soft start to the year, but quite frankly, that was a bit anticipated because, you know, based on all the demand timing with the fab openings, et cetera, we clearly see that 2024 will be much more skewed towards the second half than it is towards the first half. So, a relatively slow, slow start with net sales expected to be between EUR 5 billion and EUR 5.5 billion, and you see all the rest here in terms of, the implications for, gross margin, R&D, et cetera. Gross margin is quite a bit lower than what you saw. This is primarily because of product mix.
We expect the first quarter to be fairly light on immersion, fairly light on the installed base business, and those are typically, you know, the big contributors to gross margins. So that's the reason why gross margin in the first quarter is expected to be a bit lower than, than in the rest of the, in the rest of the year. In terms of 2024, I think Peter said it, we really look at 2024 as a transition year. We articulated the expectation that we think 2024, from a revenue perspective, will be similar to 2023, but still gonna be a big year for us in terms of preparing for 2025. For all the good reasons that Peter mentioned, we think 2025 is gonna be a big year of growth. Why is that?
Well, first off, because the secular trends are very clearly intact, right? With, you know... Without a high-end logic, there's no AI, AI. Without high-end memory, there's no AI. Without ASML, there is no high-end logic or high-end memory. So it's very, very clear that, you know, the AI development will have a very significant contribution to our, to our business in 2025. So that's a big one. The secular trends are very much intact. Secondly, you know, you saw the fab openings that Peter was sharing with you, so that's, that's a big driver. And we do believe in 2025 that we're, you know, nicely underway in the, in the upcycle within the, within the industry. So that's the way we look at 2025.
We have to prepare for that, we have to be ready for that. So we're expanding capacity, we're very much investing into the high-end AI technology, both here in the factory and also in the, in, in the field, to make sure that the entire organization is, is ready to deliver on that. And that is a, you know, big commitment that we're making, towards expansion of capacity within the industry. That said, how do we look at, at 2024 then? So as I mentioned, at the top line, similar. If you look at the moving parts, if you look at the end markets, we believe it will be, we will see growth on the, on the memory side.
You know, Peter said it, with DDR5, and with high bandwidth memory, you know, we clearly see that technology investments of the memory players are significant. You also saw in the order intake, that about 50% of the order intake was related to memory. That is very, very high, and that just tells you that indeed, these investments in the technology transitions are being made by our customers, and we think that should be visible in the, in our business this, this year. Logic, we, we think will decline a little bit in comparison to last year, because already so much CapEx has been added, to, and capacity has been added to, to logic.
So we think, you know, our customers will probably absorb that addition in 2023 before they're really going to massively invest again. So that's why we think memory will be up this year, and we think logic will be a little bit, little bit down. In terms of the, i n terms of the way it is being distributed over the year, as I mentioned, we think the second half will be, will be stronger, and there we think, you know, momentum will continue to be built second half, and then really going into what we think will be a very, very strong 2025. Dear friends, that is it from my, from my end. So, with that, I think we're going to Q&A.
Yeah. Thank you, Roger. I think your place is in the middle.
I know.
I got instructions. Thank you.
Politely follow instructions today.
Yeah, you all think this is spontaneous and easy-
Oh, well,
... but, hey, I'm getting instructions here. This is my place. So I'm sure there are questions in the room. Let's start there because, you know, you're here. I have my colleague, Mark, here to with a microphone. For those online, I've got a colleague, Kelsey. She is taking your questions, and she'll send them through to my... Oh, Ting-Fang's already there. Yes, hang on there, Ting-Fang, we'll get, we'll get to you. So she's sending out further questions, so to my iPad. So Mark.
Good. Introduce yourself.
Sandra Olsthoorn, Financieele Dagblad. I have two questions. The one is on the gap between the orders and your outlook. I'm wondering how much has to do with the market and how much is ASML, as you're saying, preparing for the big upturn, but maybe you're taking it easy on the execution or taking your foot off the gas to prepare yourself, if you know what I mean. The second question is about the China orders. It says in the press release that the export controls, do you think are going to cost 10%-15% of the Chinese revenue? I was wondering, in the past, ASML always communicated that revenue that is being made in, not made in China will always go somewhere else. I'm wondering if that's still the dynamic, or is this really revenue that you're missing? How should I interpret?
Should I take the first one-
Yeah
... and you the second?
