艾司摩爾 (ASML) 2022 Q1 法說會逐字稿

Thank you for standing by. Welcome to the ASML Q1 2022 Financial Results Call. (Operator Instructions)

I would now like to turn the conference call over to Mr. Skip Miller. Please go ahead, sir.

All right. Thank you, operator. Welcome, everyone. This is Skip Miller, Vice President of Investor Relations at ASML. Joining me today on the call are ASML's CEO, Peter Wennink; and our CFO, Roger Dassen. The subject of today's call is ASML's 2022 first quarter results. The length of this call will be 60 minutes and questions will be taken in the order that they are received. This call is also being broadcast live over the Internet at asml.com. A transcript of management's opening remarks and a replay of the call will be available on our website shortly following the conclusion of this call.

Before we begin, I'd like to caution listeners that comments made by management during this conference call will include forward-looking statements within the meaning of the federal securities laws. These forward-looking statements involve material risks and uncertainties. For a discussion of risk factors, I encourage you to review the safe harbor statement contained in today's press release and presentation found on our website at asml.com and in ASML's annual report on Form 20-F and other documents as filed with the Securities and Exchange Commission.

With that, I'd like to turn the call over to Peter Wennink for a brief introduction.

Thank you, Skip. Welcome, everyone, and thank you for joining us for our first quarter 2020 results conference call. I hope all of you and your family are still healthy and safe. And before we begin the Q&A session, Roger and I would like to provide an overview and some commentary on the first quarter 2022 as well as provide our view of the coming quarters. And Roger will start with a review of our first quarter 2022 financial performance with some added comments on the short-term outlook. And I will complete the introduction with some additional comments on the current business environment and on our future business outlook. Roger, if you will?

Thank you, Peter. Welcome, everyone. I will first review the first quarter financial accomplishments and then provide guidance on the second quarter of 2022.

Net sales came in at EUR 3.5 billion, which is at the high end of guidance. We shipped 9 EUV systems and recognized EUR 591 million revenue from 3 systems this quarter. Net system sales of EUR 2.3 billion, which was nicely balanced between Logic at 50% and 50% from Memory. Installed Base Management sales for the quarter came in at EUR 1.2 billion, as guided. Gross margin for the quarter came in at the guidance of 49.0%.

On operating expenses, R&D expenses came in at EUR 739 million, and SG&A expenses at EUR 208 million. R&D was below guidance as spend rates in the quarter was lower than planned and will move to Q2. We still expect to be around 14% of sales for the year.

Net income in Q1 was EUR 695 million, representing 19.7% of net sales and resulting in an EPS of EUR 1.73.

Turning to the balance sheet. We ended the first quarter with cash, cash equivalents and short-term investments at a level of EUR 4.7 billion.

Moving to the order book. Q1 net system bookings came in at EUR 7.0 billion, including EUR 2.5 billion for 0.33 EUV NA systems and multiple EUV 0.55 NA systems, EXE:5200.

Another very strong DPV order intake of EUR 4.5 billion this quarter, reflecting the continued strong demand for advanced and mature nodes.

Total net system bookings was driven by Logic with 66% of the bookings and Memory accounting for the remaining 34%.

With that, I would like to turn to our expectations for the second quarter of 2022. We expect Q2 total net sales to be between EUR 5.1 billion and EUR 5.3 billion. This excludes around EUR 800 million of net delayed revenue for Q2 as a result of more fast shipments at the end of Q2 than at the end of Q1. We expect our Q2 installed base management sales to be around EUR 1.2 billion.

Gross margin for Q2 is expected to be between 49% and 50%. Relative to last quarter, we expect positive margin impact from higher volume for both EUV and DPV, offset by lower EUV ASP, DPV mix and continued cost pressure in the quarter. The expected R&D expenses for Q2 are around EUR 790 million, and SG&A is expected to come in at around EUR 220 million. In the current environment, it is also appropriate to address how we may be impacted by rising costs this year. We're not immune to rising costs. There is pressure on labor cost as the global job market for engineers is tight, and there is a competition for talent.

Costs related to components in the supply chain are also increasing due to higher material costs, including additional fees to secure parts.

Transportation costs have increased due to rising fuel costs and changing flight routes. Energy contracts and renewable energy help limit the increased energy cost impacts. We clearly see pressure on margins due to these cost increases, which we expect to translate roughly to a 1% impact on gross margin for full year 2022.

We expect the second half of the year will be strong with expected gross margins of around 54%, primarily driven by higher EUV and DPV volume as well as improved margin from installed base business. In summary, we currently expect gross margin to be closer to 52% for the year.

Our estimated 2022 annualized effective tax rate is expected to be between 15% and 16%. In Q1 2022, ASML acquired 3.6 million shares for a total amount of around EUR 2.1 billion as part of our current program.

With that, I would like to turn the call back over to Peter.

