Welcome to the ASML 2022 Fourth Quarter and Full Year Financial Results Conference Call on January 25, 2023. [Operator Instructions]. I would now like to turn the call over to Mr. Skip Miller. Please go ahead, sir.

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 fourth quarter and full year 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 the 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 fourth quarter and full year 2022 results conference call. And before we begin the Q&A session, Roger and I would like to provide an overview and some commentary on the fourth quarter and full year 2022 as well as provide our view of the coming quarters. And Roger will start with a review of our fourth quarter and full year 2022 financial performance, with some added comments on our 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 want -- like.

Thank you, Peter, and welcome, everyone. I will first review the fourth quarter and full year financial accomplishments and then provide guidance on the first quarter of 2023. Let me start with our fourth quarter accomplishments. Net sales came in at €6.4 billion, around the midpoint of our guidance. We shipped 18 EUV unit -- EUV systems and recognized €2.3 billion revenue from 13 systems this quarter. Net system sales of €4.7 billion, which was again driven by Logic at 64% with the remaining 36% coming from Memory. Installed Base Management sales for the quarter came in at €1.7 billion, which was higher than guided due to additional upgrade revenue. Gross margin for the quarter came in at 51.5%, which is above our guidance, primarily due to the pull-in of additional upgrade business as well as an insurance settlement from ASML Berlin fire, which occurred in early 2022. On operating expenses, R&D expenses came in at €906 million, above our guidance due to higher depreciation. SG&A expenses were €280 million, higher than guided due to increased IT and recruiting spending as part of our headcount growth plan. Net income in Q4 was €1.8 billion, representing 28.2% of net sales and resulting in an EPS of €4.60. Turning to the balance sheet. We ended the fourth quarter with cash, cash equivalents and short-term investments at a level of €7.4 billion. Moving to the order book. Q4 net system bookings came in at €6.3 billion, which is made up of €3.4 billion for EUV bookings and €2.9 billion for non-EUV bookings. These values also include inflation corrections. Non-EUV bookings are a combination of deep UV and metrology and inspection. Net system bookings in the quarter were driven by Logic with 66% of the bookings, while Memory accounted for the remaining 34%. Looking at the full year, net sales grew 14% to €21.2 billion. EUV system sales grew 12% to €7 billion realized from 40 systems while in total, we shipped 54 EUV systems in 2022. Deep UV system sales grew 13% to €7.7 billion. Our metrology and inspection system sales grew 28% to €660 million. Looking at the market segments for 2022. Logic system revenue was €10 billion which is a 4% increase from last year. Memory system revenue was €5.5 billion, which is a 34% increase from last year. Installed Base Management sales was €5.7 billion, which is a 16% increase compared to previous year. At the end of 2022, we finished with a backlog of €40.4 billion, an increase of 67% compared to the end of 2021. Our R&D spending increased to €3.3 billion in 2022, as we continue to invest in innovation across our full product portfolio. Overall, R&D investments as a percentage of 2022 sales were about 15%. SG&A increased to €946 million in 2022, which was about 4% of sales. Net income for the full year was €5.6 billion, 26.6% of net sales, resulting in an EPS of €14.14. Improvements in working capital contributed to a free cash flow generation of €7.2 billion for 2022, mainly driven by customer down payments following the very significant order intake this year. We continue to invest in support of our road map and planned capacity ramp. Excess cash will be returned to our shareholders through a combination of dividends and share buybacks. With that, I would like to turn to our expectations for the first quarter of 2023. We expect Q1 net sales to be between €6.1 billion and €6.5 billion. Per customers' requests, we prioritized resources towards the acceleration of deep UV shipments at the end of 2022. As a result, we expect lower revenue in Q1 and higher revenue in the following quarters. We expect our Q1 Installed Base Management sales to be around €1.5 billion. Gross margin for Q1 is expected to be between 49% and 50%. The lower margin relative to last quarter is primarily due to lower upgrade revenue and deep UV mix effect. The expected R&D expenses for Q1 are around €965 million, and SG&A is expected to be around €285 million. The higher R&D guidance is primarily due to additional headcount and labor cost increases. These investments are in support of our continuous innovation as we further expand our deep UV, EUV and applications road map and, at the same time, work to improve our installed base performance. Higher SG&A is mainly due to additional headcount and IT spending. Our estimated 2023 annualized effective tax rate is expected to be between 15% and 16%. In Q4, ASML paid a quarterly interim dividend of €1.37 per ordinary share. The third quarterly interim dividend will be €1.37 per ordinary share and will be made payable on February 15, 2023. Recognizing this interim dividend and the 2 interim dividends of €1.37 per ordinary share paid in 2022, this leads to a final dividend proposal to the general meeting of €1.69 per ordinary share. In Q4 2022, we purchased around 0.6 million shares for a total amount of around €300 million. ASML announced a new share buyback program during our Investor Day on November 11, 2022, to be executed by December 31, 2025. We intend to purchase shares up to an amount of €12 billion. With that, I'd like to turn the call back over to you, Peter.

