ASML Holding N.V. (ASML) Earnings Call Transcript & Summary

Hello, everyone. This is Adi Metuku, European semiconductors analyst at Bank of America. I'm pleased to have Skip Miller, Head of Investor Relations, with us today. [Operator Instructions] So without further ado, let's begin. Welcome, Skip.

And so initially, let's focus on the near-term and COVID-19 questions. So starting off. You -- when you provided guidance, you said that revenues could be up around 50% on a sequential basis in the second quarter but refrained from providing formal guidance given lack of visibility. Can you talk a bit about how demand has trended since you provided guidance, maybe split by Memory and Logic? And are there any logistics or supply chain issues that have affected your ability to deliver tools recently?

Yes. All right, thank you, Adi. Yes. So I think, first off, from an end market, clearly, there are some positive signs in terms of the end market drivers around data center, work from home, remote learning, which are driving increased demand in notebooks and communication infrastructure. I think there's also -- maybe to the offset of that, there's some question on how consumer electronics demand, i.e., smartphones, will be impacted by COVID. Regarding our demand, we haven't seen any significant changes to demand for the year at this point. Obviously, we meet with customers on a frequent basis to understand if there's any changes in how -- what they see in terms of demand and that will become more and more critical, especially in this environment of -- with the increase in uncertainty around COVID. On Logic, in particular, our customers are saying they'll continue to innovate, and they'll drive their technology road maps, which translates to strong demand for advanced nodes as they ramp both 7- and 5-nanometer nodes. And obviously, this requires advanced litho, and they are asking us to make sure that we continue to ship. Memory, we were covering from a downturn as we started this year. And aligned to what our customers have been saying, we're seeing the utilization increases on litho systems. And if this continues, obviously, this would trigger additional demand for systems in the second half. And we did see an increase in memory in orders in support of this demand. Something to keep in mind with litho is it has the longest lead/qualification times in the fab, advanced litho even more so, which means customers that -- they need to look beyond the second half. So they need to think of second half in terms of system shipments will really mean that it's setting up for volume wafer output in 2021. On the supply side, so far, we've been able to navigate the -- call it, the choppy waters of COVID -- created by COVID fairly effectively. We haven't seen any further significant disruptions in the supply chain. Our factories, after having to go through some of the COVID protocol and challenges along with that, are returning back to more normal levels. And downstream, I would say, at our customer sites, we're seeing some of the areas where the regions that were closed, i.e., Wuhan, is -- are now resuming operations. And from a support perspective, we're able to ship into regions. But also we're -- and able to get people, we're bringing local support levels up. But also we're bringing more remote capabilities. And we mentioned, I think, on our call using some AR, VR, HoloLens type technology, which is something that not only provides effective support in these times but something that we expect we will use more going forward because it just makes good sense from an efficiency of support but also good business sense. So in summary, we're successfully operating in this, you can call it, more challenging supply environment induced by COVID, and we haven't seen any significant changes to demand for the year. As COVID obviously is not behind us, we still have to manage any potential supply challenges that could arise in the coming quarters and -- but as we move more into the second half of the year, it's more about looking into 2021 demand in terms of what shipments can mean in the second half of the year. And although the value of the technology and in particular semiconductors is quite apparent in this COVID world, we're not completely insulated from any potential economic impacts. And so we're -- there still remains some uncertainty in terms of how global GDP could impact end demand and obviously ripple down into the end markets and ultimately semiconductors. So we'll have to stay tuned a bit on that front as things unfold.

Understood. So you said memory has been -- memory customers will need to add capacity if trends continue at this level. So I mean, how much of that is predicated on what happens to smartphones in the second half? How do you think about what consumer/smartphone demand does in the second half and how that could affect your memory demand?