Thank you, Roger. As Roger has highlighted, revenue and profitability for the quarter came in as guided with system revenue balanced between Logic and Memory. We expect a step-up in sales in Q2 as revenue from fast shipments in Q1 will be recognized. While supply chain challenges are still present, we will continue to utilize fast shipments as a means to get systems to customers as soon as possible.

Although the current macroeconomic environment creates uncertainty, we believe the fundamental growth drivers remain intact. We continue to see unprecedented customer demand across all market segments from both advanced and mature nodes, driving demand across our entire product portfolio. We are running at maximum capacity and expect demand to exceed supply well into next year. Our view of the full year revenue, therefore, remains unchanged with year-on-year growth of around 20% over 2021.

As mentioned last quarter, this 20% sales growth does not include the full shipment value of systems out for this year due to a number of fast shipments, which will result in a delayed revenue into 2023. For our EUV business, we still expect to ship around 55 systems this year. And as we plan to do fast shipments on a number of these systems, including systems in Q4, we expect some revenue to be deferred to 2023. This translates to an expected system revenue of around EUR 7.8 billion this year.

In our deep UV and applications business, we expect significant growth in both immersion and dry systems, as well as continued demand for metrology and inspection systems. In addition to advanced nodes, we see growing demand for deep UV systems supporting mature market segments such as analog, power and sensors. These market segments are part of the secular growth drivers in support of digital infrastructure, which includes automotive and green energy applications. We expect revenue growth of over 20% for non-EUV system revenue.

For the Installed Base Management business, service revenue will continue to scale with the growing installed base of systems. Customers continue to look at all methods to add wafer capacity, including productivity upgrades. In these times, a very high fab utilization, some of these upgrades are hardware-intensive and require systems to be taken off-line for installation. Therefore, available system time will dictate their ability to install these types of upgrades. And as we mentioned in previous quarters, the quickest way to add wafer capacity is to install software productivity packages. These upgrades require less system downtime and have now been installed in most of our customer fabs. We currently expect 2022 installed base revenue to be up around 10% year-on-year.

Looking at the market segments in 2022. Our view on growth is similar to last quarter. We still expect Logic system revenue to be up more than 20% year-on-year and Memory system revenue to be up around 25% year-on-year.

On High-NA EUV, we're making good progress. And we have currently started the integration of the first High-NA system in our new (inaudible) in Veldhoven. We received multiple orders for our EXE:5200 system in Q1. We also received additional EXE:5200 orders this month, April. With these bookings, we now have High-NA orders with 3 Logic and 2 Memory customers. The EXE:5200 is ASML's next model High-NA system and will provide the next step for lithography performance and productivity.

The global market trends that we talked about at our Investor Day last year are broadening the application space and providing secular growth drivers for future demand. The strong demand this year and beyond is reflected in the significant bookings over the past several quarters, resulting in a backlog of around EUR 29 billion is an all-time high. We expect the strong order intake to continue as demand will continue to exceed supply also going well into next year. With multiple countries pursuing technological sovereignty, we are now seeing a number of announcements from customers for new fabs in the coming years in support of this global trend. These announced investments are expected to have a positive impact on the medium-term demand. As this unprecedented demand is exceeding our capacity, ASML and its supply chain partners are planning to actively add capacity to meet future customer demand as communicated during Investor Day last year. But at that time, we talked about the current capacity ramp is expected to deliver an output capability of over 70 EUV 0.33 NA systems and around 375 DPV systems by 2025.

As mentioned last quarter, we see a need to further increase our output beyond this level in order to meet the stronger and longer market demand to support an industry that is expected to at least double by 2030. With the goal of adding more capacity, we're investigating the feasibility of increasing our annual capacity by 2025 to around 90 EUV 0.33 NA systems and 600 DPV systems.

For deep UV, we're planning to increase capacity for both immersion and dry, with a heavier weighting towards dry. We're also discussing with our supply chain partners to secure a capacity of around 20 EUV 0.55 High-NA systems in the medium term. Bear in mind that this translates to what we currently feel our maximum capacity goal should be and may therefore not be a final output plan. We discussed our goal recently with our supply chain partners and asked them to come back in the coming quarters with confirmation on the feasibility of our request.

Once we complete this analysis, we also expect to revisit our 2025 scenarios and growth opportunities beyond 2025. We're planning to provide an update to the capital markets in the Q4 time frame.

It is clearly a dynamic, challenging but also exciting period in our industry and it only increases our confidence in our long-term growth opportunity.

And with that, we'd be happy to take your questions.

Thank you, Roger and Peter. The operator will instruct you momentarily on the protocol for the Q&A session. (Operator Instructions) Now operator, could we have the final instructions and then the first question, please?

(Operator Instructions) One moment, please, for the first question and that comes from the line of Robert Sanders at Deutsche Bank.

Like we need to move to the next caller, operator.

So the next call is -- the next caller is Didier Scemama of Bank of America.