Thank you, Roger. And as Roger has highlighted, we had a year of record sales in a dynamic environment. Demand remains strong and we finished the year with a record backlog. And looking to 2023, there continues to be a lot of uncertainty in the market due to a number of global macro concerns around inflation, rising interest rates, recession and the geopolitical environment, including export controls. Customers are still seeing demand weakness in consumer-driven end markets, the most notable with PCs and smartphones, with some indication of softening or lower growth rates in data center demand, while the demand strength continues in other markets such as automotive and industrial. Customers are telling us they expect a rebalancing of semiconductor inventories over the first half of 2023, with business expected to rebound in the second half of the year. A potential driver of this recovery in the second half of the year could also be the post-COVID opening of China. This could have a positive effect on both supply and demand. To help rebalance inventory levels, customers are running their lithography systems at lower utilization levels and some have also lowered their CapEx plan for this year. Based on this, we concluded that most of our customers have made the assessment that the duration of a potential recession is significantly shorter than our average delivery lead time. On top of this, lithography investments are strategic in nature, which means that the demand for our systems remain strong. For instance, this year, demand still exceeds our capacity, and we enter the year with a backlog of €40.4 billion, so our focus will still be on maximizing system output. We've experienced several quarters of very strong bookings, which now provides backlog coverage significantly beyond 2023, which is almost twice the expected 2023 system sales. Based on discussions with our customers and continued improvements in the capability of our supply chain, we are planning to increase our output capability this year. We're planning to ship around 60 EUV systems and around 375 deep UV systems in 2023, with around 25% of the deep UV systems to be immersion. We still plan a significant number of fast shipments this year, which under the current way of working will result in a similar amount of delayed revenue out of 2023 that came into 2023. Looking at the growth of the business for the full year 2023 compared to 2022. We expect EUV revenue growth of around 40% and non-EUV revenue growth of around 30%. And for the Installed Base Management business, we expect year-over-year revenue growth of around 5%. And as we are coming off a strong growth year in 2022 and customers are adjusting their utilization levels, we currently expect to see a slightly lower demand in our upgrade business in 2023. In summary, for the full year 2023, based on how we see the world today, we expect another year of strong growth with a net sales increase of over 25% and a slight improvement in gross margin. On the geopolitical front, as it relates to export control, this continues to be a geopolitical discussion with different government entities, a process where ASML is obviously not in control although we continue the discussion with governments to make sure that consequences of proposals are well understood. As of today, export control policy related to lithography equipment has not changed. We are still not able to ship EUV systems to China, but are able to ship deep UV as well as metrology and inspection systems to China. As our Prime Minister recently stated, this is a multinational discussion, not only with the U.S., but with several countries. He reiterated that multiple companies in the semiconductor industry, including their supply chains, are involved and that the matter is complex and sensitive. It needs careful handling with precision. And he reminded us there is a lot of interest at stake, and it's important to find the right balance. We will therefore not speculate about the possible outcome but will have to wait for the outcome of ongoing government discussions. Looking longer term, we talked at our Investor Day last year about the global megatrends, where the broadening application space is fueling demand for advanced and mature nodes. Secular growth drivers in semiconductor end markets and increasing lithography intensity on future technology nodes are driving the demand for our products and services. ASML and its supply chain partners are actively adding capacity to meet future customer demand as confirmed at Investor Day, with our capacity plans of 600 deep UV, 90 EUV low-NA systems by 2025-'26 and 20 EUV high-NA systems by '27-'28. Also presented during our Investor Day last November, we see an opportunity based on different market scenarios to reach an annual revenue in 2025 between €30 billion and €40 billion and in 2030, an annual revenue between €44 billion and €60 billion. Part of our long-term growth opportunity, we also remain committed to our ambitious ESG sustainability goals. On February 15, 2023, our 2022 annual report will be published. As part of this report, we plan to provide you with an update on how we collaborate with our stakeholders to deliver on the ambitions of our ESG Sustainability strategy, which we can summarize as follows: our ambition is to achieve carbon neutrality with net zero emissions in our operations, Scope 1 and 2 by 2025; we aim to achieve net zero emissions in our supply chain, Scope 3, by 2030; and net zero emissions from the use of our products by our customers, Scope 3, by 2040. In addition, our goal is also to have zero waste from operations to landfill and incineration by 2030. From a social perspective, our ambition is to ensure that responsible growth benefits everyone. To maintain our fast pace of innovation and ensure our long-term success as a company, we need to attract and retain the best talent and provide the best possible employee experience. We aim to be a valued and trusted partner, improving the quality of life for all people in our communities. In summary, while there's a lot of near-term uncertainty in the current environment, our customers' demand for our products continues to exceed supply. We're working to increase our capacity to meet our customers' future demand, and we're fully confident in the opportunity this provides for our future growth. With that, we would 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 your final instructions and then the first question, please?

[Operator Instructions] Our first question comes from the line of Joe Quatrochi at Wells Fargo.

