ASML Holding's Management Presents at Goldman Sachs Technology & Internet Conference (Transcript)

| About: ASML Holding (ASML)

ASML Holding N.V. (NASDAQ:ASML)

Goldman Sachs Technology & Internet Conference Call

February 25, 2013 3:45 pm ET

Executives

Craig DeYoung – Vice President-Investor Relations

Unidentified Analyst

So we got Craig from ASML.

Craig DeYoung

Hi.

Unidentified Analyst

(Inaudible) So where shall see start, like what happened at the Q4, I mean your stock went down 3%, closing at 7% like, everyone got confused. What happened?

Craig DeYoung

I think there were several nice coincidence of how could we go clean around our results actually, and help this out like some serious guidance.

Unidentified Analyst

Yeah. So in other way, actually….

Craig DeYoung

Yeah. Mainly, that’s specifically I think was helpful to us, but again is what we’ve been telling the markets regards the intensity of the 20 nanometer nodes, it’s quite significant from a litho perspective, the move from the 28 to the 20. So the fact that you guys like TSMC and their competitors in the foundry space are going to, if and when they move there and again, it looks like that TSMC’s going to take a leadership position at least in the installation of capacity, it’s going to be quite significant and actually it will drive to a large part our second half year-over-year growth this year.

So just to recap the Q4 results and then maybe we can build that just to remind people. we were able to guide fairly confidently a year-over-year a flat year. So last year, we did about €4.7 billion coming into this year, Eric Meurice and Peter Wennink, our CFO and CEO suggested our view of the world today is roughly equivalent to what it was last year in terms of litho demand.

Now the composition will be slightly different. I have to remind you that in the flattish number, we’re considering about $450 million to $500 million EUV revenue. We will over the course of this year and potentially early next year recognize greater revenue than that out of the 11 orders that we have. but being a bit conservative on the revenue recognition timing in the $4.7 billion, let’s say to make the math easy, it’s about €500 million in EUV revenue.

So that makes the year-over-year base business down roughly, whatever 15% as we view it today. And that considers a very, very nominal spend in the memory sectors, both NAND and DRAM. So we just do not expect a lot of equipment acquisition there in support of supply, bit supply levels to about 25% bit supply in DRAM and about not above 40% in NAND. So the installed capacity through this year will support in the two different sectors those levels of bit supply. And I don’t think today there’s a forecast that’s above 5% bit demand in the memory in the DRAM space, so again we don’t expect much upside there.

And there could be as we go through this year a growing demand above the 35% to 40% in the NAND space that is supported by the current installed capacity. So toward the end of the year, we’ve suggested there could be some upside in the NAND, we’ll have to wait and see we know nothing. I mean our customers certainly are talking to us about it. Although if you look at the customers on an individual basis, both in DRAM and NAND, at least three out of the four, our data suggest are going to be very tight on the supply side to those levels of bit supply. So anything incremental above that could well drive equipment purchases.

But again in the first half we don’t expect that potentially in NAND and in the second half, we’ll wait and see. And it will probably be largely dependent upon on what their view of 2014 looks like. So if they see certain growth expectations for ’14, as we get into the second half of this year, they’ll probably start to gain some visibility, and it’s above these nominal levels I spoke of, and let’s say the 40% level, and then it probably will trigger additional capacity additions again, because they’ll want to get equipment in the place such that when the calendar year turns over, they are prepared to supply whatever the forecasted demand is, but in our 4.7, it’s very, very nominal assumption of spend just in support of the existing technology no transitions that are going on. So we are talking about a couple of handfuls of immersion tools, let’s say to each of those spaces at the most this year, so it’s not a very significant spend.

As we started out by saying, those guys that are going in the logic space, specifically the foundry players that are going to the 20-nanometer node with some aggression in the second half are going to their spend will increase half over half, and that will drive our second half along with the €500 million in EUV revenue to be significantly higher than the first half, against that €4.7 billion kind of target.

Who else do we have? So if we step back a bit and, maybe we talk about logic in more detail. Maybe three, four months ago if you are spending Samsung spending CapEx, Intel no CEO not spending CapEx, TSMC reporting CapEx slab, and now we are thinking, actually Intel is going to spend the same, Samsung maybe going to spend maybe not exactly as much, but still a big number, and TSMC could spend more, the stock wit the number which could end up at €10 billion on everything. So it’s quite healthy on the logic side.

