艾司摩爾 (ASML) 2020 Q4 法說會逐字稿

Thank you for standing by.

Welcome to the ASML 2020 Fourth Quarter and Full Year Financial Results Conference call on January 20, 2021.

(Operator Instructions)

I would now like to turn the conference 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 is ASML's CEO, Peter Wennink; and our CFO, Roger Dassen.

The subject of today's call is ASML's 2020 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 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 2020 results conference call.

And I do hope all of you and your families are healthy and safe.

But 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 2020 as well as provide our view on the coming quarters.

And Roger will start with a review of our fourth quarter and full year 2020 financial performance with 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 will?

Thank you, Peter.

Welcome, everyone.

I will first review the fourth quarter and full year financial accomplishments and then provide guidance on the first quarter of 2021.

Net sales came in above guidance at EUR 4.3 billion, primarily due to additional DPV system revenue and upgrade business opportunities.

We shipped 9 EUV systems and recognized EUR 1.1 billion revenue from 8 systems this quarter.

One system was shipped with a new configuration that needs to be qualified at customer site as the revenue will be recognized after site acceptance staffed in early 2021.

Net system sales of EUR 3.2 billion was again more weighted towards Logic at 72%, with the remaining 28% from Memory.

The strength in Logic drives both DPV and EUV revenue, and recovery in Memory business is mainly driven by DRAM.

Installed Base Management sales for the quarter came in at EUR 1.1 billion above guidance, showing continued strength in our service and upgrade business.

Gross margin for the quarter was 52% and was above guidance due to the additional DPV immersion and upgrade business.

On operating expenses, R&D expenses came in at EUR 556 million and SG&A expenses at EUR 152 million, which was slightly above our guidance.

Net income in Q4 was EUR 1.4 billion, representing 31.7% of net sales and resulting in an EPS of EUR 3.23.

Turning to the balance sheet.

We ended the fourth quarter with cash, cash equivalents and short-term investments at a level of EUR 7.4 billion, which is significantly higher due to customer down payments and early payments, which materialize in 2020.

Moving to the order book.

Q4 net system bookings came in at EUR 4.2 billion, including EUR 1.1 billion for EUV systems, net 6, and very strong DPV demand.

Order intake was largely driven by Logic with 78% of bookings.

And Memory, the remaining 22%.

For the full year, net sales grew 18% to EUR 14 billion.

EUV system sales in 2020 was EUR 4.5 billion, which is about a 60% increase from last year.

On EUV margins, we continue to drive profitability in both the systems as well as the service business.

We achieved our 40% system gross margin in 2020 and delivered a positive margin on EUV service.

We expect the upward trend for both systems and services gross margin to continue in future years.

The Installed Base Management sales was EUR 3.7 billion, which is a 30% increase compared to previous year.

In 2020, we had a total of EUR 11.3 billion, reflecting customers' strong demand for EUV and DPV technology.

DPV booking value was at a record EUR 7.3 billion, with demand from both advanced as well as mature market technologies.

Our R&D spending increased to EUR 2.2 billion in 2020.

While we continue to invest in DPV and applications, product innovation, the increase was primarily driven by the acceleration of our EUV road map, both low and High-NA.

Overall R&D investments as a percentage of 2020 sales was about 16%, SG&A was about 4% of sales.

Net income for the full year was EUR 3.6 billion, resulting in 25.4% of net sales and an EPS of EUR 8.49.

Improvements in working capital contributed to a free cash flow generation of EUR 3.6 billion as we continue to invest CapEx in support of our road map and planned capacity ramp.

Excess cash will be returned as per our policy.

With that, I would like to turn to our expectations for the first quarter of 2021.

We expect Q1 total net sales of between EUR 3.9 billion and EUR 4.1 billion, which is a very strong start of the year and a reflection of the current market demand.

We expect our Q1 Installed Base Management sales to be around EUR 950 million.

Gross margin for Q1 is expected to be between 50% and 51%.

The expected R&D expenses for Q1 are EUR 620 million, and SG&A is expected to come in at EUR 165 million, reflecting a continued investment in the future growth of the company.

The higher R&D is to support road map plans to drive further innovation of our EUV, DPV and apps products.

The SG&A increase is driven by higher IT, security costs and general organizational growth.

These quarterly run rates are a good indicator for the expected full year operating expenses.

Our estimated 2021 annualized effective tax rate is expected to be between 14% and 15%.

Regarding our capital return, ASML paid total dividends of EUR 1.1 billion in 2020, made up of the 2019 final dividend and 2020 interim dividend.

ASML intends to declare a total dividend with respect to 2020 of EUR 2.75 per ordinary share.

Recognizing the interim dividend of EUR 1.20 per ordinary share paid in November '20, this leads to a final dividend proposal to the general meeting of EUR 1.55 per ordinary share.

This is a 15% increase compared to the 2019 dividend.

The 2021 Annual General Meeting of Shareholders will take place on April 29, 2021, in Veldhoven.