Yeah, that's what I thought.
So on the first one, by default, we never take the foot off the accelerator, right? So that, we're not-- that's not what we're doing. If you look at the orders that came in, a lot of those orders, some of that was for 2024, a lot is also for 2025, right? So, it is a very good start into the growth year of 2025. For 2024, we're essentially covered with orders, so, you know, for the, for the amount that we're, that we're guiding, we have, we have the orders in. Of course, it could be if, you know, Peter talked about the slope of the ramp.
If the slope of the ramp is going to more, is going to be more aggressive, if indeed customers see that demand is going to, to come back stronger in the course of 2024, then I'm pretty sure customers will start knocking on our door and say, "Hey, could we accelerate some of the, some of the demand?" And as Peter said it, we look at 2024 and 2025 combined, and we know 2024 and 2025 combined will be two big years. It's just a matter of, is it gonna fall into one bucket, or is it gonna fall into the other bucket? So you're absolutely right. I mean, there is a, there is a few, there are a few strong indications of recovery. The utilization of our tools is improving.
Inventories at customers and end customers are actually going down, which is always a good a healthy sign. And indeed, the order intake is strong. So there are some very positive developments, but we said it's a bit early days. You know, three months ago, we said, you know, we expected that the year is gonna be flat. It's a bit too early to now say that we're, that, that we're gonna, that we're gonna adjust our guidance. So we stick to our, what we call conservative guidance for the year. It—Could it be that, you know, some demand is going to be accelerated from 2025 into 2024? Absolutely, but that's not what we're guiding today.
If that happens, then from a capacity perspective, we would be able to ship it more. So-
To ship earlier.
Yeah. Yeah, to ship earlier.
Yeah.
Yeah.
I think also you need to realize that the order lead time for, for instance, EUV systems, is between 12 and 18 months, and for a high-end system, even longer. So a lot of those orders are also with an eye on, for instance, the end of the year or, you know, 2025? So that's also the case. Now, on on our China business, strong. Yeah? But, as I said it before, the 2023 is really a reflection of our under-shipment in 2022 and 2021, where we basically had the orders, so we shipped the orders in 2023 for that we, that we got in, you know, 2022.
Yeah.
The rest of the industry went through a downturn, but we undershipped China significantly. I think it was. They got what we call the order fill rate was less than 50%. So that's all clearly create a situation where you can ship in 2023. Now, the 10%-15% you referred to, we just gave this as an indication of the size of the impact of the export control. It's about 10%-15% of the 2023 China sales. Yeah? So that's kind of to give you an indication that, yes, it has an impact. Clearly, it has an impact. But what's our China business? Our China business is significantly in the area of the mature chips and the less of the mid-critical chips. And why is that?
It goes back to you know, our slides that I showed. When you talk about connectivity, you talk about AI, it's all about advanced. That's not in China, yeah? But it's when you think about the rest of these trends, whether it's energy transition, EV transition, industrial IoT, the rollout of smart grids, you know, life sciences, that's where the Chinese industrial capability is pretty high, and they need a lot of chips. So this is why they are buying that stuff, and they will keep buying that stuff. But you're also right that, say, what we don't sell to China, we'll sell somewhere else because our Chinese customers or the end customers in China still buy chips. They buy chips from Korea, they buy chips from Taiwan, they buy chips from the United States.
So the advanced chips, they still buy. They don't make them. They make the, you know, mature chips, huh? So that will actually reposition itself, so you, so you are right. What we don't sell in terms of advanced chips, still the capacity is needed, we'll sell somewhere else.
Okay, next question.
Hi, Pieter Haeck from POLITICO. I had a question on Taiwan. There were recent elections there, that gave the victory to a more China-skeptic candidate. There were also some strong remarks from the Chinese president on reunification with Taiwan. You have a sizable presence and a sizable revenue also in Taiwan. Are you increasingly cautious or concerned about the business side in Taiwan?
No. No, I can give you a very short answer, but it's... No, I think these are political elections and, you know, you can read the political analysis of the people who really understand this, which, you know, you shouldn't ask me. But what I read is that, you know, as always, yes, there is this new president that is from this particular party, and but they don't have the majority in parliament, so big law changes need to go through parliament, which basically creates this balanced situation where, you know, there's always going to be political gives and takes across the globe, you know. That's true for Taiwan. That's, I think it's true for many, many other countries.