I was wondering in your 2023 outlook if you could kind of give us any color on how you're thinking about Logic versus Memory growth this year.

Yes. I mean we've not made that -- Joe, thanks for the question. We've not been explicit on that. We think that in the current environment, with all the dynamics and all the uncertainty that are out there and also the fact that we still look at 2023 as a year where we are supply constrained. We don't think that, that makes a lot of sense to guide around that. So that's why we said we're going to guide on [indiscernible] so EUV versus -- EUV and non-EUV, but we think it is not very constructive and meaningful to provide guidance on Memory versus Logic. I mean clearly, if you look at the recent order intake, you do see that Memory is far from that. As a matter of fact, you see Memory actually picking up in the quarter and quite healthy in the year. And that tells you that the Memory buyers are also making strategic investments because they recognize that if at a certain point in time, the market is going to come back, they need capacity. So that's what you see in the order intake. But again, for the full year, given an indication of the distribution over Logic and Memory, we thought it was a meaningless action in light of the dynamics that I just talked about.

Got it. That makes sense. And as a follow-up, you talked about the inflation-related adjustments in your bookings. Are you able to reprice kind of your entire backlog or portions of your backlog? And then is there any sort of quantification that you can help us with on just the average increase of ASP that we should be thinking about in our models?

Yes, Joe. As you know, we've been through this before. We're having good discussions with customers. As you know, legally, the way the backlog is construed, we've agreed on a price. So this is just a matter of discussions with customers about a fair distribution of the burden. We're fairly advanced in that discussion. A number of customers have already made commitments to us on how they're going to contribute, with most other customers who are fairly advanced and are -- we think that we're going to find a solution. It's only a minority of customers that are not open to this conversation. So the lion's share is open to the conversation. But it really is in the spirit of finding a fair distribution of the inflation burden. So that has helped. And we're only putting into the backlog those inflation adjustments that have been explicitly agreed with customers. So more of that is to come because, as I said, we have a number of customers who are still in negotiations. So all of that is coming. I think the way it's going to pan out without really quantifying it, but just to give you an indication, as you know, last year, we said that we were having about 1.5% drag on the gross margin coming from inflation. I would say most of that is -- we expect to recover during the year. So most of the inflation that we incurred over 2022, I think most of that we will recover. But of course, there will be inflation in 2023, and that will remain a drag on the gross margin. So the gross margin impact of inflation will be less so than it was last year. But in comparison to 2021, you will still see a bit of a drag on the gross margin coming from inflation.

Yes. I think on your sales price you want to put into your models, I mean we guided EUV sales up with 40%, non-EUV with 30% and then installed base with 5%. I mean that's in the end where you need to end up with if you're going to change your model and you come to a different outcome, then something went wrong in your calculation. So...

And for the people that weren't really carefully listening, what that really means in terms of ASP for EUV, we talked about it before. Originally, we were looking at €160 million. We've then been talking about €165 million to recognize also increased functionality, I think, with the increases on ASP on the inflation. I think it's good to go somewhere between €165 million and €170 million. I think that's, on average, I think, the right way to go.

And our next question comes from the line of C.J. Muse at Evercore ISI. C.J. Muse: I guess first question, Peter, I was hoping you could take a step back and maybe talk about your discussions with customers. Obviously, reducing utilization to clear inventories. But at the same time, given your lead times, you're continuing to have great visibility well beyond 2023. So I guess, can you kind of speak to the moving parts there? And I'd love to hear kind of your thoughts on your visibility for all of 2024.

Yes. Good question, C.J. Yes, I think in the discussion with customers, it's exactly what I said. It's very clearly, we concluded from the discussions that customers believe. And I think it's also you can corroborate those statements with their public statements and basically saying some of the customers feel that they see a recovery towards the second half of this year. And that actually means that they tell us, "Listen, we know you guys are short of the demand we put on you." So that means that we don't want to risk our strategic investments, which go beyond the first half of 2023, moving into '24 and '25. So this is what that -- we're really having this midterm to long-term discussion with them of what's needed. And that's why they keep saying, "No, we need those machine." Having said that, of course, last year, we kept informing you that the demand on us significantly exceeded our build capacity, sometimes to 40%, 50%. Now that demand did come down. I mean that's also clear that some of that demand disappeared because of the market situation. But the end result is that the demand -- the cumulative demand is still higher than what we can make. So this is where we are. And I think that is driven by what I said that I think the average expectation on the duration of the downturn of, let's say, working through the inventory is significantly shorter than the average lead time of our tools. I think -- and that gives us a lot of visibility into 2024 also. Customers give us orders throughout the year, very significant levels of orders, which actually have over €40 billion in the backlog, which is almost twice the system sales that we expect to have in 2023. So yes, we have the visibility. And it still means that the customer expansion plans on adding that capacity in 2024 are still very real. Otherwise, you wouldn't give us these orders with down payments. So this is the level of visibility that we have. Now all of course hinges on the macroeconomic situation. Finally, you could say the expectation that the duration of any recession, if it would come, would indeed be short. I mean that's the big question mark that is out there for all of us. But this is where we are today. And we're just telling you what the discussions with our customers are currently at, and they are exactly what I just told you. So this is the level of visibility that we have, and always just a matter of watching the macroeconomic situation. C.J. Muse: Very helpful. As my follow-up, just two quick housekeeping. What percentage of EUV bookings in the quarter came from Memory in December? And then secondly, how many e-tools do you plan to ship in 2023?