Through December 31, 2020, ASML acquired 3.9 million shares under the 2020-2022 program for a total amount of EUR 1.2 billion.

Given our strong cash position and positive outlook, we expect to execute a significant share buyback in Q1 2021.

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

Thank you, Roger.

As Roger has highlighted, we had a very strong quarter, resulting in another solid year of growth in both sales and profitability, driven by strong Logic, recovering Memory demand and a significant step-up in our Installed Base revenue.

We were able to achieve an 18% top line growth and 37% growth in profitability despite some unique challenges we're having to continue to run our business through the pandemic.

I think this is all thanks to our employees and partners who have done a remarkable job executing in this challenging environment.

However, we continue to remain vigilant as this COVID-19-induced crisis is not behind us yet.

Following a strong 2020, we currently expect another year of good growth in revenue and profitability in 2021.

In Logic, we expect another very healthy year driven by a further broadening of the application space, fueled by the global digital transition.

Customers continue to see strong demand for advanced nodes, which includes the secular growth drivers, such as 5G, AI and HPC.

And in addition, and also driven by the digital transformation, we are seeing a strengthening demand for the more mature nodes across a wide variety of markets such as consumer, automotive and industrial.

While we are still very early on in the year, we think that with these demand drivers on full throttle for advanced as well as mature nodes, we expect Logic revenue to be up at least 10% from an already very high number of EUR 7.4 billion in 2020.

In Memory, customers have indicated that inventory levels continue to come down and expect a further tightening of supply throughout the year.

As is the case with Logic, the digital transformation is also fueling memory demand across the broadening application space.

Customers continue to see healthy demand in data centers with increasing Memory content in consumer electronics.

With customers indicating stronger bit growth this year for DRAM around 20% and for 3D NAND around 30% to 35%.

And taking into account lithography 2 utilization already at high levels, we expect a recovery in lithography demand for Memory to continue this -- through this year.

Therefore, we expect Memory revenue to be up around 20% this year from EUR 2.9 billion in 2020.

Although DRAM primarily uses DPV technology today, we do expect our EUV shipments to DRAM customers to increase in the coming years.

On our Installed Base revenue -- or Installed Base business, service revenue will continue to scale with the growing installed base.

We expect an increase in contribution from EUV service revenue as these systems run more and more wafers in volume production.

Customers will continue to utilize upgrades to increase capacity and improve imaging and overlay performance required of future nodes.

With this continued growth in both service and upgrade business this year, we expect Installed Base revenue to be up around 10% this year from EUR 3.7 billion in 2020.

All in all, we started the year with robust demand across the entire industry and across all geographical regions.

This should bode well for a double-digit upside from our 2020 revenue numbers.

We feel comfortable with the levels of potential growth expectation for business segment, but clearly see potential upside to these numbers, where we can disregard any further impact of export control regulations resulting from the current geopolitical situation.

I would now like to update you on our products and business starting with EUV.

EUV is making strong progress and continues to mature as we execute our road map and grow our Logic and DRAM business.

As Roger mentioned, we shipped 9 systems and recognized revenue of 8 systems in Q4, bringing the total to 31 systems in 2020, with a revenue of EUR 4.5 billion for the year.

This translates to about 60% growth in EUV systems revenue, reflecting the expanding use of this technology in high-volume manufacturing.

Based on customers' growing EUV demand in advanced nodes, we currently expect a growth of around 30% over last year, translating to around EUR 5.8 billion in EUV system revenue for 2021.

We continue to improve the EUV manufacturing cycle time to enable the capacity in our factory to meet the growing EUV demand.

We will continue to drive the EUV 0.33 NA product road map, which is aligned to our customers' note cadence.

Our goal is to deliver value to our customers via performance improvements in imaging, overlay and productivity.

As customers continue to shrink on future nodes, the performance improvements of our 0.33 NA systems road map will also enable cost-effective double patterning solutions before customers reach a point where it will require High-NA to reduce process complexity.

We are aligning with customers on the road map timing of High-NA, insertion and volume production, currently estimated to be in the 2025, 2026 time frame.

To meet this time line, we will start integration of the modules this year and plan to have first qualified system in 2022.

We plan initial installation of the first systems at customer site in 2023, and plan to provide a more detailed update on our High-NA program during our Investor Day this year.

In our DPV business, we're focused on meeting our customers' increasing demand for all of our DPV products by maximizing factory capacity, reducing installation cycle time and optimizing performance of our systems in the field.

As we mentioned earlier, the application space for Logic is expanding rapidly, which also has an effect on the demand for DPV products across our entire product offering.

The demand has actually been stronger than we anticipated some years ago, which means that we have increased our investments in R&D to provide our customers with ever more powerful and productive litho machines to help them deal with the increasing demand and lower cost per chip challenges.

Our DPV R&D plan, therefore, includes a significant program to bring DPV to our -- from our XT platform onto the NXT platform, thereby seriously boosting productivity and lithographic performance such as CD and overlay.