So, we'll just follow what's happening, and we don't feel that what happened in Taiwan is a reason why we should lower our activity in Taiwan. No.
Okay. Thank you. So let's move to a question from Asia. Ting-Fang from Nikkei Asia. About Japan, "Please, could you share how you think of Japan's plan to bring back their chip industry and supply chain? Do you think they will be successful, and what are the challenges and opportunities?" I guess she's talking about Rapidus. "One of your major customers' expansion went quite well in Japan. What are ASML's expansion plans and status in Japan, and how do you observe Japan's opportunities?
Well, I think, you know, Japan has a very strong history in semiconductor manufacturing. For all kinds of reasons, the market share of the Japanese manufacturers went down. I think there's a very clear focus of the Japanese government to reinstate that, you know. It's not only they're historically very good in memory, but, you know, we now have a new initiative, Rapidus, which is basically focusing on advanced, you know, semiconductor manufacturing-
Mm
... which of course, we are going to supply very advanced machines. So I think, it, it's going to be a challenge. You know, where is the, you know, challenge? The challenge is, you know, you have to rebuild some of that, some of what was actually slowing down over, or has slowed down over the last couple of years, or last couple of decades. So that means people, so you need to invest in people, you need to invest in the ecosystem, and this is also why your industry partners, like ASML, will also invest. Where, this new enterprise, you know, Rapidus, is going, we will go. But with us, others will go. So it's gonna take some time, but with the right focus, with the right incentive programs, it means that you can reinstate that.
I think that's the goal, and I think we all should be... I'm an optimist that, of course, worries a lot from time to time. But my optimism clearly tells me that if you have the right focus, and you also, you made reference to one of the initiatives that is building out, that's, that's a Taiwanese company, building out, excess capacity or extra capacity in Taiwan, sorry, in Japan, that's going according to schedule. There's also very much a Japanese characteristic, you know? When they do it, they plan it, they execute. So, you know, everything's there. Yeah. We just need- It's going to be not gonna be easy. These things aren't easy, yeah? But we're there to help.
Okay, thank you. Questions in the room? Yeah.
Good morning. I'm Sarah Jacob from Bloomberg News. Can we expect the contribution from the Chinese market to ASML sales to go back to single digits in 2024?
Go back to single-.. It's, it's, it's.
Single digit.
Yeah, it's-
Single digit.
For the last couple of years-
Perce- percent
... it hasn't been single digit, and I don't expect it to come, to go to single digit either. Like I said, you know, if you look at 2023, and you take the, and the impact of the export control rules, it's 10%-15% of our 2023 business, which the rest of the business is mature technology that are needed for all the-... transitions that we're currently having to deal with. So I don't expect that. Like I said, the last couple of years, it, it hasn't been single digit, and I don't think it's, it's going to be.
Yeah. Not for anyone in the industry, right?
No, that's not only for us.
I mean-
Mm-
... yeah, without just taking a lithography tool, you can't make a chip. You know, there's our U.S. peers and our Korean peers, they're-- Just look at their financial statements. They're just, well, going very strong, especially in this domain of the more mature or the mid-critical part of the semiconductor manufacturing. Yeah.
Okay. Anyone else in the room? Yeah. Merlijn.
Thanks. Merlijn van Dijk, Eindhovens Dagblad. During 2023, you slowed down the hiring of new people, significantly, significantly, also here in Veldhoven. Could you elaborate on your plans for 2024? And maybe also give a number as to how many people you are planning to hire this year.
Yeah, I think we're not gonna give a number of the people we're going to hire. And the slowdown is indeed that we slowed down from hiring 10,000 people per year to 3,000 people per year. So that's a slowdown. But we're still hiring people, and we need to hire people because we will grow. And that's gonna be a reflection of our forecasted growth, and our forecasted growth, as you've, if you just follow the the indications we have given during our capital markets days, is going to be double-digit. Yeah? Now, that's going to be not a straight linear line. It's going to be somewhat, you know, bumpy, you could say.
But we're gonna hire people, and in a period where our industry, like we said, 2024 is going to be a, it's around a flat year, yeah? It also means that, yes, while we're hiring people, we're also going to use that time to actually make those people that we hired, to make them more efficient, to make them more effective, and to work on processes, and to improve things that you can improve in the company. But longer term, there's gonna be growth, and significant in terms of, you know, FTEs people. Yeah.