So the -- it's roughly 25%-75%. So that's true for the entire bookings. It's also true for the EUV tool. So about 25% would go to Memory, around 75% comes from Logic. And I didn't quite get your last question. I couldn't really understand that. C.J. Muse: How many e-tools should we plan to ship in 2023?

Okay. That's a very limited number, a very limited number of tools. Less than a handful, I would say, C.J.

Our next question comes from the line of Didier Scemama of Bank of America.

Quick clarification, if I may. I think there seems to be quite a lot of confusion around fast shipment. So maybe Peter or Roger, if you could set the record straight. Number one, is it fair to assume that your calendar year '23 guide does not include any fast shipments, either on revenue or on gross margins? And then number two, if that's correct, when do you think you could get clarification from your accountants that, that could become the norm? I mean is that something tied to the sign-up of the accounts for 2022? Or is that a completely unrelated decision? I've got a quick follow-up.

Thank you, Didier. Thanks for the question. I think this is helpful to indeed clarify that. So what you saw is that we're having -- that we had fast shipments for an amount of €3.1 billion at the end of 2022. So the revenue for that will be recorded in 2023. However, we also assume that a similar amount will go from '22 to '23 -- from '23 to '24. And what that means, Didier, is that the guidance or the more than 25% growth over 2022 that we've given you, that in fact, treats the fast shipment effect as neutral, right? So we assume that, that will be neutral, i.e., the amount that comes in to '23, the €3.1 billion, we also expect that to leave the year into '24. So it's neutralized for the fast shipment, so that's the way to look at it. The question on can you change the revenue recognition. As we mentioned before, there are 2 key elements in here. One is an accounting element, so to what extent can you get this done. But pivotal for the accounting treatment of fast shipments is that customers are actually upon shipments are going to accept the full risk on the tool. That's what is pivotal. So the conversation that we need to have with customers and need their final blessing on is that based on a far more limited testing protocol, because that's what fast shipment in essence is. It's a far more limited testing protocol where there's a couple of weeks of testing that we actually omit out of the sequence, that based on a far more limited testing sequence, they still assume the full risk of the tool upon shipment. So that's the conversation that we need to have. And you will appreciate that that's a conversation that, of course, customers need to really look into and become comfortable with. So that's not a 5-minute conversation. We're now done with the accounting analysis, so the accounting analysis in and by itself is clear, but it all hinges on the premise of being able to get to an agreement with customers on what I just told you. So those conversations were now starting. Cannot tell you when that is done. In all likelihood, I would say, by mid this year, we should have clarity on whether customers are willing to do that or not. It could also be that some customers are going to accept it, others could not. And then once we have that clarity, of course, we will share that with you, and we'll also clarify to you what that means. To the extent that we would be able to recognize revenue again upon shipment for these fast shipments, of course, that would be additional to the more than 25% revenue growth that we've mentioned.

Yes, that's very clear, Roger. And as a follow-up to that, so if fast shipments becomes the norm, is it reasonable to assume that your actual capacity for EUV actually is increased by 5 units because you shave off effectively a month of cycle time? So 60 divided by 12, 65 -- I'm sorry, 5, plus 5, so 65, plus whatever deferred from '22 into '23 if fast shipments becomes the norm. Is that the way to think about it?

I think that's directionally correct. That's directionally correct, Didier. And bear in mind, in the output that we had for this year, of course, we already started to have that benefit, right? So there's already that benefit. But then to the extent that we're really able to get all the supply chain issues sorted, get back to a normal cycle time. As you heard before, we're eyeing a regular cycle time of 17 weeks. If we're then able to shave off 3, 4 weeks of testing, then that's the unit number that you could see it increase with. Correct.

Brilliant. Congratulations. Not too bad for a recession.

Thanks, Didier.

And our next question comes from the line of Francois-Xavier Bouvignies at UBS. Francois-Xavier Bouvignies: So the first one is on, Peter, you mentioned in your remarks that the demand is still above what you are able to do in terms of supply. So in a situation where -- and it's quite often a market share shift within your customer base, in deep UV and EUV. How do you plan to treat that? If you see pushout, let's say, by a couple of quarters so it slips to 2024, do you keep the slot for those customers? Or are you going to reallocate directly to the one that can take it straight away? So just how do you deal with pushout in this kind of allocation market if some customers have some delays? That would be interesting to have your point on that. And I have a quick follow-up.