In our application business, we had a record year for YieldStar shipments and shipped the first YieldStar 385 to a customer in Q4.

The YieldStar 385 offers the latest overlay of focus metrology with enhanced throughput and accuracy to its customers' future node requirements.

We also shipped an additional 2 e-scan 1,000 multi-beam systems in Q4, bringing the total number of shipments in 2020 to 3. And with 9 beams and high-speed stage technology, these ASMLs provide up to 600% higher productivity in single-beam systems.

In summary, 2020 was another great year despite the challenges presented by the pandemic.

For 2021, taking into account that we're coming off a higher 2020 revenue base, we still expect a year of double-digit growth.

This is driven by strong demand in Logic and continued recovery in Memory with potential upside to these numbers where we can disregard any further impact of export control regulations.

The build-out of the digital infrastructure across multiple markets drives demand for both advanced as well as mature process nodes.

This is expected to fuel demand across our entire product portfolio.

Although there are, of course, still some near-term macro and geopolitical uncertainties, the long-term demand drivers only increase our confidence in our future growth outlook towards 2025.

And with that, we will 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) The first question is from Mr. Joe Quatrochi.

It's Wells Fargo.

I was hoping to get some color on your Memory business and the demand you're seeing there.

I think last quarter, you had talked about 30% growth for this year for Memory revenue, but you came in a little bit below that.

So I guess, did something change there?

And then how do we think about that follow through to that into 2021?

Thanks, Joe.

No, I don't think anything really materially changed there.

In fact, the fact that we kind of missed the 30% was, in essence, the reallocation of a few tools from Memory into Logic.

So that really had to deal with where do the tools go by the end of the quarter.

So really is a spread over Q4 versus Q1.

So not a systemic reason.

The momentum that we saw building up in Memory in the course of last year, we think, is continuing.

And that has to do with the things that Peter talked about.

We see the bid growth development, the 20% for DRAM, 35% for NAND.

And of course, that's driven by demand in data centers.

It's driven by what we see in terms of Memory being design and used in consumer electronics.

So we see those underpinnings continue.

We also see that the utilization of the lithography tools is at a very, very high level.

So the demand momentum that we saw already in the second half of 2020 we believe is sustained into 2021.

And that's the reason why we, again, forecast a 20% increase there for 2021.

That's helpful.

And then as a quick follow-up, in the prepared remarks, you talked about some supply chain limitations potentially on the EUV side that you're seeing.

I guess does that change your expectations for producing 45 to 50 units or your capacity to do that number of units this year?

I think our capacity, our capability internal in the Netherlands, in Veldhoven, to build 50 systems is there in terms of people and square meters.

Now we have to build those systems out of modules, which we don't produce.

It's in the supply chain.

And it's just a reflection of what happened last year in Q2 and Q3, where as -- clearly, our key foundry customer came back and said, listen, our key customer for N3 is now blacklisted, so we cannot ship, so we need to adjust our 2021 outlook for EUV systems, which was followed by another customer that said, well, we're going to delay the road map, which also means that this will be pushed back once again, which actually led to a situation where we actually reduced the number of planned systems 2021 for EUV, because customers said, these are the 2 reasons.

And they were 2 big customers.

So what we did, we went to the supply chain and said, sorry, we need those lenses and lasers.

Very expensive pieces of equipment.

We need them actually later.

You need to realize that the integral lead time between the installation of an EUV tool and the start-up of module production is 20 months.

So when you push back 6, 7, 8 months, yes, then -- and then finally, customers come to the realization that it's not as bad as they thought, and they want those machines.

Then we have an issue with getting the modules on time, and that's the only issue.

The only issue.

It's just a result.

It's a function of the fact that our customers changed their mind in Q2 and Q3 and then rechanged their mind back in Q4, and today.

But there's nothing we can do about it, which actually means that we're all prepared to do 50 units next year.

That is in 2022, it will just shift to 2022, yes?

So it's the -- so it's -- you could say it's supply chain limitations by design because it's -- because the customers told us, we don't need them.

And coming back there, oops, we might have been wrong.

Next question is from Mr. Alex Duval.

It's Alex from Goldman Sachs.

Congratulations on the very strong results.

A quick question on the CapEx spending backdrop and how that feeds into your guidance.

Obviously, TSMC is a very large CapEx spender.

They've guided over 30% CapEx growth this year.

Yet, your guidance, although very strong, it's closer to 12% on revenues.

Clearly, you also have other verticals, do you also have other customers in the mix.

But I wondered if you could give a little bit more color as to any areas where at the moment, you're being more prudent on your guidance, but where you could be a bit more positive over time and what you need to see to get more constructive?

Yes.

We gave a range, EUR 25 billion to EUR 28 billion.

Hey, great.

We plan our business based on what they ask us.

And as you know, TSMC has been asking us, in 2020, on several occasions, to ship very different numbers for 2021.

So this is -- what I'm saying is it's in constant flux.