Yeah. So speaking of growth, I also got a question in from, and I'm hoping I'm pronouncing this correctly, Ridha Loukil from L'Usine Nouvelle. French. Your CapEx plan in 2024 compared to 2023, where does it go?
The CapEx plan?
Mm-hmm.
I think that will be sort of in line with what we had in 2023. That would be my expectation. So, you know, as we mentioned before, we are really building capacity against the targets that we've communicated before. By 2025, 2026, we want to be able to do 90 EUV and 600 DUV, and mid-term, we're looking at 20 High NA. So that's a big jump, and of course, that requires quite a bit of CapEx. And, you know, these are last year and also this year, are the key years in which we're building that. So I think it's prudent to expect that that will be more or less-
Yeah
... at the same level.
I'm guessing the question is also in what regions are you investing in CapEx?
Oh, it's everywhere. So of course, the lion's share of investments is here, because this is where most of the manufacturing happens. But y ou know, we're also investing in the United States, we're investing in Korea, we're investing in Taiwan. I mean, those are the places where we have the highest investments. Some in China. So, you know, we are investing, you know, across the globe.
Yeah. Thank you. Anyone here? Yeah. Paul.
Paul van Gerven from Bits&Chips. Can you comment on the High-NA order intake, please?
Yeah, on High-NA, we've we've never, we've never quantified that because our customers don't want us to do that. Because they believe High-NA is so critical and strategic to them, that they don't wanna, you know, give away to the very small world in which they operate what the what the order profile and what the order profile is. But, you know, we did say that we had multiple High-NA orders in the order intake for the last quarter. So it's a healthy journey. We're happy, we're happy with what's going on there. We communicated a couple of quarters ago that we were in double-digit territory in terms of orders that we had there. And of course, we added a couple thereafter. That gives you a bit of a perspective. Healthy, in line with our expectations, but we cannot put a definitive number on it.
But it's nicely supporting our 25 and 26.
Absolutely. Yeah
... you know, build plan. So yeah, we're, we're happy.
Okay, and Peter, do you think Europe should follow Japan's example in establishing leading-edge manufacturing capacity of its own?
Absolutely. You know, I think you have two initiatives. You know, we have initiative that in, particularly in, Germany, two major semiconductor makers are there, have been, been very clear about their plans to, you know, establish semiconductor manufacturing. I think it's needed. You know, there's a concern by a lot of people that said basically all these incentive programs, they will create this big overcapacity because all, you know, capacity will be built in all places in Europe, in Japan, as we just discussed, in China, in Korea, in the, you know, United States.
But you have to realize that, well, not one of those charts, if our industry, our customers' industry, is going to grow over $1 trillion by 2030, it took us 40 years to build the capacity to do $550 billion, then we only have less than 10 years to do another step up and doubling it again. Trust me, we need that capacity, and if that capacity is only concentrated then in only one or two places on the planet, you get a level of dependency that people, after COVID, realized it's not healthy. So I think we need to do that, yeah?
But we also need to do it, going back to one of my other slides, is not only on the connectivity part, on the high-end stuff, but we have to do it also on the mature side, yeah? The mature and the mid-critical, 'cause that's where the volume goes, yeah. If you take an electric vehicle, a handful of very advanced chips, hundreds, hundreds of chips that are more of the mature nature, that's more silicon, yeah? So I think this is where, if I was—but we've always said this. You know, we've—if you might even know, we issued a white paper just before the presentation of the draft EU Chips Act, where we made that point, is that you cannot only focus on leading edge.
You have to also focus on the mature chip, 'cause those are the engines, the workhorses, of these major transitions, yeah? And of course it all ends in high-power compute and access to data, very fast access to data, high bandwidth memory. Yes, yes, yes, yes, all true, yeah? But it's combined. It's an ecosystem. You cannot do one without the other. And this is where I think that's what governments need to realize, and Europe is not alone in that. I think there are more governments around the globe that are focusing solely on advanced chip manufacturing, while I think real issues are going to pop up in the mature space. And that's where Europe should also invest.
Okay. Let's go to a question online. Kunal Kamal from Mergermarket. I think this is for you. Surprisingly, the question is on M&A. With the semiconductor landscape undergoing significant changes and given the company's strong cash position, are there any regions or technologies where ASML sees strategic opportunities for M&As or joint ventures to support its technology portfolio and/or to expand capacity? And would you be looking at other partnerships like the one you announced with Samsung in December?