Yes. Yes. I think let me be very clear. I mean we are capacity constrained. So if a customer says, "Sorry, we want to reallocate, I want to push it back to 2024," then that slot will be taken by a customer that raises their hand and says, "Ship it to me." And that's what we will do. So that slot will be taken. And that means that the slot that the customer says, well, I have an order for -- okay, please pull it -- push it back. We just need to negotiate when that pushback is because that pushback could fall into a period which is already fully booked. So this is then fine, if you don't want to do it then. It actually means that it's a kind of a negotiation when it's the first open spot in 2024. That's how it works. So yes, we are just filling it up, and that means customers need to accept the fact that, that tool will not stay here in our premises, in inventory or in work in process until they can ship it. No, it will go. Francois-Xavier Bouvignies: Okay. That's very clear. And my quick follow-up is on memory China since we had the restriction on memory China not impacting a lithography based on your previous comments. I mean do you see any pullback or spend pullback on the memory China? Or you still see strong demand on that front?

Yes. I think it's not only applicable to China, what I'm going to say. I mean it's applicable to all our Chinese customers and, as a matter of fact, non-Chinese customers also. You need to realize that -- and you actually know that, Francois, I mean, planning and building and executing a new fab is a matter of years. So that means that anything that we shipped this year has been planned in 2021, 2020. So these plans are there, and we just execute on those plans. Now having said that, it's also clear that we have a demand -- we have a capacity shortage. So we have an over-demand that indeed, when we reallocate, which is true for Chinese and non-Chinese customers, where we have a bit more space to reallocate, then we will reallocate because the demand has been higher than our capacity. So yes, that's true for Chinese and non-Chinese customers. We treat all the customers the same in that way. So I think we haven't seen any acceleration with just the execution of the plans because those plans are planned for years. You cannot just think of as semi fab and it exists in 6 months' time. So this is why it's just planned. But yes, you will probably see open spots, if you could call it this way or pushbacks of it. There are many customers that raise their hands, it's not only China, not only China, it's across the globe.

And the demand from China remains very strong, right? So we reported last time on this call that the percentage of China in the backlog was around 18%, and that's remained throughout the quarter. As a matter of fact, it's even gone up a little bit. So the demand from China still remains quite strong.

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

I have two of them. First one, Roger or Peter, just a clarification. On your backlog, can you give us some color if you did not -- if you did, I forgot, I missed it, between EUV and deep UV, Memory and logic foundry?

Yes. So in the backlog, it's around 25%, 75%. So if you look at the backlog for EUV, around 25% of that is for Memory customers, around 75% of the EUV backlog is for logic.

And of the backlog, the total backlog, EUV and deep UV?

Of the total backlog, let me see, it's around 55% of that is EUV.

Yes.

Got it. Got it. And then as a follow-up, I just wanted to touch again on the fast shipment. I understand last year and into this year, customers are scrambling to get litho tools and, therefore, they're willing to take fast shipments. In an ideal world, you'd like fast shipments, too. But if things do slow down a little bit, do customers really want fast shipment? Wouldn't they rather have you test the tool in your factory before you ship it to them?

Well, customers, you could argue, they don't care whether it's a fast shipment or not a fast shipment, they want a tool at a certain moment in time. And that's whether it's fast ship or a regular ship, they don't care. Now we did the fast shipment because we're late, and we were late and the demand was higher than what they -- what we could make. And this actually is a driver for still the fast shipment this year because our demand is higher than what we can make. That has changed. So if the customer says you can ship the tool, as long as they know the date of which the tool will arrive, it can be installed and can be signed off. That's the date that they're actually interested in. And as we see it today, there is still a higher demand than what we can make. So it's the same situation.

And our next question comes from the line of Alexander Duval at Goldman Sachs.

Congrats on the strong results. Firstly, I wondered if you could just update us on how much bigger demand is than supply. You're obviously still positive on the demand backdrop, even though you talked about utilization and machines having gone down. I think in the past, you talked about more than 30% excess demand versus supply. So is that still the case? And then secondly, a housekeeping item, just on OpEx. You've guided for 1Q, which is perhaps a little bit higher than expected. Can you help us put in context what that means for full year OpEx? And specifically, a bit more color on what you're investing in and what that means for harvesting future opportunities and your margin potential?

Yes. I'll do the first part of the question, and Roger will take the second part. I think on the 30% excess, the amount of supply actually is more like 50%. So now that has come down, but it's still significantly double digit above our capacity. So -- and like I said earlier in an answer to an earlier question, yes, of course, we also see the reflection of the demand curve because of the weakness in, for instance, the consumer demand clearly. So some of that over-demand has gone away, but it's still there. And it was actually not 30%, it was more like 50%. And it's come down now to -- it's still significantly double digit. Roger?