So what they are asking us and telling us that they would like from that point of view, has a range, and we need to be able to respond to that.

So I don't think you can draw any direct conclusion from the TSMC CapEx numbers.

Directionally, yes, but not in absolute terms.

Now having said that, I also said that I do believe that we see upside to the -- your calculated 12%.

And an upside is clearly there, but I think that's upside that comes out of the Logic space in China.

And of course, that will happen if the current export control regulations stay as is.

Now you could argue also, why do you -- why are you so conservative?

Well, simply because what we've seen over the last 2 months in terms of regulations that we had to deal with and that basically we're issued rather suddenly, there is a level of conservatism at -- and on our side.

I said, we're not going to add that upside yet to your 12%.

Because we've been -- I wouldn't go per it, but I mean we've been surprised on a regular basis by all these new regulations that do have an impact on our business.

So where nothing changes, it stays as is, there is a significant upside to what we told you today, but then everything needs to stay as is.

And I think we made that clear in our prepared remarks.

Next question is from Mr. Mehdi Hosseini.

It's Mehdi Hosseini from Susquehanna International.

2 questions.

Peter, when you look in the longer term, looking at 3-nanometer transition, I understand opportunities in the near term.

But I want to hear more about what your thinking has been, migrate to the second generation of entry.

And I want to better understand how change of transistor architecture, like Samsung migrating to gate-all-around.

And also introduction of High-NA is going to impact your overall system shipment.

And I'm -- in context of what happened in -- during 2014 through 2016, when we migrated from planar transistor to FinFET.

And there was some slowdown, and I want to see if you see the same kind of pattern happening, and I have a follow-up.

Yes.

I'll follow up on 4 questions.

So on the transistor architecture, yes, I think what we know today on the patterning side and especially, I would say, on the geometric side, I don't think it has a major impact.

That's not what we expect.

So whether it's the FinFET or the gate-all-around, it is in the 3-nanometer realm.

And customers need that lithographic capability.

And it's an architectural choice, as you pointed out.

We don't think it has a major impact on our business.

The longer-term 3-nanometer transition.

I think you were probably referring to the transistor architecture change.

We'll just have to see how that pans out.

An architecture change, even when we had FinFET, you also know that wasn't completely flawless because it is new, yes?

And if there is a slowdown, it probably could be a slowdown because the technology is indeed new.

And that the ramp is potentially slower.

But that's speculation at this moment in time.

From a, let's say, lithographic point of view, there's not much difference.

How does that impact High-NA system shipments?

Well, High-NA is not slated for the N3 node.

It starts to be used at N2 and N2+.

I would say N2+ and N1 and beyond.

So in -- for the 3-nanometer node, it doesn't have a major impact.

I have a very short follow-up.

China as a mix of your overall revenue has been going up 12%, 19%, 17% 2020.

How do you see that trending in '21?

And I understand this could be a source of upside, but what is your current projection?

Yes.

I think what we see -- we see that trend keep going up.

It has to do with the fact that there is a significant amount of investment plan in China, Memory and Logic.

In our -- what we gave you in terms of growth for 2021, 10% Logic, 20% Memory, 10% Installed Base.

Our assumption there is that we -- that the indigenous Chinese business has about the same euro level, yes.

But it's a different type of customer, yes?

I mean as we told you also 3 months ago that we expect 2021 China business to grow, but largely in Memory.

So it's in 3D NAND and in DRAM.

Those are the big drivers.

And that's in what I would call, what we gave you to calculate and somebody calculated it to a 12% growth number.

On top of that, there is a significant upside in Logic.

How big can that be?

It's significant.

But like I said, we're a conservative company.

We've experienced the unpredictability of the legislation over the last couple of months.

We don't want to put that into your forecast right now, but when things don't change and they stay as is, there is some significant Logic upside in China.

Next question is from Mr. C.J. Muse.

It's C.J. with Evercore ISI.

I guess first question on gross margins.

You gave a pretty solid outlook for the March quarter.

And as we move into the second half of the year, you're going to start shipping the EUV tools, I would assume.

The installed base would start to come off warranty on the EUV side as well helping.

So how should we think about the trajectory of gross margins?

And is 52% plus or minus doable for the full year now?

Thank you, C.J. So you might recall in the last call, in the Q3 call that we had, we talked about a bandwidth for the year between 48% and 50%.

We also reminded people at that stage that in comparison to 2019, we also started the year with a 1% negative as far as that is concerned on High-NA.

So that was one of the reasons why before doing your bridge between 2019 and 2020, you first need to dug that 1% out there.

I think bearing that in mind, but also bearing in mind how sales seems to be firming up in this year, my expectation would be that we're going to see gross margin trend towards the upper limit of the bandwidth.

So the bandwidth of 48% to 50% that I gave in Q3, my current expectation based on the composition of sales would be that that's going to trend up towards the upper limit of that bandwidth.

Okay.

That's helpful.

And then I guess a quick follow-up on a prior question around EUV supply constraints.