Third one?

Okay. So the next question will come from the line of Joe Quatrochi of Wells Fargo.

Looks like it's more structural.

Mike, I think there's a connection issue.

Just one moment, I'll just have a look into this. It does look like the line is open, but yes, that would be 3 questions in a row. That looks like it's possible that's the connection. Let me just have a moment.

(technical difficulty)

Okay. Apologies at the latest. So I'm just going to test this. Joe from Wells Fargo, are you able to hear us and speak?

Yes. Can you hear me now?

Yes. We can hear you. Apologies.

Yes. Correct. We just have to make sure perhaps we can squeeze in the other -- a few that were kicked out because of...

Yes. Let's cycle back on both Rob and Didier exactly.

Yes. Rob and Didier. So we just have to make sure that they somehow get back in the queue.

Yes. Okay. Go ahead, Joe.

Joe, please.

Okay. Great. So I wanted to ask a question on the capacity expansion plans, specifically on the DUV side. When I look at your prior expansion plans relative to like the high end of your 2025 model, it looks like you had been kind of putting in about a 30% higher manufacturing capacity. So with the updated capacity plans, are you looking to maybe increase that kind of softer range? Or how should we think about that?

Yes. Joe, that's a good question. Basically, the number of 600 is a result of 2 things. One is, you have to remember, when we did the September Capital Markets Day, we prepared this a couple of months before. So I think you're looking at market intelligence that, as compared to last year, is about a year old. So we have -- has a bit more insights. And I think that's translated into that number. And on the other hand, we are currently looking at the demand for this year and next year, where it comes from. Looking at the customer base, which effectively means that every customer we almost ever sold a machine to is coming to us to ask for a DPV tool, which was just particularly true for dry.

So it's the current demand level plus our, let's say, insights that have progressed over the last 12 months or so. And those insights are basically across a number of lines. I would say one is the mature market upside is higher because the visibility has increased, especially from Asia. And when we look at the type of semiconductors that are in demand now and the capacity that's been installed and especially if you look at the lower end, when you talk about power ICs, you talk about sensors, optical and non-optical sensors, analog. You talk about microcontrollers. There is a demand for those type of semiconductors almost in every industrial area, which when we look at it and we look at the growth numbers and the demand, talking to our customers, it's a very wide range, just the mature market upside is really driven by a significant increase of IoT-type opportunities. Now that's one.

Number two is we also see the need for more silicon for the PC and high-power compute applications. It has to do with the fact that you actually see that in order to be more energy efficient, the device's clock speed, the clock frequency is decreased but actually has a significant impact on the energy usage. However, it also has an impact on the performance. So how do you compensate that by adding more transistors. Where you have more transistors, you add more silicon square millimeters. So die sizes are growing. That's what we very clearly see has also been confirmed with many customers.

Then we see also in non-semi litho applications, for instance, goggles and waveguides. We see an increase in the demand for lithography equipment. And then on top of that, we see the demand medium term for technological sovereignty demand, which is basically an incentive program that actually helps our leading-edge customers to build out fab capacity quicker. So I think all in all, the -- and also the -- I would say, the market visibility has increased. A very important issue, of course, that our customers are finding out, that our order lead time is one thing but the most important lead time is the capacity lead time. So our customers now sharing with us much longer capacity improvement plans, which actually extend way into, let's say, the 2024, 2025 time frame. So all that together gives the confidence that the numbers that we're talking about, which are 90 EUV numbers and the 600 deep UV numbers have actually a basis. Now a lot of the detail we would like to share with you once we get the confirmation from our supply chain partners that, that feasibility request is -- can actually be executed on. And we think that will happen in the next couple of quarters. So in Q4, it's the reason why we want to update you to a Capital Market Day, where we can give you some more details on what I just said, which basically -- I think I want to give you some details, but that then we can spend a bit more time on it.

And then specifically also, Joe, to your question on the buffer because you are right in the -- on the Capital Markets Day, we did have a buffer because there also we indicated that we were looking at a capacity of 375 units in DPV at that stage. And if you took our market share in the worldwide demand, then, of course, that was lower than the 375. So that's the buffer that you're referring to.

There, I would say that's typically something that we will discuss in the Investor Day that Peter was referring to because Peter is talking about what we're looking into in terms of capacity. At the Investor Day, at the Capital Markets Day, we're obviously also going to articulate our expectations on demand. And there, I think you will see how that capacity will relate to the demand that we see for 2025. And that will then give you a bit of an indication on the buffer that we foresee for 2025.

That's some really helpful details. Just as a quick follow-up. I was wondering if you could maybe just give us the puts and takes on the EUV ASP this quarter, what drove the upside?

Yes. It's primarily accounting, Joe, because, as you know, with VPA accounting, you get these corrections based on estimates and based on revisions of estimates that you make for the VPA.