Yes. Alexander, on the OpEx question, I think the numbers that we gave you are, I think, a pretty good proxy for the full year. I think what I -- if you take them together, it's close to 19% that I would model for OpEx, so for SG&A and R&D combined for the full year. Of course, we had significant hiring in the course of Q4. Those people are added to the year, if you like, in terms of headcount. Of course, we had increases in wages. And of course, that kicks in from Q4 to Q1. But that's the rate I would assume for the full year. In terms of what do we get for that, very good question. So on the SG&A side, obviously, this is in line with the growth of the company. So you continue to see us operate around 4%-ish SG&A percentage of sales. As you also would have seen, if you look at the 2025 and 2030 scenarios that we've talked about, you see it coming down a little bit, and that's clearly the intention to get some operational leverage there. But for the fast-growing company that we are, investments like these are necessary to make sure that the organization has run in a very professional way, that security is up to scratch, that the IT support for our professionals is up to scratch, et cetera, et cetera. So that's what we're doing on the SG&A side. On the R&D side, this really is the very, very aggressive road map that we have on all cylinders. So this is deep UV. And as Peter mentioned earlier on, this is not just a deep UV on the immersion side where we continue to drive immersion but also, this is on the -- on the dry side, KrF. So for KrF, we continue to drive road maps for smaller customers on the XT platform for larger customers operating large fabs on an NXT platform. We're looking at i-Line road maps. We have a very broad road map that we have on metrology and inspection. And obviously, we continue to drive both low-NA and high-NA. So it's a very extensive road map, more extensive than we've ever had before in the history of the company. And of course, all of this is conducive to the goals that we've been talking about and the scenarios that we've been talking about at the Investor Day for 2025 and 2030. So yes, definitely significant investment, but definitely rewarding if you look at the potential that we see in the market for those products.

Yes. I think it's what Roger said. I mean our relentless focus on innovation and R&D has paid off handsomely in terms of value that we could create with our products, but also the extension of our market shares, the integration of our product offering into a kind of a holistic approach and value to our customers, that will only increase going forward. I mean things in semiconductor manufacturing will not get easier. They are getting more complex. It actually means that the entire product portfolio focused on patterning, patterning the 2-nanometer, the 1.5- to 1-nanometer and beyond, that is going to be significantly valuable to our customers. And that means we need to spend on R&D that Roger just talked about to make sure that all the ingredients for that value recipe are actually there. And that's what we need to do. And it has served us extremely well in the past, it will serve us extremely well going forward.

And our next question comes from the line of Mehdi Hosseini at SIG.

A couple of follow-ups. Peter, based on your conversation with customers over the past month or 2, has anything changed with your assumptions since the Analyst Day? I hear -- what I hear from you is pretty much the same as what we heard in November, which is also consistent with what you said in October. On the surface, it seems like there hasn't been any change to the customer zone or how they're planning despite the fact that end market demand is weaker. And I'm just wondering if there's anything else you can share.

Yes. So I think what we definitely see in the discussion with the customers, they are addressing the short-term challenges, it's just clear. I mean they basically say, "No, we need to -- we see inventories rising. We need to rebalance the inventory. So how do you do that?" So you just lower the utilization of the tools. That's short term, they're very clear. They're also very clear about their confidence in the long-term growth trajectory of the industry and of the need for significantly more semiconductors in all kinds of applications. I think it's without any exception, customers are talking about the medium- to long-term growth trajectory of the industry. That hasn't changed. It is also why, of course, there are short-term concerns like always short-term concerns. They say, "What am I going to do short term that will not impact my long-term targets?" That's basically the question, what they have. That's why, yes, some of them have adjusted their CapEx plans, but they say, "What is absolutely essential to secure our long-term growth path?" We will still do, which of course, this is why litho is a very strategic tool. This is why the litho orders keep coming. So in that sense, yes, things have changed short term, and they are reacting. But longer term, and in the CMD, the Capital Markets Day, we didn't talk about Q4 or Q1. We talked about 2025 and 2030. And that's exactly the discussion we're having with customers. I would even say that the longer-term road map this time with our large customers have intensified to a level we have not seen before. So like I said earlier, it's not getting easier, it's getting more complex. But it also means that the [cooperation] models that we have with our key customers are actually intensifying. And that just confirms not only the technology road map, but it also confirms the capacity plans that they have and the confidence we both have in the growth of this industry. So in that sense, Mehdi, nothing changed in 2 months' time.

And just quickly regarding your calendar year '23. If the non-EUV revenue is growing by 30%, how should I think about the mix of metrology and inspection?

Yes, it is in there. Yes, I think it's the 30%. I think it applies roughly to all of it. I mean you have to remember that our metrology tools have a very strong attach rate to our deep UV tools. So yes, they're in line. So it's for both.

And our next question comes from the line of Amit Harchandani at Citi.

Amit Harchandani from Citi. Two questions, if I may. My first question again goes back to the non-EUV business. The significant growth that you have guided for in 2023, which admittedly is aligned with some of the points you made at the Capital Markets Day, but clearly, that's where I feel investors also need the greatest -- need to see the greatest level of comfort. So could you give us a sense on the DUV drivers, your level of confidence in the demand, the level of discussions you are having around lagging edge? And simply reverse engineering your 30% guide does suggest that you'll get pretty close to the 375 capacity number that you've talked about. So any thoughts on DUV would be appreciated. And then I have a second question.