That's a very broad question...b ut in general-
Why don't you explain our M&A strategy?
Exactly. Why don't I do that in detail? So bottom line, and I think that's consistent with also what we said at the Capital Markets Day, if you look at the opportunities that ASML has ahead of it, it's phenomenal, right? If you're looking at the EUR 44 billion-EUR 60 billion that we've said we believe is possible in the 2030 timeframe, it tells you that we have a lot of opportunity that we see for ourselves in our current, in our current market. And we also believe there's still a lot to be done, because in order for that to be successful, we have to build capacity. In order for that to be successful, we have to make sure that High NA is up and running.
And we think that, you know, the company is best off, and we think our stakeholders are best off, if we really focus on that. So that's, you know, the emphasis, I would say, for the foreseeable future. Never say never, right? I mean, obviously, there could be a point in time where you're gonna say, "Now we have more capacity to look at, to look at other product market combinations." But for now, we believe the opportunity that we have ahead of us is still so phenomenal that there is no action required on that, on that front. That said, you know, we are very active in partnerships. We are very active in partnerships with our customers, and indeed, you know, Samsung is a great example of that.
And, you know, we have a very specific one there, but I can tell you with all of our customers, we have very intense collaboration, and we double down on those efforts to make them successful, 'cause that, that's, I think, what it ultimately boils down to. We have very significant initiatives within our ecosystem, with universities, et cetera, to develop stuff. So we do a lot on that, on that front. And finally, you know, on the supply chain, and there I think, again, very consistent with what we've done in the past. We're very happy, you know, with a supply chain that is so dedicated to our ecosystem that works really well for us. The last thing we wanna do is buy up all of those suppliers, make them part of ASML.
I think the model we have today works extremely well. We have a level of focus and dedication in our supply chain and professionalism in our supply chain that I think is, is unique. We would only do an acquisition in the supply chain if we really feel that there is a supplier out there that cannot keep pace with what we do. If that's the case, if we see a supplier that is critical to what we do, but for whatever reason they cannot make the commitments that we require, then we, then we make an investment. Yeah.
Thank you. Questions in the room? 'Cause I have some more on my iPad. Yeah, okay.
ASML has been pushing at the boundaries of Moore's Law for decades. When do you see Moore's Law sort of expiring?
Yeah, I think, yeah, Moore's Law, said it earlier, I think it's an empirical law of economics, so it will slow down when it doesn't make sense to create this next transistor or function, whereby the value of the transistor of the function is the same or lower as the cost. Then you stop. And and the value of the transistor or the function is not only driven by lithography, it's one of the scaling engines. But very important scaling engines are system integration, system architecture, material choices, that altogether creates this ability to keep scaling, huh? And that together is what you do with the customer. So I think... And with our peers.
So if we and our peers and the customer come to the conclusion that the added value, that the incremental value of the next generation of transistor or function in the chip is not going to cover the cost, it stops. We're very far from this. So it's the combination of all of it, that at least we think for the next decades, we will see Moore's Law continuing, because it's a combination of everything. And I think when we talk to our customers, we see this value accretion per transistor or per function of the chip still higher than the cost. But that's our challenge. That's the challenge that we have, together with our peers, to manage that cost, yeah?
... Okay, let's go to Asia again. Peilin Liu from Caixin Media. "The new message is that," I think it's a clarifying question. "The new message is that you are now expecting for a handful of fabs not to get export licenses for China for the 1970. In last quarter's conference call, you didn't mention anything about the 1970. How does it happen?" And then a follow-up question, which we also got from Thijs Rösken from NU.nl: "Will we see more models banned in the future?
Well, you know, just a confirmation, yes, it's the 1970-
Yeah
... 1970 , 1980 . There's not a-
Yeah
big difference between the two, so you can-
Yeah
Mention them in, you know, in one sentence.
Yeah. I think that was the-
So yes, I mean, that was indeed-
Yeah
... also for this handful of fabs, they cannot access-
U.S. rules, yeah.