And of course, then if you only have 3 EUV units in sales as we had for this quarter, then, of course, that has a big impact on the ASP. But it's not -- nothing specific. It's not that all of a sudden the ASP has risen dramatically. I think in your models, I think it's still prudent to take the 160 million as the ASP for 3600 going forward.

And apologies for the issue earlier. We are now able to go back to the line of Robert Sanders at Deutsche Bank.

It was just around the backlog for the EUV tools. If I just do some sort of probably simple math, it looks like your backlog for DUV only is sitting at around 500 tools or so, while you're shipping around 60 tools per quarter in DUV only. So does that mean that the waiting time for DUV tool today is more than 2 years?

And related to that, as you think about how you -- your increase of DUV capacity, where would you like that sort of waiting time to go back to? Would it go back to 9 months in a perfect world? Or is there a kind of new reality where backlog will be extended for quite a while?

Yes. Two good questions, Rob. One, yes, the DPV backlog is quite significant. You mentioned 500, it actually is a little bit more. But -- yes, we're shipping around 60 units. But you have to realize that while we're doing that, we are increasing our shipment capability. I mean, although we are not at the 600 units as we mentioned, we will be able to increase more deep UV shipments throughout this year and next year for the simple reason that we're basically pulling out all the stops in cycle time reduction, in process efficiency. So I think that number will go up. So it will not be stable.

Having said that, you are right. If you would come in and you want a deep UV tool now, it's going to be second half of next year because for the simple reason that the PO, the purchase order lead time is completely irrelevant now. It's the capacity lead time.

So you need to build capacity in order to be able to get more, which ultimately means that, yes, by we think 2025, let's take another 2 years at least to build out that capacity. Then we will go back to, let's say, 9 months and less PO lead time. Yes. So this is indeed the case.

This is also why the demand that we currently see in deep UV, but we can probably only ship at this moment in time this year, only 60% of the deep UV demand. So anything that comes on that, that's driven by our maximum capacity. And anything beyond that, indeed, is a matter of capacity lead time, not the only time.

Got it. Just a quick follow-up to that. I'm assuming suppliers will want prepayments, guarantees in order to build more capacity for you guys. And they may look for higher pricing. But so should we think about pricing of individual tools going up on top of the inflation that we're seeing from the various different cost drivers just from the fact that suppliers will want some kind of guarantee from you guys that you'll take delivery?

Yes. I think this is -- it's always been the case when you go to the supply base, hey, we need to add capacity. I think we have a pretty convincing story for them that this industry will grow. Like I said in the prepared remarks, we think there's a very good chance that this -- that the industry of our customers will more than double before the end of the decade. And if that's the case, we're just going to need more capacity, which they all agree with.

Now if we're going to help suppliers, and I'm not excluding that at all, it's just a matter of sometimes the financial means that a supplier might have or run into a certain financial ceiling that we might help and step in like we've done in the past, that's not a major issue.

That is not to take away the risk but just to make sure that people execute faster. And I think that, that might happen. But generally, our suppliers are big enough to pay for their own CapEx. So I don't think that's going to be a major issue.

And on pricing, I think the cost increase that we've seen are really short-term cost increases. We do value pricing. So it is not cost plus, what we are doing. Having said that, the value of the wafers coming out of our customers over time will also go up, which were just going to be a function of that cost increase. So it also means that going forward, value pricing will also lead to different prices that we will charge to our customers. I mean that's all connected, not short term, but definitely over the next 12 to 24 months when the volume purchase agreements run out, we have new volume purchase agreements that will be reflected.

And our next question comes from the line of Didier Scemama of Deutsche -- Bank of America, sorry.

I've got 2 quick questions. First of all, maybe a question just to understand a bit better the '25 capacity plan. I know you don't like to talk about WFE, but I would like to understand what's baked in your numbers because when I just sort of reverse engineer what you've guided for, I get to WFE roughly in the $140 million to $150 million range. I just wanted to understand if that's what you're thinking about.

And related to that, when we make WFE intensity numbers, assumptions, et cetera, you get to a global semiconductor demand in 2025 in the range of EUR 800 billion-plus to EUR 940 billion-plus, which I'm a believer, but that's an aggressive growth rate, especially in the context of the current macro environment and worries about researching, et cetera. So I just wanted to understand if that's the right way to think about it? And if not, why not? And I've got a quick follow-up on High-NA.

Well, I think there's -- a lot of what you said is the right way to think about it. But I think the only comment I would make is that you pin everything on 2025. When we basically say we want to build our debt capacity by 2025, we do this also with the intent and with the focus on what we believe is going to be a very strong secular growth for our customers. So even if you look at the latest Gartner data, actually, they are above EUR 1 billion for 2030.