Okay. Yes, I think you have to realize, and it goes back to what I said earlier, that the biggest shortage in terms of demand -- or let's say, shortage in terms of capacity or shipment was in deep UV. So there is a backlog of deep UV investments that were planned by customers that we're now able to fulfill. You should not underestimate the size of the gap in deep UV, which is one element. And that's especially true for those geographic areas where we cannot ship EUV, but deep UV is the, as you could say, the workhorse for the mature and less critical semiconductors, i.e., China. So that space, let's say, is strong. On top of that, if you have more EUV sales, well, you don't make a chip for a semiconductor device with only EUV. I mean so if you grow the EUV base, we need a lot of deep UV layers. While the layer growth, nodal node is there, as it's always been there. So this is also an extra driver against the background of the fact that we can't ship enough. So this is what is happening. There are very few customers, which by the way, when you look at the industrial domain and you look at, for instance, some of the smaller IDMs, smaller -- I don't need to -- but it's important customers, but have just simply don't buy the same quantities as the 3 or 4 large customers. There in the industrial domain, they still call me and tell me, "Where is my deep UV tool?" And up to tools, multiple, yes? That's also happening. So it's the shortage that's still there. And it's, of course, I fully understand from an investor point of view to say, yes, but we also see the weakness in the consumer markets, we see inventories going up. But in a specific domain, anywhere between 45- and 20-nanometer, it's still short. And definitely in certain industrial domains like automotive and just industry in general. So this is still there. And this is why there are a couple of drivers. Like I said, it's, of course, China. It's -- EUV will need deep UV, and it's the deep UV OEMs that are in the sweet spot of where the shortage is. And then these are the reasons.

Understood. Very helpful, Peter. And as a second question, if I may go back to the topic of the gross margin evolution over the course of 2023. You've talked about a gradual improvement over 2022. At the same time, looking at your top line guidance, it seems like your ASPs are going to rise over the course of 2023. You obviously have greater operating leverage. Is the mix effect so strong, that's triggering this incremental step-up in gross margin? I guess I'm just trying to get a sense of what that incremental step-up is going to be in '23? And what are the headwinds which are offsetting some of the tailwinds you have highlighted earlier?

Yes, Amit, let me take that one. So many moving parts to the gross margin. We already alluded to one, which is the inflation one. So on inflation, it's important to recognize, yes, on the one hand, we will get some compensation. Yes, that will lead to an increase in ASP, but we're also getting inflation on the cost side, both on parts and on labor. So net-net, that will still be -- that will be an improvement over the 1.5% negative that we talked about last year, but it's still a drag on the gross margin in comparison to, for instance, what we had in 2021, but still an improvement over the year. So that's one element. The second element, of course, the fast shipment. Last year, we had a drag of 1.5% on the fast shipment. Of course, that drag is gone because as I mentioned, we expect the fast shipment amount coming into the year to also lose out of the year. So that 1.5% drag is gone. So that's also a positive in comparison to last year. Then volume goes up, both on the EV side and on the deep UV side, so that gives you a bit of a positive. Mix effect in the deep UV business is slightly negative because of the increase in the dry business that we also talked about. So that's a slight negative. And then there are 2 big ones to bear in mind. First off, as you would have seen, we're pretty cautious, if you want to call it that way, on the installed base business. And of course, you can assume that the service business will continue to grow. So then if we only guide 5% increase on the total installed base business, that implies that the upgrade business that we have in there is actually contracting. So that might -- that's up for speculation because that's the one that's hardest to plan for reasons that we talked about before. But we assume that the upgrade business will be lower and that, of course, is a big drag on the gross margin. And then the last element in the gross margin, and that's important to recognize, and we've been talking about this before, but I just want to emphasize it. We are incurring significant costs in our operations as it relates to preparing for both high-NA and preparing for the big step changes that we're making in our capacity. Do not underestimate that. That's a significant number. It's an even bigger number than it was last year. And of course, that continues until the point in time where the capacity is built and will be fully utilized or -- and/or high-NA is going to be utilized. So that's a drag that we continue to have, and it's bigger, as I mentioned, in '23 and -- than it was in '22. So those are the moving parts. That's why we say, if you add it all up, we're talking about a slight improvement over last year. I think it's important to note, Amit, that if you look at all of that, a number of the negatives will be gone, let's say, in the '25 time frame. And that's why Peter also reiterated in his video that we believe the 54% to 56% is still there because at that point in time, the capacity is there and should also be utilized. And the same high-NA at that point would be up and running. And also potentially, the installed base business at that point would be normalized. So I think that's the way to look at it.

And our next question comes from the line of [Sabas Aria] of JPMorgan.

This is Sandeep Deshpande here. I just want to touch base back again at the gross margin. I mean you've talked about the €3.1 billion of fast shipment being recognized in '23 and then potentially as much going into the following year. I mean does that have an impact on the gross margin? Or is it completely neutral to your gross margin? And secondly, in terms of the cost crossover on the gross margin, which was expected to happen in the second half of this year between EUV and deep UV, is that happening? And I have a quick follow-up.