Unidentified Analyst

Is it something, which could have an impact on all the (inaudible) basically your Q1 (inaudible) and then it’s improving as we go year-to-year.

Craig DeYoung

Certainly the orders will improve in time to support the stronger second half and shipment focus standpoint, so that's kind of a given.

Again you might remember, we’ve said on multiple locations, we’re not so focused on the orders. We understand that you guys are, but that's why our guidance, our revenue guidance is intended to give you a feel for what orders have to do, but because we are talking about three big customers that are really driving our business. You can imagine that they are very lumpy and we are talking about leading-edge tools that are €40 million each, so you have to be little bit careful as do we and how we manage the order flow such that it's not doesn't mislead everybody.

Unidentified Analyst

So 20-nanometer versus 28 nanometer what's the difference, because of course for you it’s great, because if I do the calculations 28-nanometer if I build big production line cost the €500 million of your tools, if I do the 20, it cost me nearly €1 billion, so for you it's correct?

Craig DeYoung

Exactly, yeah.

Unidentified Analyst

What is it on the customer on the customer's customer side, because I heard what Qualcomm was saying this morning, and (inaudible) cost is not going down, Broadcom, of course it took the own book, because it won’t like 20-nanometer on the sale of (inaudible) someone’s goods, so of course on a technology side, it's important to go to the next node, because we need to continue to more. Then on the other side if transistor cost will go down, it’s not that great, for you it’s good…

Craig DeYoung

Right.

Unidentified Analyst

Maybe for the industries are not good...

Craig DeYoung

You are absolutely right, I mean [Francois], we are – let's call it for lack of a better term, I can’t think of a better one other than crossroads here relative to cost and continuation of shrink for cost, but there is another purpose of shrinking that shrink for more functionality per square centimeter of silicon, so it's apparent, I think that some of our customer’s customers are able to create enough value in the new device at the newest nodes that they can pass whatever cost is – that incremental cost there is, it belong to the consumer.

If you can’t do that then there is no absolutely no reason to make the strength. So either you’re going to have to absorb the cost yourself or within the supply chain or you’re going to have be creating enough value that you can charge for it. So, there are certain functionalities that you and I and the audience who will pay for. But it has to be value, value-added. But we’re at point…

Unidentified Analyst

That’s a change compared to what we’ve seen in about 10 years…

Craig DeYoung

Well, because the values come at a cheap price. So, it’s justifying the value by cost reduction is value. So that’s a easy way to do it and fortunately the industries – once they had it easy, but where we’ve been able to perpetuate the decrease in cost or the increase in value in time without much thought. So today, I mean this is the whole reason that EUV has to happen.

I mean the bad thing for our customers are they’re confronted today with the significant increase in cost to go from the in-logic, let’s say to go from the 28-nanometer node to the 20-nanometer node. The good news for us as you suggested in the short-term is they have to have it. They have to do it, at least in the near-term.

The advantage for us in the longer-term is that they need EUV, because EUV even at €70 million a piece instead of €40 million a piece reduces the cost per function and allows the continuation of Moore’s Law on double transistor density. So again, the lithography were kind of in a sweet spot, but the last thing we want to do or can afford to do is stress our customers. And so, I mean EUV is vitally important to our customers today and what we need to do to bring EUV to the marketplace in a economic fashion of course, because we’re talking – it’s all about economics, is to get the throughput to the systems.

Unidentified Analyst

So, that’s why you bought Cymer.

Craig DeYoung

That’s why we bought the or in the process of attempting to buy Cymer. Yeah.

Unidentified Analyst

So on this topic, actually if I move to the Cymer acquisition which is pending et cetera, today what are we waiting for exactly for the transactions to close? It’s complicated – it’s a very complicated process.