And so it -- so it is capacity that is there for the second half of the decade. It doesn't mean that if we would have that capacity by 2025 that the -- we're not saying that the semiconductor industry will gobble up all our capacity by 2025, but it will definitely do that in the second half. And this is why it's important because you add capacity, you need lead time. You need to start now because it takes a couple of years. And that might be that we don't need it in 2025. We might definitely need it in '26 or '27 or '28 to your point.

Now -- and if you then say if you would look at the latest Gartner data, which is just above, I think it's EUR 1.1 trillion. If you look at that and you apply your historical WFE numbers, you're coming pretty close to what you just said. So it is needed. This is an investment in the growth of the company and the growth of the industry of our customers. That's what it is. And I'm not telling you what the 2025 revenue number for the semiconductor industry is going to be, but it's going to be higher. That's absolutely certain. But this is the reason why. So I think your line of reasoning is correct or you don't pin yourself onto 2025 as a year.

Very clear. Perfect. So my second question is on High-NA. So at the time of the CMD, you were sort of cautiously telling us about 5 units, give or take, 2025. Now you're saying that you're thinking about building capacity for 20 units midterm. So I take it beyond 2025. I just wondered if you could, a, give us a sense of how many units have you booked so far given the support you get from your 5 customers in the quarter? And give or take, I mean are you thinking about more than 5 units already for 2025? Or sort of what's the linearity of that 20, what's the trajectory? I mean is it really taking off in '25? Or is it taking off a little bit after?

Good question, Didier. I think at the Capital Markets Day, what we said is the 5 is just the 5 that we take into revenue. And the reason -- and you might also recall that the 5 was in there for the high market scenario, but also for the low market scenario. And the reason we did that is that we said we're a bit unclear at this stage as to how revenue recognition is going to look like for those systems. Also take into consideration that we don't know exactly how in '24, how in '25, how the installation process, how exactly that will work because we have different models how to do that.

So that's why we said we're going to be pretty conservative. The 5 is a pretty conservative estimate of what we might be able to -- what we might ship in '24, and then we take that revenue. Clearly, we're looking for '25, we're looking at a number north of those 5 in terms of shipments that we expect for 2025. That's the reason why we were so conservative, if you like, on the number of 5 in 2025. So more shipments expected in '25.

And that's also why we said for the 20, we set medium term. Medium term is a little bit after '25, so that could be in the '26, '27 time frame. That's where we think the number of 20 will be necessary to cater the demand that we have there.

I mean the buildup of the order book is progressing very nicely, not just on the 5000 tool, which will initially be used for research purposes. But as we mentioned to you, now we also have clear orders for the 5200 coming in last quarter and also coming in, in this quarter. So the order book there is building up nicely and in line with our expectations.

And our next question comes from the line of Krish Sankar of Cowen and Co.

I had 2 of them. Peter, just to follow up on the macro question. It seems like in the last 3 months, your FY '22 outlook has been pretty much stable. Still in supply terms, demand has actually grown. Kind of curious, based on your forecast, it looks like it's (inaudible) through 2025, but the market seems to be worried about macro recession. So I'm kind of curious, do you think investors are too negative? Or do you think the semi cap industry or semi industry is too positive and the truth is somewhere in between?

Well, I'm with the company 23 years and I've always tried to read the minds of the investors, and I always got it wrong. So don't ask me that question. So no, it's a place that can only read the minds, not only read the minds, but talk with our customers and see where we are today. And the issue is, and I tried to mention this before is that the demand that we're currently seeing comes from so many places in the industry. Technology-wise, market-wise, geography-wise, it's so widespread that we have significantly underestimated the -- let's say, the width of the demand. And I think that I don't think is going to -- is going to go away.

And it just -- it's an anecdote, but I met a very large -- the executive of a very large industrial company, a conglomerate, last week. And actually, they told me that they're buying washing machines and to rip out the semiconductors to put them in industrial modules. I mean, that's happening these days.

Now we could say that's an anecdote. But to be honest, it happens everywhere. It is -- like I said, it is -- it is 15-, 20-, 25-year-old semiconductor technology that is now being used everywhere. It's got -- it's all driven by IoT-type applications. I don't think that is a fluke. I don't think that is real. And I think our customers are currently demanding, and I said it before, 600 deep UV systems this year. Now if we can do 60% of that, I'm pretty happy this year.

Now that will also grow next year. We will ship more than 375 deep UV systems next year because we're pulling in that 2025 target that we actually had and then putting out all the stops to ship more systems.

Now the demand could go down from 600, yes, another 25%, which is still above our maximum capacity. So you tell me whether we're too optimistic. I'm just looking at the data points and the data points just point into a market that is significantly short of semiconductor manufacturing capacity this year and next year. And what macroeconomics will do, I don't know. We'll just have to wait and see. But currently, we see no signs of any weakening in our customer base, 0. And even if the demand weakens, there's a big gap between the demand and our capacity.