Well, I don't -- I think if you listen to the -- again, the end markets, there's the data centers [ hype ] and the work from home, remote learning as it relates to communication infrastructure. And all those items, obviously, have a positive. The question is that how much on the smartphone. I think it's too early for us to tell in terms of what could occur there on the smartphone demand and how, in fact, that will then, in turn, potentially negative offset the positives around data center and others. So I think we'll just have to see how those 2 play out as we get further into the year and then we get more visibility. Obviously, the second half -- as we went into this year, we said we saw a recovery in the second half of the year on memory. Utilization trends are supportive of that. And the orders, as I mentioned, are also supportive of that. But ultimately, there will be -- we will have to look at the end market demands and what could occur with smartphones and how that could -- and what that will, I should say, have a impact on demand and is that still a positive view on the recovery in the second half as we get deeper in the year.

Understood. So when you talk about orders, this, I take it, you're referring to the orders we saw in 1Q as opposed to more recently.

Correct.

Okay. Understood. And just to clarify, so if you managed to execute well in the remaining 3, 4 weeks of the quarter, then your revenues could be up around 50%, as you said at 1Q results. Is that still the case?

Correct. Yes. We -- I think we, again, made it clear. We didn't provide formal guidance in Q2. But we said, if we are able to navigate through some of these challenges that were brought about by COVID, that Q2 could be a nice growth quarter. We indicated it could be on the order of 50%. In addition, we could see gross margins improving as well. So again, as I said, knock on wood, we continue to navigate fairly effectively through these choppy waters. We still have another roughly month to go. But overall, again, I think things are looking favorable in that respect, and that provides some opportunity for growth in this quarter.

Understood. Now one of the concerns recently that's come up is around Chinese foundries stockpiling tools in the more recent past, ahead of the expected increase in trade and tech tensions. We -- I was listening into my U.S. colleagues recently chatting to AMAT and they don't think that is the case. I just wondered if you could give your view on what you're seeing in terms of shipments into Chinese customers and how that compares to the -- to what you'd expect given your knowledge of the technology ramps, et cetera.

Yes. So we still are shipping into China, and there's a number of customers that are ramping their technology or qualifying their technology as we go forward. And we again expect that to be a -- has opportunity for another growth into domestic China. And I would say we do get this question, I'll say, over the past quarters and years as you see the U.S.-China tensions increase or change. But in general, I think it's fair to say that we don't see a stockpiling or pull-in of systems in response to the U.S. as it can be tied to U.S. trade tensions. Customers are just executing to their technology and their ramp milestones and the equipment demand is reflected in these planned ramps.

Understood. And again, kind of following on from that question, one of the things we've heard recently is around Huawei stockpiling chips for their products. And again, this is one level removed from you. But I just wondered if you could give us any color on what your customers are saying in terms of what they're seeing in terms of stockpiling at Chinese customers of the final end products as opposed to the tools you guys make.

Yes. So again, I can't comment because this would be a -- this is a customer of a customer. So it's hard for us to put any certainty around what's going on in that front. We don't have any visibility of something better for our customers. I think in general, again -- and I think the way we look at the region in general is end market demand is driven by migration as we move to this more digital economy. And with that, you see expanding applications and secular drivers such as 5G, AI, high-performance compute, industrial automation, so on and so forth. And all this fuels demand for both high-performance logic as well as high-performance memory. And although it is clear this is global, we can't speak for our customers' demand, but we see this coming at a more global level versus just one particular region.

Understood. And then moving on slightly -- taking a slightly different viewpoint. Obviously, COVID has impacted R&D plans or it's led to some cost cuts at most companies. I just wondered if you could comment on how COVID has affected your long-term R&D plans. And does cost control mean that R&D into high-NA will be affected? Any thoughts around that?

Yes. So I...

And I was just going to add maybe more generally also around your product road maps not specific to High-NA.