And for the rest, you know, I'm the last person that's gonna speculate on on any additional, you know, export control rules. But we have to realize that, you know, export control rules as a as a mean to control national security interests, very valid, yeah? That's that's pretty normal, yeah? But where we are today is that when we have done that, and the rest is for mature and mid-critical chips, which we need for day-to-day life and for all these transitions that are coming, that's an economic-
Mm
... you know, decision to actually help society and to actually solve these big problems. I think, you know, I would be of opinion that we should absolutely keep that open. We should just make sure that the world, yeah, if we want to grow to a semiconductor industry that's over $1 trillion, you know, because that's going to support all those major transitions, then we better make sure that there is enough capacity. Thank you very much, yeah?
Mm-hmm. Okay, thank you. In the room, any questions? Yeah, Pieter.
You mentioned in your presentation that you need that international population, international students coming to the Netherlands. How do you look upon the formation of the new Dutch government, which may not be very open to knowledge migration to the Netherlands?
Yeah. Well, I always say, you know, you should know what you wish for. So if you, if you don't want this in the Netherlands, then you should accept the consequences. And the consequences are, just to give you an example, 65% of all the international students in a technical university in Eindhoven lands in this region, the Brainport region. That is a very significant source of talent that we need to drive innovation. Well, this industry is not gonna stop. We're going to have 1 trillion+, yeah? And we have to expand across the globe. We need to support it with R&D, and with people, and with innovation. If we cannot get the people here, we get the people somewhere else. It's very simple, yeah? We will go where we have access to the means to grow the company.
If the, if the Netherlands shuts down because we cannot get immigrants, or foreign students-
Mm-hmm
... yeah, fine, you know, you have to accept the consequence. You should know what you wish for, yeah? So we are a company, we're a global company, yeah? We will go where we need to go to make sure the company can grow and, you know, service our customers.
Okay, any questions here? Then we go back to my iPad. Again, L'Usine Nouvelle: "You seem to be moving against the grain of the semiconductor market. The market was down, according to Gartner, that was -11% in 2023, and you end the year with a 30% jump. For 2024, the market is forecast to see strong growth, +17%, according to Gartner, and you are counting on a flat turnover.
Yeah.
Why these contrary developments?
No, it's, yeah, that sounds intuitively, yes, but it was. It's what Roger said.
Right.
One, part of the 2023 sales is a catch-up of, of customers that, "You undershipped us, more than 50%, so can you please, give us the machines that you obviously reserved for other customers? Can you, you know, ship them to us? Thank you." And these are very long lead time items, so they get the machines in 2023, so they have the capacity. That's why Roger said logic is a bit down probably this year, because they're gonna use that capacity. Don't need to invest twice. I would love them to invest twice—but they're probably not gonna do it, yeah? So this is why there seems to be a contradiction, but this is why if you chop up life in, in, in periods of January 1st to December 31st, you get these very strange views, yeah?
Mm.
That's why we look at 2024, 2025 together, 'cause it's a continuum, yeah? Whereby long lead times of ASML, short lead times of our customers, all play a role in how the dynamics within such a period actually happen. So it's not a contradiction, it's perfectly in line.
Yep. Okay, I have one more question on my iPad, but we already covered that. So if there's nothing else here, then I have one more slide for you, which is thank you. Thank you for being here. Thank you, Roger, for again explaining the figures so well and make them come alive.
Pleasure.
A really big thank you, Peter. This is, our last press conference together. In January, you'll be here for the next coming months, and I'm sure you're not going to slow down.
Right.
I just want to thank you here for all the times you were here to talk to everyone and taking the time for interviews. It's been a pleasure and a privilege to work with you on this way. Thank you.
Hear, hear.
Thank you.
Pleasure was on my side. An absolute pleasure to work all these years with ASML, and then I got a question on one of the interviews. I said to you, "How do you feel?" I said, "It's a bit sweet and sour." You know, it's... it's the sweetness is to see where the company is and to see what we've done, and the sour part is that, you know, you have to say goodbye, which, you know, is logical.
Not yet, huh?
Not yet.
You're still here till-
Not, not yet.
... April 2024.
But I'll tell you, I got the confirmation this week that my state pension, my AOW, will start on June 1st, so... Well, what it is-
So you can afford, right?
Okay, thank you all.
So thank you all. Thank you for all the years that we've you know exchanged our ideas and our thoughts, so thank you so much.
Okay.
Yeah.
Well, thank you all for coming, and see you next year.
Thank you.
Bye-bye.
Bye.