Got it. Super helpful, Peter. And then a quick follow-up. You also articulated how the strong demand for trailing edge technologies. And when I look at the China sales, it's been a record high for the last 2 quarters. From your vantage point, is there a way you can segment it and see if China is actually buying for true end demand? Or do you think they're just stockpiling deep UV tools? Any color into the thought process there would be helpful.

In fact, Krish, I don't think that China is moving at a faster pace in general than the rest of the world. Also for China this year, we're expecting about 20% growth, which is similar to what we see for the rest of the company. So I think China is in line. The fact that you saw such a high percentage for China in the Q1 sales in particular is because, as you know, we only had 3 EV systems in the total mix. So as a result of that, the China sales as a percentage of the total was so high. But I think that's an anomaly that will rectify itself in the course of the year. So we're still expecting domestic China to grow 20% just as the rest of the customer base.

And an answer to your question, yes, we do believe that what they're adding in terms of capacity is being used. It's being used for manufacturing. So it's not like they're piling up lithography tools, they're really using these as well.

Yes. And we can see that just based on the utilization numbers of our machines across the entire installed base is Logic and Memory. It's never been as high as it is now. So if they're stockpiling anything, they're not stockpiling our machines, they're stockpiling semiconductors, which when you look at the shortage and the fact that people are desperately trying to get their hands on mature type of semiconductors, I don't think they're stockpiling anything.

So it's -- they're being used and that's what we see. As a matter of fact, one of our Chinese customers, big customer told us, "Hey, I'm sold out. I've sold out all the wafers and my -- and the wafer capacity that I'm adding sold out until the end of next year, '23. Not '22, '23."

And our next question comes from the line of Adithya Metuku of Credit Suisse.

Firstly, again, just to follow up on Krish's question there. You -- when you say you're going to ship -- you see demand for 375 DUV units next year, we're all kind of on the investor side playing these war games around what could happen next year to see demand really go down. And if you look at consensus next year for you, consensus modeling DUV units to decline next year given these concerns around the recession.

My view is, even if you saw a demand decline, your customers are unlikely to cut tool orders. But in your view, can you give us some scenario as to what would need to happen for customers to really cut orders for next year given the lead times, et cetera? If you have any color around that, that would be super helpful for investors.

And secondly, I just wondered if you could give us the figure for High-NA EUV tools in your orders this quarter and in the backlog that would be so helpful.

Well, what could happen? If we would do next year, let's say, I think it's going to be over 375 because we're pulling out all the stops, like I said, which will -- that basically be our absolute maximum capacity. I mean, we're doing overtime, hiring people, whatever. There's not only us, it's particularly in the supply chain. So that's where we want to go, yes.

Now what needs to happen for that demand to go down, the demand for next year is 600 units. So you need to lose 35% to 40% of your demand. And then you hit our capacity, our maximum capacity.

Now of course, we can all dream up horror scenarios whereby the world collapses and all kinds of crazy things happen, it's -- we'll deal with it when we are there, when it's such a disaster. But for, let's say, a macroeconomic slowdown, well, needs to be significant to take 35% to 40% out of the demand curve. And then we're still shipping more than what we shipped this year.

So yes, no. And on horror scenarios, I don't know, I'm not the kind of guy that you should ask the question, but probably all the people better suited to do that, but that's the way that I look at the world.

Whatever. He was just looking at me, but I've not going way in there either. So I will try and give you a non-answer to your question on the High-NA orders, Adi. So we did disclose that we have 5 orders for High-NA 5000, for the 5000 system. That's an R&D system. We've also indicated we have multiple orders for the 5200, and we've also said that we have these orders from 3 Logic customers and 2 Memory customers, and that's really what we can share. Our customers for a host of reasons do not want us to be specific as to the order intake for High-NA for the commercial system. So that's why we stick to this to the labeling it multiple.

Understood. That's very clear. Just a quick clarification on that 375 number. Is that shipments? So will that not be revenue recognized? Will you recognize revenue on all of those tools? Will there be some different...

These are deep UV systems and that generally recognized upon shipment. So I think it's the 375-ish which we're trying to, like I said, to push a bit higher that's on shipments -- revenue, revenue shipments.

I mean if you talk about next year, if you talk about next year, Adi, in all likelihood, you're going to see a bit of a balance in terms of fast ship, right? So at that stage, I think the impact of fast shipments on the delta between output and revenue should be very small.

And our next question comes from the line of CJ Muse at Evercore ISI.

I guess first question, it certainly sounds like you're pulling in High-NA or have greater expectations sooner. And so I guess curious, how are you on the supply side ramping optics in particular, particularly given the supply constraints we're seeing on DUV and low NA UV?

And then the second part of the question for High-NAs. How are your customer conversations evolving? And do you think we could see high-volume manufacturing insertion as early as 2 nanometer?

Yes. I think High-NA, sooner. Generally, that's not the case because it's -- like you pointed out, I mean, it's a supply chain issue. I mean it's a different machine. It's sort of different modules that needs to be built. So this is really focused on executing according to plan, which doesn't give us a lot of room to do more and sooner. So it is really our shipment plan.