It looks like implied planned revenue units is 40, your backlog is 42.

Curious if you think your supply chain can offer any upside to that 40?

And then if not, what does that tell you around EUV tool demand in '22?

And do you need to start thinking about having sufficient capacity above 50 units?

Yes.

Okay.

Good question.

I think, of course, we'll push the supply chain.

And -- but don't expect miracles there.

I mean if you get -- at the end of the year, you get 1 or 2 tools extra, okay, fine.

But it's not going to give you 5 or 10 tools extra.

It's simply not possible.

I mean -- so what actually means is that the demand that we cannot fulfill this year, we will fulfill next year.

And your point on the 50 capacity.

I think it's efficient.

The 50 capacity has to do with the fact that although customers are buying units, they're buying basically wafer capacity.

And don't forget, we have a higher productivity tool coming out of the 3600D in the second half of this year, which has a 15% higher productivity.

So with 2022, only being these, you already get a 15% higher wafer capacity out there.

So with the 50 that we feel comfortable with.

Also, from a supply chain point of view, we'll be able to manage that for next year, yes.

And the 15% higher productivity on the tools compared to the C, you actually see that we have quite some -- we have ample opportunity to help our customers build wafer capacity, yes?

So I think it's enough.

Next question is from Mr. Adithya Metuku.

Yes.

This is Bank of America.

So my first question is just on your IBM customer.

There's been a lot of discussion around whether the foundries are including something from them or whether they're not.

I just wonder if you could give us some color on what you've factored into your FY '21 outlook for -- from your IBM customers?

And secondly, I had a follow-up.

So if you could answer that, I'll come to the follow-up.

Yes.

I think we're not going to be specific on any customer as you can imagine because of the fact that we only have very few.

So the issue is that when we look at 2021, 2 things are impacting our shipment schedule to our leading-edge customers is, one is what you refer to is effectively -- has there been a transition from tools that we originally planned for customer A potentially to customer B and C. I think, yes, that has happened.

But on top of that, I think there is the increased demand for advanced nodes, yes?

So it's the combination of the 2 that actually caters for maximizing the shipments out of our shipment capability, and it's the redistribution that has happened.

So yes, I think we're, in that sense, sold out for this year, but perhaps with a one or two upsides, we are referring to the previous question.

But I think no impact on this foreseen in our 2021 numbers.

Understood.

And just as a follow-up, just on the metrology side, I just wondered if you could give us some color on how you're thinking about revenues in metrology in fiscal year '21?

Yes.

I think you're probably referring to the multi-beam tools, yes?

Yes.

I think those tools, 3 have been shipped to R&D centers.

They need to be qualified.

So that is going to be the key decision point.

So customers are putting them into their metrology architecture and basically needs to be qualified with -- there's no loyalty to itself, but it's, of course, the software that actually drives the tools.

So when that's done, we will recognize revenue.

That's how it works.

So it's this year for those 3 tools.

Understood.

I just meant if you could give us some color around the growth in that bid, given these multi-beam tools are now shipping.

So it's what we're -- what we expect for 2021 is that we will have positive evaluations.

And those positive evaluations will be followed by orders.

So orders for HVM shipments.

When that will happen, it's still a bit unclear because it depends on where we get the sign-offs, yes?

But we could be able to -- we expect sign off in the first half of 2021.

So this year, then we could see orders for shipment towards the end of 2022, yes?

But I think -- so end of 2021.

But I think that we would see an acceleration of that in the year 2022.

So I think 2021 will be characterized by the qualification.

And the decision of the customers to put multi-beam tools in their metrology strategy, which then will probably lead to first shipments towards the end of this year and then accelerating in 2022.

Next question is from Mr. Sandeep Deshpande.

JPMorgan.

I'd like to just go back to the question on the Memory market.

You had guided to 30% growth last year.

You did around 20%.

This year, you're guiding the 20% growth.

There is also the added -- there is going to be the shift to EUV-based DRAM at some point end of this year or into '22.

So how should we be looking at the overall Memory outlook for ASML here?

I mean are we going to see an even more accurate outlook because of what we saw last year in terms of what you reported?

Or is this -- that this just goes along and it's just something shifted, and that is why it has happened?

And I have 1 quick follow-up on the EUV associated with Memory.

Yes, I think Memory outlook, it actually is developing the way that we expected and the way that we told you.

You may remember that since the middle of last year, we've told you that we see utilization is going up.

And that there will be a point where if we are at the theoretical maximum utilization that our customers will want more capacity.

And that's actually happening.

So now on the -- and I think Roger explained that on the 20% to 30% growth.

The only reason why the 30% isn't 30% -- not 30%, but 20% is because some of those shipments that were earmarked Memory actually went to Logic.

Why?

Because Logic was on fire, and Memory was getting into fire in 2021.

So just the choice of a customer that, hey, don't ship me to A, but ship me to B because that's why I have more business.