Yes. So I would say in general, although we are always looking for opportunities to control costs and, I would say, in this current COVID environment, we also have been mentioning that we're taking prudent steps to reserve cash. I would also say that we've continued to invest in our R&D as we have done in the past through these challenging times. This is because you shouldn't think that we're going to change our long-term R&D or High-NA plans due to COVID. Execution of our technology road map, as I've said, not only High-NA but EUV and deep UV as well, is not only our future but is also critical for our customer's road map. We've indicated a quarterly spend rate of around EUR 550 million per quarter. And I think Roger has mentioned a number of times that we would expect -- you should just expect inflationary increases going forward as we still work towards the longer-term R&D number that we outlined in our 2018 Investor Day, that around 2025, we'd be around 13% of sales for our R&D spend. That's kind of how more near term and longer term in this environment, especially.

Understood. Moving on to EUV in the near term. Now one of the things you've talked about is 45 to 50 tools in the 2021 time frame. Now I just wondered if you could give us an update on how the cycle times are trending. Has there been any recent progress on the cycle times? And where do you think they'll end up at the end of this year, end of next year? Any color around those?

Yes. So from a -- what we indicated on our cycle time is that we have started in this year around 30 weeks, and we plan to exit the year around 20 weeks in terms of manufacturing cabin cycle time. The significance on the 20 weeks is that sets us up for a capacity of 45 to 50 systems next year utilizing the same 4 walls that we produced, the 35 units in this year. I think, in terms of the progress so far in the year, we started the year, and in the quarter, we had the impact from some supply chain disruption, but also we ran with some -- when we went to the modular vessel, and I won't get into all the technical detail here, but as part of the modular vessel, one of the components with the tin refill, we had to make some tweaks in that respect. And we are in the process of completing that implementation. We largely have that resolved in such that, as we exit through this quarter and into Q3, we expect to get back on our cycle time reduction plan. And we still plan to ship 35 units in the year. And then as you look out longer term, again, as we're setting up for this capacity for next year, of this increase from 45 to 50, we see -- and why we're building that is because we see all of our logic customers ramping next year in addition to our first memory customer moving into EUV.

Understood. And just on the multi-beam e-beam tool, you recently made the announcement that you had shipped the first tool to a customer. Can you give any color on this? What tool is this? Is it the 3 by 3 tool? When will your 5 by 5 tools come out? And what is the road map there?

Yes. So just -- so a bit of comment on the e-beam, why e-beam, why multi-beam and plans going forward. First off, e-beam, the value e-beam provides to our customers is it provides increased resolution or sensitivity. So as you keep going down smaller and smaller nodes, it becomes more and more challenging to see these small defects as the lines and the features get smaller and smaller. E-beam provides that sensitivity. The challenge today with e-beam is the productivity and the speed. So with multi-beam, you're effectively adding more beams and increasing the speed, the productivity, which is -- enables and make it more, call it, production worthy. In this quarter, in Q2, we shipped our first multi-beam, which is a 3 by 3 multi-beam system. We plan to ship these systems or, call it, development systems that we'll ship this year. And we plan to further scale the number of beams in time such that as we go to the 5-by-5 beam system tool next year, that will provide additional productivity increase. The example we're just providing for the 3 by 3, the 9-beam system. In our press release, we indicated that the productivity increase will be around 600%. That's a fairly significant step-up in productivity, obviously, which will help make it more volume manufacturing ready. But 5 by 5 will even scale that further. So that will make it more prepared for volume manufacturing. And I think as you look towards the end of 2022, into 2023, that would set up for the 5-by-5 beam to start entering into volume manufacturing. And on our technology road map, we have plans to further scale the number of beams beyond the 5 by 5.

Understood. And just on these 5-by-5 e-beam tools now. I know some of your semi cap peers also produce these similar tools. So are you addressing a greenfield market when you expand the beams? Or are you gaining share from one of your peers?