And Roger just talked about the difference between revenue recognition and shipment. But the shipment plan hasn't changed that much because it's all very much linked to the capability of the supply chain to actually forward that. But I think in HVM insertion, we always have 25, 26. I think that's very likely what it's still going to be the case.

Now having said that, it could very well be that you see the first 2-nanometer chips being produced in 2025 as a kind of a first set of devices that show the capability. But I think HVM is probably the '25, '26 time frame.

Very helpful. And then just a quick follow-up. As you contemplate the capacity additions, can you kind of walk us through the impact to your cost structure, both COGS and OpEx?

Yes. I think we'll do that in Q4 in a bit more detail. But I think, yes, I mean, we need to -- we, us and our supply chain and especially the supply chain, they need to invest in CapEx. Now CapEx for us is not the major issue in our cost of goods. And I think for our suppliers, they, of course, need CapEx. But I mean, it's CapEx on a higher volume. So I don't think currently, it's a bit early and then would expect big impact on the margins or on cost, Roger?

No, exactly, CJ. We'll come back to that in Q4 when we talk a little bit more. But to Peter's point, I wouldn't expect a significant impact on the gross margin expectation that we gave for 54% to 56% for 2025. I think it will still be in that ballpark.

And our next question comes from the line of Aleksander Peterc of Societe Generale.

I'd just like to come back briefly on these numbers you gave us for 2025, the capacity expansion you're looking at. So the 90 EUV and 600 deep UV. Is that your estimate of what unconstrained demand will be by 2025? Is that what the industry needs and you need to guess to an equilibrium between offer and demand or something a bit more reasonable terms of lead times? So in other words, if suppliers come back to you and say, "Yes, this is possible. It's going to be feasible, we can get there." Is that the kind of number you then drive to and that will become your revenue guidance? Or is this too optimistic and we should [cover] that in any way? That's my first question. I have a quick follow-up.

I'm probably going to answer that question for the third time. But I think it's basically what we feel the capacity is needed to actually be available in 2025 for the rest of the decade. Now if for whatever reason, the demand that we could see today and for next year would still be the same demand as we see -- that we would see in 2025, we would need it in 2025 to actually ship, but that we don't know that. I mean this is really driven by how the market is by 2025. It could be -- it's the same as it is today. It could be higher, it could be lower. We don't know that yet, but this is definitely something which we are going to discuss in Q4 where we show you a couple of scenarios where we can show that this could happen. If this happens, then you would see that result. If that happens, you would see a different result.

But it's really about the capacity buildup. So it's not that we are saying, well, the unconstrained demand for 2025 on us is 600 DPV and 90 EUV. We will have the capacity there that if that would be the case, we could do it, but we don't know whether it's going to be the case, but we definitely need it. We definitely need it for the second half of this decade given the growth of our customers. And let's keep our fingers crossed and see what 2025 brings us.

Okay. That's very clear. And just a quick follow-up just on 3D DRAM. In light of the multiple orders you now get from DRAM customers, not just 1 but 2 customers now, could you share your thoughts on the potential timing of 3D DRAM and that 2025 number we saw being branded about? Is that now a clear indication that's actually happening or later?

Well, I think we've made a point a couple of times. 3D DRAM is an option. We think it's not going to be 2025. It is an option. It could hit somewhere in the second half of this decade. But we also know, and we know that because our customers are telling us this, and actually, our customers are telling this in a public way, that whatever their road map is, what is clear is that EUV and High-NA has a particularly important role to play in that road map. And the latest, very clear and public data point, and that was Samsung at their conference in November of last year, where they clearly demonstrated that EUV was there, was on their road map for the next 10 to 15 years. That's what they indicated. They said even to go to DRAM, I think it was 8 nanometers or 5 nanometer that I talked about, less than 5 nanometers. They need EUV and High-NA in order to get there.

So that's what's important to us. That's what's driving our road map also on the -- on Memory. Could that go in conjunction with 3D DRAM? It's possible, but it will be always a combination. That's what our customers are telling us and that's what we stick by.

Our next question comes from the line of Rolf Bulk at New Street Research.

I had a question on your Memory business. The order intake of memory in the quarter was again very strong, close to EUR 2.4 billion. And I was wondering, can you talk about what is driving this order flow? Is that NAND or DRAM? And is it primarily for capacity being added this year and next? Or is a large component also EUV and High-NA orders for long-term capacity additions?

Yes. I think it's largely these capacity additions in Memory are for current capacity demand. I mean, like I said earlier, the utilization of our machines in the memory space is very, very high. So we even say it's the highest point ever. So if there's a high-teens, 20% bit growth number expected for this year, you just need to add capacity. So I think the bulk of what you're seeing there is just to add capacity.