So this was the only reason, which actually means that we are seeing -- Memory is coming in, especially DRAM.

I mean we need to distinguish between DRAM and 3D NAND.

We don't see as strong a recovery in 3D NAND.

You could argue because we're not that sensitive to it.

But we simply don't see it.

We see it in DRAM stronger.

So this is really is a DRAM game in 2021, whereby 20% bid growth which is the expectation of today, simply not -- simply too much to be dealt with in the current installed base.

That's why we see the orders coming in.

So yes, EUV will be used in DRAM, especially in 1 alpha.

But that is going to be limited, as we also mentioned last time.

This is not going to be a node on node, let's say, full transition from DPV to EUV.

Don't forget that EUV has a -- doesn't have a maturity level of DPV.

So there will be a part of the wafer capacity will be allocated to EUV with a limited number of layers, which will grow over time.

We'll see the first application of it end of the year and moving into 2022, yes?

So it's going to be a gradual adoption of EUV in DRAM.

I mean just following up on that DRAM and EUV.

How do you -- I mean you've had a very strong year last year in DPV and even this year looks very good in that technology.

I mean do you expect this -- even as EUV ramps up in Memory, et cetera, in DRAM, whenever it does, however slowly does or however fast, does this level of this baseline effect of DPV remain as part of your revenues?

And if we just add on EUV over time as we've seen over the last couple of years?

Yes.

I think the number of EUV layers will be relatively limited as compared to Logic.

So DPV will be the bulk of the layers and will stay the bulk of the layers, needing better lithographic performance and productivity.

So this is why we have these extensive R&D programs still in DPV.

I think DPV, in general, will be a bigger part of our business going forward than we anticipated a few years ago.

And it's not only Memory, it's like we said in the prepared remarks, it's very much also the mature markets whereby 90-nanometer, 65, 45, 28 are all growing in terms of wafer capacity, for the simple reason that there are applications or device for applications in that technology realm that are basically supporting in IoT solutions.

And that's a trend that we underestimated a couple of years ago.

I think this has been a big driver for our DPV business.

So it's not only Memory, where it will stay strong, it's very much also the broadening application space in DPV.

Next question is from Mr. Pierre Ferragu.

New Street Research.

Peter, a question from you, maybe like long side forward-looking, but also looking back maybe 2, 3 years ago.

So TSMC is going to spend between $25 billion and $28 billion in CapEx this year.

And they expect to grow revenues in the back of that in mid-teens, which means to me that that's a new normal for CapEx, like it's probably going to be a number that keeps growing from here.

And I can't imagine a word in which that many leading-edge Logic chip get into data centers, into pieces, into firms without volumes at DRAM following.

And so that feels very, very good in the long run for the industry.

And my question is first slide.

In 2025 -- so like 2, 3 years ago, you gave us a 2025 outlook with a fairly wide margin.

Did that kind of work in which TSMC spends so much in 2021?

Was that part of your range?

Or does that exceed what you were looking and then secondly, in -- so -- and then yes, of course, so my question behind that is your 2025 like kind of guide, do you think there is today a chance we go above the higher range of that range?

Well, you're basically asking me to give you a big review of our Capital Markets Day.

But let's take it from a 30,000-feet level.

3 years ago -- what you're basically asking is, Peter, how do you think about the industry today as compared to 3 years ago?

Yes.

I'm more positive for all kinds of reasons because, like I said in answer to your previous question, I did not expect DPV to be as strong as it is today.

And everything that we know, talking to our customers, will stay strong, yes?

I think that is -- I could call it a surprise, but that's something that we didn't understand well.

We understand that better today.

There's another thing that we assumed, as you know, when we talk about as a base case or a mid-market scenario, starting from the 16-nanometer node, we say basically every node that's 10% lower wafer capacity.

So it's minus 10%, minus 10%, minus 10%, so by the time that you're at 5, you've had almost 4 times minus 10% reduction of that wafer capacity needed for that node.

That seems very conservative at this moment in time if we listen to our customers.

Because as a customer like TSMC doesn't tell us that they're going to spend 25 billion to 28 billion, if they believe that their wafer capacity that they need for those nodes is going to be minus 10%, minus 10%, minus 10%.

So they obviously have now a different view as to the size of that market.

And I think we all understand drivers.

Perhaps not all.

But we have -- but most of them, yes?

So I think, all in all, I think there is a different basis for our assessment, I think, of where we can be in 2025, 2026 or 2027, for that matter, yes?

And I think it hasn't worsened.

I think it has gotten better.

Definitely on the Logic side, right?

It's definitely on the Logic.

But also to Pierre's comment, Memory is a derivative of Logic and Logic we need in all these applications, Memory will follow.

So yes, do we have a more positive basic view as to the growth perspective of the industry?

Yes, I think we have, yes?

Our next question is from Mr. Krish Sankar.

It's Krish from Cowen.

Congrats on the strong results.

First question I had for you, Peter.

I think you did answer this question in many ways.

So sorry for beating it up again.