Well, there's a number -- to our knowledge, I think we're the only ones that have the multi-beam system. There's also some work we're aware of on multi column, but we're the only ones that have a multi-beam system. And again, the competitive aspect of it is, obviously, multi beams equal higher productivity, which makes it more compelling for volume manufacturing in the inspection arena. Now there are multiple -- when you get into e-beam detail here, there are multiple segments that the e-beam market lands in. And so I think the fair way to say it is -- today is that the multi-beam provides the higher productivity, enables the capability to participate and realize opportunities in multiple of these market segments.

Understood. And then moving on to some more medium-term and question to Huawei and the trade and tech war. Obviously, you have the new rules from the DoC that have been announced. And I just wondered if you could comment a bit on how you think ASML will be affected by these tools. How much -- by these rules, sorry. And how much visibility do you really have into where the chips produced by your customers go and how onerous these new approvals will be versus the existing approvals you have to get for your products?

All right. Okay. So yes, first, let me just -- maybe a quick level set in terms of the facts regarding current process. We currently ship deep UV tools to China without license. They're not restricted. We've been doing shipping to China for 25-plus years. EUV does require license, and we've applied for export license renewal with the Dutch government. So that's currently pending and we need that to ship EUV. But DUV can ship and have shipped without license now for a number of years. The -- there -- the DoC communication, rule communication, at the end of April, that's planning to go into effect at the end of June. There are currently a number of questions on implementation of these rules. And we're working with industry groups, think SEMI and SIA, along with our peers, to get clarification or guidance on application of these rules, which includes due diligence as it pertains to military end use or end users. As we are still in the process of obtaining clarification, it is still, I think, too early at this point to comment on potential impact, if any. Now again, the plan is to have this really go in effect at the end of June. So in the coming few weeks, we'll be in the process of completing this process and understanding what implementation requirements need to be put in place, again, if any, with respect to our change in way of working. I think in the big picture this doesn't change the fundamental demand for expanding end market applications. As I mentioned earlier, you have AI, 5G, high-performance compute, along with the baseline of many others, that are driving technology today. Innovation will continue because the value is so significant and this semiconductor technology capacity will need to be built somewhere in the world. So as long as innovation and end markets demand continues, it'll likely be the key driver for semiconductor demand going forward and thus ASML products.

Understood. Maybe delving a bit deeper into that answer. One of the things that AMAT said in one of the sessions at our conference this morning was that they now are a bit more comfortable about how they'll be compliant than they were a couple of weeks ago when they reported their numbers and that they think the onus will not be on them but on their customers. Do you have any thoughts on this? Or is it still a bit fuzzy to really say anything along these lines for you?

Yes. I think I'll leave it this way, Adi. It's still a bit premature. We're still going through the process. So let us work through the process in these coming few weeks. And then we can say with more certainty in terms of what the implications are, if any, of this whole military end use clause and the definition to make sure that we're implementing in a way that we're compliant with the rules. And so I would say, give us the time we need to get through this before we make any definitive statements.

Understood. And then, obviously, Huawei recently has been [ talked ]. The U.S. government seems to be blocking Huawei from getting products even from TSMC. I'm sure you've seen the news. I just wondered if you could give us your thoughts on how you see Huawei getting blocked affecting demand for your products. There's been some news flow around TSMC potentially postponing 5-nanometer capacity additions, 3-nanometer risk production, which they then obviously said wasn't true. But I just wondered if you could give us your views around how you think this will affect you. And also, more broadly, as the trade and tech war escalates, do you see the possibility that all semicap vendors -- and this is more a question for the industry by space. But do you feel that there is a possibility of semicap vendors being blocked from shipping anything to the -- to domestic Chinese foundries?