Craig DeYoung

Yes and no, I mean but what we have to go through is we have to go through regulatory approvals in multiple geographies; multiple countries and we have to go through shareholder approvals. Well, that’s done. So on February 5, the shareholders of Cymer okayed the deal. So we’ve been able to pass the regulatory approval for anti-trust related issues mainly. Sorry, so you have anti-trust in fair trade if you will in Germany, in Israel, in Taiwan, in Korea, in Japan and in the U.S. So we’ve left the hurdle in Germany and Israel. So we’re in the process elsewhere in terms of regulatory approvals. Probably, the next one we’ll be looking for and hopefully in the not too distant future, meaning a matter of weeks will be the Department of Justice U.S. approval for the deal, so from an anti-trust standpoint.

We have also – there is one other regulatory approval process that we’ve gone through, knock on wood, in a fairly streamline fashion and that’s the CFIUS committee, which is the committee on foreign investment in the US.

Unidentified Analyst

This one is done?

Craig DeYoung

This one – this is what’s done.

Unidentified Analyst

Okay.

Craig DeYoung

No provisions or no requirements, so it’s clean approval process. So we still expect the deal will close in the first half of this year, but we do have to get over the hurdles for lack of a better word in the processes in the U.S., Japan, Taiwan and Korea. Okay, so those are outstanding, but again when we guided the first half of the year closure, we anticipated process timing of all of these, so we still anticipate as we said months ago that the first half should close.

Unidentified Analyst

Okay. If I understand today, obviously the industry is waiting for EUV, you’re working hard on it, hopefully in time, things will get may be quicker as we discuss this whole (inaudible). So it looks like you are trying to get a bit more of the value chain from a later source point of view also from mythology point of view, I think this morning you showed me this new magic tool which you can put at the back of the lithography tool where you basically, you’re trying to get more of the value chain. I think that that’s very interesting that you need to get your margins up from where we’re today.

Craig DeYoung

Yeah, I would put it a little different. Exactly, I’ll put it a little different than you. We are not trying to grab different parts of the value chain; we are trying to bring more value to the customers. It’s the same time kind of in the reverse fashion, but I like the more customer sensitive presentation of it. So no it’s okay. No, it’s very important. The acquisition of Cymer is really driven by the need to manage better the development activity in the UV. Just recognize that Cymer is a relatively small company. The technical challenges are huge.

Unidentified Analyst

Huge, we’ve learned in time.

Craig DeYoung

They’ve learned in time and we’ve learned in time. Again trying to be very fair, I think they underestimated the technical challenges involved in EUV, and we overestimated their capabilities to be quite honest. And again, I don’t mean it in a negative way, but they’re a small company with a huge challenge. So they need not only – and this was agreed sometime ago between both management teams of Cymer, and that the way to maximize the value and get the fastest progress in this area was the merger of the two company.

Now they have as you know a very successful b-to-b business, which we intend to leave as it is. It’s very well managed, very good return on that business, just the b-to-b; the ArF, KrF that they supply to the industry to us, and into our competitors. So that’s the intend to just leave that as it is. And to learn what we can from it and hopefully optimize the operation a bit, but not much needs to be done there. We don’t believe. But we’re going to fully integrate the EUV development activity into our organization such that we can work more effectively and more efficiently against this very challenging 70 wafer for our goal we have set for ourselves in conjunction with our customers for roughly this time next year.

So there is a lot of work to do. Obviously in bringing in a high value supplier into our fold, which is not our business model, by the way we are an outsource company. So it’s not our nature to acquire suppliers. We’d rather not do it. But in this particular case with a very – what really end up being a very high value part for the system. We will appreciate a significant decrease in the cost of our systems, which will enhance the margin opportunity, and the long-term for the company and certainly for that particular product. So that’s a byproduct as well.

To touch on the other issue related to metrology. So a few years ago, we brought a company called Brion here in Santa Clara, California which does; it’s a software company that’s what’s considered today computational lithography, so they’re very good at predicting performance of optical systems. So I’ve been able to optimize with the mass to the source, I mean in the light source, given what, knowing a good deal about the lens that is in between the mask and the wafer plane. They can set the system up and designed the mask for the optimum output of this machine. So that in and of itself has great value and that’s how they built their own business. Our intention for them in the acquisition some five years ago was to figure out how to use that, not only upfront in the design of mask and the setup and the optimization of the system, but in managing the system, so managing it back to that optimum setup if you will.