And yes, there will some of it will be higher, so you will -- that's not translated into an immediately bit capacity. But with EUV, it is, because don't forget, EUV is now HVM technology also in DRAM. So most of it, the vast majority of it is for current capacity additions, which I fully understand when I look at the utilization rates.

And as an unrelated follow-up, if I may, could you please give us an update on your EUV service business? How large is the business today on a run rate basis? And what growth do you expect this year? And also, where do you stand on gross margins versus the corporate averages?

Yes. So the EUV service business is around EUR 1 billion this year. And the gross margin, we talked about gross margin in the past. So gross margin was around 30% last year. We think it's going to be around 35% this year. We talked at the beginning of the conference of this call, we also talked about some pressure on labor cost. So obviously, that's heading us also a little bit on the service side. But still, we believe 35% gross margin is possible this year. And as we said before, we're trying to get to approximately 55% in the '24, '25 time frame. That's the trajectory that we're on.

And our next question comes from Houri of ODDO BHF.

Yes. Actually, I would like to ask a question again about the -- your supplier to really understand what you are talking about when you say you are discussing with them. Does it mean that they are all okay to increase their capacities? And if you're -- and you're just talking about the way to do it? Or if there is some discussions to be held about the possibility to raise their capacities?

No, I think we just had our, let's say, supplier day a couple of weeks ago and was very clear that all our suppliers confirmed that they said this is the target that we should go for. It's really a matter of them creating the executable plans to actually get to that capacity number by the time frame that we asked.

And that has to do with the plans that they need to make to build square meters, to build factories to hire people, to order specialized machinery. That needs all confirmation in their supply chain. So it's a supply chain of the supply chain. They need to find the builders. They need to get the square meters. They need to get the permits.

So just a very practical stuff that if you want to expand capacity, how quickly can you do that? And what are the normal, let's say, business hurdles that you would have to pass in order to make sure that, that happens. This is the type of discussion that we're having. And this is -- that's basically -- all these things are gating the conformation.

So they work on this and they will get -- as we go along over the next couple of months, quarters, they will get all the confirmations and say, yes, we can do it. Probably not 100%, but 90% or 85%, yes, we can do it, and then will probably confirm that --- not very likely, will confirm that to us and says, yes, we're going to be there at that point in time when you need us. That's the situation.

Okay. And on your side, can you communicate already on the cost of such an expansion?

I think most of the cost will be in the supply chain. I think, our CapEx is -- our CapEx on the balance sheet, it's roughly 10% of the total balance sheet. I mean, it's not huge. So yes, it's a big number. But when you look at the size of the company, it's not the biggest cost item. It's in the supply chain. So I think it won't be an issue for us. And I think when you look at the upside opportunity of having an extra capacity and the forecasted growth, then it's a relatively fast payback time.

All right. Sorry. We have time for one last question, and I apologize for technical difficulties at the start of this Q&A. If you were unable to get through on this call and still have questions, please feel free to contact ASML Investor Relations department with your question. Now operator, may we have the last caller, please?

And that will be Francois-Xavier Bouvignies from UBS.

I just have maybe 2 quick ones. One is on the -- basically, you work out dealing with your suppliers. You mentioned, Peter, that you don't bear a lot of cost to add capacity, but these suppliers will probably have a lot of risk or more risk to you than you in terms of balance sheet and under (inaudible) charges and you want to build this capacity well ahead of time for the second decade.

So my question is, do you expect to share somewhat the risk of the investment with your suppliers somehow, I mean, with maybe a payment or anything like that we should expect to help your suppliers to get to this kind of level and bear the risk of overcapacity?

And the second one is you talk about suppliers, but what about the peers, your peers as well? It seems that your capacity expansion is quite big, at least since last year. Do you expect the peers for other parts of the supply chain to increase as much? Is there a risk here that you build too much and the others don't, and therefore, you can't unlock this value?

Yes. On the first question, yes, I mean, we've done that in the past also. I mean if a supplier, let's take ZEISS, for instance, it comes to us and says, hey, can you do some shared risk investment and it goes beyond what you would say, the let's say, average or normal financial capabilities of that supply. We have done that in the past and we will do it again.

Having said that, over the last 10 years, our supply base has actually grown quite significantly also in terms of size and in terms of financial capability. So I think it's going to be limited, but if it's needed, it's needed and we'll do it. So I don't think it's going to be a major issue. So risk sharing is in our genes. So that's not the point. On the peers, we're coming out because our capacity lead times are probably the longest in the industry. So we go first. And I would expect our peers, if we do that, then the market is there and we believe the market is going to be there, they will follow. It's that simple.

All right. Now on behalf of ASML, I would like to thank you all for joining us today. Operator, if you could formally conclude the call, I would appreciate it. Thank you.

Thank you. This concludes the ASML Q1 2022 Financial Results Conference Call. Thank you for participating. You may now disconnect.