The upside to calendar '21 numbers, is there a way you can simplify and say, do you think it comes from either Logic or Memory?

Do you think it comes from DUV or EUV?

And then I had a follow-up.

Yes, I think it comes from DUV and Logic.

That is a very short answer for Peter.

That is what it is.

And then a follow-up for Roger.

On the EUV service gross margins, 2 quarters ago in September, you turned positive.

Is it fair to assume from here onward those margins should start and keep improving because all the 2019 tools coming off warranty keeps adding to the service gross margin?

Yes, Krish, I think that's a fair assumption.

So we -- indeed, we did turn positive during Q3.

We were positive for the entire year.

And as I mentioned also on the video, I believe that within about a 4-year time frame, we should see the EUV service margin sort of approach the corporate gross margin.

And that's the trajectory that we're on.

And it's a matter of, on the one hand, to your point, seeing tools getting out of warranty, seeing tools produce more and more wafers.

So the throughput is going up.

And as a result of that, the number of wafers and therefore the paperware for -- going up for us.

On the other hand, us being better able to control the cost.

So that's the trajectory that we're on.

That's the goal in 4 years' time, and we'll continue to develop towards that goal.

Next question is from Mr. Aleksander Peterc.

So this is Alex from Societe Generale.

I just have 2. One is on -- if you could comment a little bit on EUV average selling prices, which are quite firm over the past couple of quarters.

So should we expect this firmness going forward as well in the first half?

I suppose you have seen an uptick in the second half with the new model shipping.

So if you could maybe comment on that a little bit.

And then just secondly, the R&D to sales is a bit higher than your longer-term targets.

So at what point in time do you think we should...

What...

R&D to sales...

Oh, R&D to sales?

Okay.

Yes.

Yes, R&D intensity, is that going to decline maybe once High-NA is shipping or even before?

Just to sort of know what time frame that could decline a little bit?

Yes.

That's fine.

So you are right.

I mean the ASP this quarter, but also in Q3 was a little higher than what you typically have in your models.

But it is, as we also explained on Q3, to a very -- it is really driven by configuration.

So what options are on the tool already when they leave the factory and specific customer requirements.

So that's what drives it.

And the composition that we have for the EUV sales, both in Q3 and Q4 were kind of the richer configurations as a result of which you saw the higher number in there.

On go-forward basis for the 3600D, I think we mentioned there that you should look at mid-teens in terms of increase over the ASP for -- in comparison to the C model.

In terms of your question on R&D, you are right.

I think R&D as a percentage of sales is comparable to where we were last year, actually, so it's high 15s, approaching 16.

It's a little lower actually than what we had in 2019.

As Peter and I already mentioned in the -- in earlier comments, there is so much opportunity that we see in the road map, and this is both low-Na, High-NA.

But also and Peter made reference to that, a number of developments that we're having in terms of DPV and the throughput, imaging and overlay potential that we still see there.

There's a lot of potential that we see.

And that's why we keep the -- keep it a little bit higher than the 13% than some of you might have in your mind.

That is still clearly the goal for 2025.

So we're working our way towards that.

But at this stage, we believe the company's value has increased by spending this amount.

So I think the approaching 16% that we saw for this year and maybe also for next year would still be appropriate.

But then from that point onwards, gradually model it towards the 13% for 2025.

I think that's the way I would do it.

And you also need to remember that we do these R&D investments, particularly for DPV, where we have underestimated the size of that market and also the need for DPV tools, bringing them on to the NXT platform, which is quite a significant R&D program.

But once we do that, and we see a significant increase in DPV demand, which we think will last.

Then we give our customers the capability to actually have a very productive tool with a lower cost per wafer competitiveness, yes, which, of course, drives a higher sales price.

So these are investments we -- and actually going back to an earlier question on the 2025 model, where we're doing things today because we see upside to that model and especially in DPV, for instance, and to give you 1 example, yes?

Next question is from Mr. David Mulholland.

It's David from UBS.

I just wanted to come back on the change that happened in terms of your plans and your build through the supply chain for this year or for 2021.

Because my understanding previously was you were willing to essentially bear the cost on your own balance sheet if demand didn't end up reaching the 45 to 50.

So I'd love to understand what changed because it seems like that's potentially not constraining the outlook for this year.

And then secondly, on cash flow, very strong quarter, obviously, in cash generation in Q4.

Was there any -- was that all just pure cash coming in the door from customers?

Or was there any seller receivables like we saw a few years ago?

Well, to answer the first question, David.

You might go back to the conference call script, but we never said that we were going to commit ourselves in the supply chain to 45 to 50 units, we -- actually, we said we were not going to do that.

We said, we are taking as a bit of a buffer above what our customers at that time, the end of Q3, said that they would need for 2021, which is a lower number than we're currently planning to ship, yes?

So we did take that buffer into consideration.

Yes, can we squeeze out another 1 or 2?

Probably.

But we never indicated or we never meant to indicate that we would cover our customers for our full capacity.