So what are they important, so we had the scanner itself, and then we had the software and what we were missing in the equation than was a metrology tool, a measurement tool, that can measure the performance of the scanner, meaning what we call the critical to the mention of the resolution, the line which that we were …

Unidentified Analyst

Printing

Craig DeYoung

Printing, thank you, and other important parameter being the image placement of the alignment. So we needed to be able to measure that very quickly and if possible in line to give us what you would call real-time process control. Because today what’s done is you run a number of lots through the scanner or through any piece of equipment and then you take a selection of those wafers off-line, you take them to another corner of the fabric you measure them and then you go, oh no, it’s screwed up here. So they can do it three signs basically. We will try to reopen, but certainly on a lot by lot basis. So that has great value to the customers, especially as you’re going from – you’re reducing your feature sizes, because the smaller the features, the more critical the alignments become and more difficult it is to manage that. So to wrap that all up, so now we have the scanner. We have the computational lithography that can setup the optimum condition, but when we feed it, information about how the process is varied in the end product, meaning their resolution, critical mention and the overlay, we can feed that into the black box and it can tell us what to adjust back in the scanner, because we have a lot of knobs, we can turn in the scanner.

So this again has great value for the customers in terms of originally increasing the process window and then in the end, improving the process control within that window. So that’s all, it all talks to the yield issues and as we are getting this, Ill continue on to a product strategy of the company. So what this does is, it enhances the ability of the customer to employ the near-term multi-patterning solutions that they are forced to use to do the 20-nanometer node at high cost, but they are able to maximize the yield out of that, because those are very complicated processes. It’s not only that they are costly, but again when you are trying to bring multiple images together to create one image with multiple passes of the wafer and multiple radicals,

Unidentified Analyst

It’s complicated.

Craig DeYoung

It is complicated.

Unidentified Analyst

And it’s bit of the – was (inaudible).

Craig DeYoung

Exactly, so this holistic lithography package which includes both hardware and software is likely going to find a very – actually it is now, finding a very high attach rate to the scanner itself because of the probably almost absolute requirement for this kind of controller processes. So we’re taking advantage of the needs of the customers and our ability to bring the software, hardware and this new metrology in to help them control the processes and it has value to them.

And so again, we expect that if we can prove in the 20-nanometer processes that are going to be implemented and ramped over the course of the next 12 months, if we can prove this finally the value of this tool, then there will be a higher attach rate within a broader customer base and it will lead to the similar attach rate in EUV as we get into volumes. So that, in the 2015 and 2016 timeframe. So it’s a hugely valuable product package to the customer. But it’s also quite margin enhancing for us because there’s a lot of software in there, so.

Unidentified Analyst

I mean it’s actually big numbers of the company because I think this year, it’s going to be all in our application and services and these types of things, whatever you call them, it’s €1 billion of revenues.

Craig DeYoung

Yeah, wait the factors would probably confuse you a bit, not intentionally about how we bucket these things. So today, this is all bucketed in our service revenue reporting and that’s why last quarter we said, they expect that to go up because again, the attach rate is getting much higher and just for historic reasons, we put it in service. It’s not intuitive, of course, but and that’s the way we do it.

At a point in time, I think, we’re not decided on that, but there is no reason for us not to make that more transparent to you. So, yeah – but again, this is a part of our dual product strategy. So we recognize that we need to get EUV in place as soon as possible and as economically viable as possible, as soon as possible. But we recognize that that’s going to take some time and the interim period of time, our customers are going to have to continue to use immersion in more volumes for multi-pass patterning, and what we’re trying to do to try to help mitigate cost is one; an improve performance of these packages I discussed, but we’re also increasing the throughput of the tools. So today, the tools out of the factory are at 230 wafer per hour. Later this year, there will be 250 capable, and we’re also going to improve the overlay performance, which again, because very important this interlayer within a layer alignment capability, because you bring in multiple images together to make one is highly important.

So we continue a dual product strategy immersion, not at all abandoning and that is matter of fact again, improving the performance as fast as we can in these two areas, and then the EUV in parallel.

Unidentified Analyst

Any question in the room, even for the late comers? Yes, please.