Yes. I think as -- again, to your last question, I think, as it relates to this, they're different, maybe the DoC rules that we talked about earlier. It was more focused around this military end use and user. And this seems to be targeted at a specific customer. So I wouldn't draw a speculation between this particular customer or other customer and what it means to Chinese customers and Chinese foundries in that respect. I think, with respect to your earlier question, I think it's still too early to know. This is, again, as a -- not our direct -- it's a customer of a customer. So it's too early to know potential impact. And they are likely still working to understand what this means for them. So I again don't want to speak for our customers. I will say, just like we discussed on the DoC rules, it doesn't change the fundamental end market drivers or the secular end market drivers around 5G, AI or HPC. They still require advanced semiconductors and advanced litho that is supplied by ASML. And it has to be built somewhere in the world. And so does it mean you could see regional or company share shifts over time? Those are all possible outcomes, but I don't want to get into speculating on scenarios that could occur in this environment.

Understood. And then just kind of sticking with this topic but slightly differently. TSMC talked about building a fab in the U.S. recently. SMIC seems to be raising funding potentially for deploying capacity in China. And there's talk of the American industry lobbying for subsidies to bring onboard -- to reshore production. So how do you think of these news articles? Will this generate incremental demand for you? Or will this just be a [ resharpening ] of capacity? Or is the answer somewhere in the middle? Any thoughts around this?

Yes. Obviously, any new fab means additional demand, yes, for semiconductor equipment. And with TSMC being our customer, this obviously would mean additional business for ASML. So in that respect, you can say it's a positive for future demand even if it's still a few years out. As you build additional fabs, in general, I think you can make a comment that you do have some inefficiency that comes with that. And therefore, you could conclude that some portion of this, at least, would likely be incremental due to some of this resulting inefficiency with the new fab build even if it's still part of their planned overall capacity build. I think that's how we would summarize the U.S. fab announcement in terms of potential, call it, incremental even if it's still a part of their overall existing capacity plan build due to the inefficiency factor.

Understood. Understood. And again maybe -- I don't know how much you can say on this, but one of the questions, one of the things that investors are wondering is whether this reshoring trend is for real or whether this is some kind of token move. Do you have any thoughts around -- based on the discussions you have had with industry experts, with the government bodies in different countries, do you have any thoughts around whether this is -- whether this can truly drive some significant shifts geographically in terms of where production is based? Or is it again a bit too difficult to say from your perspective?

Yes. I don't want to state what we currently have today, and I don't want to speculate what the future may or could hold, but I would leave it like this. It's we don't manufacture today our deep UV, EUV tools in the U.S. We manufacture the components in the U.S., in Wilton and San Diego, but they -- the manufacturing of the systems, the integration and assembly, that's all go and manufactured in the Netherlands. And so how things could change in supply or manufacturing or whatever down the road, I don't want to get into any type of speculation because it would be only bad speculation today.

Understood. Understood. And okay. And one last question maybe. We're almost coming to the end of the session. So just on DUV and EUV, maybe could you talk a bit about what proportion of these tools are based on American technology? With EUV, you previously said, I believe, high -- mid- to high single-digit proportion of the machine comes from the U.S. Maybe if you could elaborate a bit on that, what the part is that comes from the U.S. and how you come up with these numbers. And then if you could again give us some color on DUV, that would be helpful.

Yes. As I mentioned, I think, again, our manufacturing is done in the Netherlands on these tools, our deep UV and EUV tools. However, we do get some components from both San Diego and Wilton in the U.S. With respect to percent of U.S.-origin content on the system itself, we're substantially below the 25% number, which is in the de minimis agreement as it pertains to the U.S. discussion here. And so from that perspective, yes, I'd say it's significantly below the content and as it relates to that U.S., but that, I'd say, is not the key topic of discussion as it relates to DoC rules.

Understood. Understood. And just on DUV, what proportion of your content comes from the U.S.? Can you say anything on that?

I [ would like to ] say it's substantially below 25%.

Okay. Understood. So we're nearly at the end of the session. We're nearly out of time. So let's end the discussion here. I'd like to thank Skip for making the time and -- for your insight, Skip, and thank the audience for attending our conference. So thanks all. And take care.

Thank you, Adi. Take care.