That's not what we said.

And this is not what we did.

So what we actually -- we're probably ending up somewhere in between, yes?

Which is good because if we wouldn't have done that, then there would not be a 30% increase in our DPV -- or sorry, in our EUV planned revenue for this year.

We are able to do that because we built this buffer capacity in the supply chain, but not everything.

David, on the free cash flow, as we mentioned on the previous call, this is the year where you see 2 effects in that regard.

So on the one hand, you see the effect of extended payments that some customers were entitled to, based on all the contracts.

And you see the effect under the newer contracts of down payments that are coming in.

And they balance out nicely.

Although in Q4, we saw most of the effect of the down payments.

And in the earlier quarters, you saw the effect of the standard payments.

But for the year, it's nicely equaled out.

So that's in fact what you see.

So very, very strong free cash flow in Q4, making up for some of the things in the previous quarters.

To your question, is there an anomaly in there?

Not really an anomaly, just as we had it in previous year, there was some customers choose a factoring solution for some of the payments and some of the down payments, but that -- on their request, but that's their discretion and their request doesn't impact us whatsoever.

But that hasn't really changed, I would say, as a policy from last year.

And I think the anomaly, if you like, that I just talked about, I think it's really something that we had in 2020.

For 2021, I would expect a more regular buildup of the free cash flow also over the quarters.

Next question is from Janardan Menon.

It's Janardan from Liberum.

I just wanted to go to your gross margin guidance where you said that you are likely to be hitting the high end of your 48% to 50% range.

But since you're doing a midpoint of 50.5% in Q1, and you will have the 3600D shipping in the second half of the year, which is a corporate average gross margin, which presumably is close to 50%.

I'm just wondering what is holding you back from getting to a north of 50% margin for the full year.

Is there something specifically in Q2, which could lower that?

And also, I would put it in the mix, presumably your DUV shipments are going to be quite strong this year, which should also have a positive effect on that.

Well, first off, we shouldn't forget that there is a very strong improvement over gross margin if you compare it year-on-year, right?

So let's not forget that.

As I mentioned to you, of the 48.6%, you should start by deducting 1%, you get to 47.6%.

That's really the starting point for the year.

So even if I say that we're approaching the 50%, I think that's a very strong improvement over last year.

If you then say what is going to be different in the next quarters?

To a very large extent, it's mix, right?

So in Q1, we have quite some immersion sales.

And as you know, immersion has a very strong gross margin.

We have more immersion sales in there, relatively speaking.

We have relatively low EUV sales in there.

So while you're right that the gross margin per tool in the second half on the 3600D is going to be a bit bigger, I think it's also fair to say that in the second half we'll see more EUV sales relatively speaking than you have in Q1.

So in that way, it kind of creates an equilibrium there.

I think it's also fair to say, when I say approaching the upper limit, approaching the 50%.

When Peter talks about potential, the potential over the 15.7, I think it is fair to say that the potential that we see there, as Peter mentioned, is in DPV and in Logic.

So that would lead to a further uplift of the gross margin.

So it's 50% that I was -- I think we're approaching is for the 15.7 indication that we gave or expectation that we articulated in terms of sales.

If we're going to see sales uptick, I think there should also be an uptick in the gross margin percentage.

Yes.

Understood.

Can I just -- got a brief follow-up on that.

Just on the potential upside for this year that you talked about.

You said it's all coming from, if it does come, it's DUV and from Logic.

Is that entirely the China sort of geopolitical upside that you're talking about?

Or is there potential upside outside of that?

And normally, when Memory prices are going up, in past cycles, we've seen that the memory vendor, the DRAM vendors tend to increase their orders as well.

If they were to come out with additional DUV orders, would you have the capacity to meet that demand?

Do you have sufficient buffer there, if that were to happen?

Yes.

I think we would be able to deal with some of the additional demand in Memory, and you are right.

I mean you've been around, Janardan, even longer than I am.

So I mean you know how that goes.

So yes, we would be able to do that.

How much, it would be difficult at this moment in time, we're January '20, we're talking about, and we're trying to guess the entire year.

So yes, I think we'll be able to do that.

The upside that we were referring to is DPV Logic.

That is where the very large part is upside in China.

Which like I said, if we can ship under the current rules, then that upside would materialize, yes?

And which would have the impact on the top line and on the gross margin, as Roger mentioned.

So yes, I think we will be able to do that.

So there's some upside on Memory if they would come.

And it would be DPV, to your point.

All right.

Okay.

We have run out of time.

So 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.

Before we sign off, I'd like to remind you that -- remind everyone that we are targeting to host our Investor Day on June 23 this year in London.

Now we hope we can have a face-to-face meeting.

But of course, this will depend on the progress against the virus.

We will provide more details in due time, and we hope you'll be able to join us.

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, sir.

This concludes the ASML 2020 Fourth Quarter and Full Year Financial Results Conference Call.

Thank you for participating.

You may now disconnect your lines.