Okay. So on the effect of the fast shipments, if indeed we're going to have the same number going into the year as we see leaving the year, then the impact is 0. Of course, the impact, and as I mentioned before, Sandeep, if indeed we're going to see some change in our revenue recognition, of course, that would be a potential plus for the gross margin. But we're not planning on that in the numbers that we've now shared with you.

Do you have another question, Sandeep?

Yes. No, sorry. My follow-up on that on the gross margin was the crossover.

Yes, the crossover. So EUV system gross margin at this stage around 50%, so more or less the corporate gross margin. What we said is there will be a point in time where you will also see EUV system gross margin getting closer and closer to the immersion business. There, I think the introduction of the 3800 is going to be a big one. So the introduction of the E model, the 3800E, by the end of this year even though, as we mentioned, it's only less than a handful this year. So the impact on the year will be small. But once the 3800E is going to be the lion's share of the EUV tool, then you will really see a good boost to the gross margin that will have on EUV.

And just one quick follow-up on China, Peter. I mean I know you don't really want to talk about it, given everything is still up in the air. But when we look at what your competitors have done in other markets, there could be a future impact to ASML from whatever is negotiated in China. How should we look at it into 2023 year such really? I mean clearly, without quantifying it, will there be an impact? Or do you think that you have enough in your backlog that you could recognize through the year, which would mean that there should be no impact to ASML at all?

It's a good question, Sandeep, but it really depends on what the outcome of the export control negotiations are because we can speculate all kinds of scenarios, what that would be and what the impact would be. The only thing is that we have a safe buffer, you could say, between the demand and our output capability. So it fully depends on what the outcome is. And it's not that I want to talk about that. I don't want to talk about China. I want to talk about China, it's no problem. What I cannot do is just tell you what the impact of the potential outcome is because I don't know the outcome. That's the only thing that -- so I just don't want to speculate. That's the only thing. But I think I find comfort in the significant buffer, you could say, between the demand curve and our output capability.

All right. We have time for one last question. If you were unable to get through on this call and still have questions, please feel free to contact the ASML Investor Relations department with your question. Now operator, may we have the last caller, please?

That comes from the line of Aleksander Peterc at Societe Generale.

So just a final question would really be on your indication on EUV ASPs in 2023. Now you obviously highlighted it's slightly down versus '22. Is that a conservative assessment? And if you could maybe think about the elements that pushed the EUV ASP to such high levels in '22, those will not be achievable this year. And then I have a very quick follow-up with the E system, obviously. Should we also assume that if we look at the increased current for that machine, that you will basically keep half of the increase for yourself and the other half will be given to your customers as per the usual?

Yes. Thanks, Aleksander. So as I mentioned, gradually, over these calls, we've increased the ASP of the D model that we've given to you. So it started with €160 million and we said use €165 million, and I just said, use somewhere between €165 million and €170 million. You're right that if you just calculate the ASP in the quarters in 2022, sometimes you've got to some higher numbers. But the reason that, that was the case, as we've also explained on previous calls, those were one-offs as a result of revenue deferrals that were in there. So sometimes, that led to a remeasurement of those deferrals. And that then sometimes, because we're also looking at distributing those adjustments over a very limited number of tools, that could actually have quite a significant impact. But we've always said normalize that, don't take for granted that those numbers are the right numbers to use on a go-forward basis. So it's not that anything bad happened. As a matter of fact, something good happened because systemically, I think the ASP on the D has gone up. As it relates to the pricing of the E, yes, I mean, definitely, we're still sharing value with the customers. I mean that's a fundamental principle in the way we do business with our customers. We want to make sure that the incremental value that the tool provides over the predecessor tool is being fairly shared between the customer and ourselves. So that promise has not changed. If you look at the way that pans out because you need to recognize value is a whole bunch of things, value is better imaging quality, value is better overlay, value is higher productivity. If you add it all up, then in the past, you saw a pretty strong correlation between improvement in productivity and ASP increase. And that historically, on EUV, there was sort of a near 100% correlation there. As you know, ultimately, on the 3800E, we'll see an improvement in productivity, 160 wafers per hour to 220. That will be in stages, but ultimately the 220 is the stage that we'll get to. So that definitely means that we're going to have a pretty significant improvement of the ASP. And I think that, that also should play out in your models.

And in addition to that, it doesn't mean that this goes without any cost increase. I mean it's not that go down straight down to the bottom line. Why? Because the 3800, as we mentioned in many occasions, actually sees the first introduction of technology that we are applying in the high-NA tool. So there is also a sort of new stages, so it's also a cost element there. So it's not a one-to-one increase in ASP going to the gross margin. But yes, you'll see gross margin increase, but also a cost increase because we are using the latest and greatest technology out of high-NA backported into the 3800. So just...

That's correct. Yes.

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

Thank you. This concludes the ASML 2022 Fourth Quarter and Full Year Financial Results Conference Call. Thank you for participating. You may now disconnect.