Question-and-Answer Session

Unidentified Analyst

So how long do you think it’ll take EUV to the mass produced customers in that?

Craig DeYoung

Yeah. So we talked about 70 wafers per hour is our goal for this time next year. There’s nothing really what I call magic about that, and this is the challenge in answering your question probably, if you ask me the question for five years out is probably easier to answer than it is for two-years from now, exactly what it’ll look like. And that’s because every customer and every process and every layer within that process specifically in logic, although across all applications has a certain either technical and/or economic threshold.

So depending on what he is doing today, the EUV may cut in at different levels in time. So let me just use the 20-nanometer logic foundry process, average process. It’s driving this 2X consumption of a litho, €1 billion to build 45,000 wafer starts of litho in capacity.

In those processes, they do about 19 critical layers, and only eight of those 19 can be done with the single image, which were all used to one flash, one image, easy stuff, only eight of the 19. 10 of the 19 have to done with the double patterning meaning litho can’t do this, got a decomposed into two images, and now I’m introducing the etch step in between two, not the etch this going to create the device, but etch is going to create an image that could be create a device. So we’re introducing etch into it.

So when you talk to Applied or you talk to Lam, they’re very happy about this and well they should be because it’s going to be required, but it provides more difficulty. But so you have a certain economics for that layer. And then of the 19, 8 straight, 10 double and one triple in the average process.

So triple patterning three exposures, three mask and three etches. So let’s say at 70 wafer per hour, which we’re going to try to bring to the market next year, and maybe it will knock off that first layer that that triple patterning layer because we have economics work out. And as we go up to 80 and 90 and 100 wafer per hour, we’ll start to probably grab some of the 10 double patterning. But each process, I’m talking about an average price, but each guy’s process is different in what it’s cost and then how many double patterning. So that’s were it’s very difficult to say exactly how it will ramp in time.

Over the last six months, we, our CEO, CFO and IR team have been talking about a dozen tools this year hence that which we have a 11 actually orders for. About a dozen next year, which will be a transition year. So the early adopters, which will likely be memory just for a couple of layers not the complicated logic guys, because they are already into the 20-nanometer building capacity with the multi-pass patterning. But so you will see memory and then maybe a little bit of logic late next year as they start to look out into their next nodes to the sub 20-nanometer nodes.

And then we could see in the 2015, 2016 timeframe as many as 30 systems as the memory guys get into a DRAM guys and dig it into a pretty fall and then you get the real early adoption within the logic space. But ultimately in about at least five years or not more than five years time, we anticipate that market will approach about the need for about 60 of these tools per year. So that's kind of – and then whether that's in the 2016 or 2017 timeframe that's about the best picture we can paint for you right now.

But we’ll be working in time to be more specific about the potential implementation points with some real scenarios that will address certain EUV throughputs if we get to X throughput where you know how many EUV tools versus how many immersion tools, because there will be a balance between as I suggested in the scenario, it will be a balance between EUV and immersion. The more EUV the less immersion and vice versa, which by the way in the end provides this roughly the same opportunity for us, that's why we said and our CEO said on multiple occasions. In the near-term it doesn't matter so much for us, our opportunities roughly equivalent either way you go, but it's highly important to the customers, because of the cost involved.

So for the next two or three years, we’ll try to help you with see what it look like exactly, but the top line will probably be very similar in any scenario.

Unidentified Analyst

A very quick question because we have got only one minute left. Currencies, is it important for you because it’s a big publisher, the Japanese Yen is 20% weaker than it was one year ago. If Nikon tries to set an equipment 20% cheaper, does it make a difference to you?

Craig DeYoung

It can, but I wouldn’t say in this environment it does. It would only make a difference if they had a highly competitive tool, which meant, they would have to have performance in all areas of the tool. Remember a resolution overlying throughput. Without resolution nobody can buy the tool no matter how cheap it is without the appropriate alignment capability same thing and then throughput doesn’t matter if the tool doesn’t work. So throughput is usually used in cost justification. But in any environment the foreign exchange effects that are probably most important to us are the Euro-Yen relationship as you pointed out and it will impact us mostly and only really at the point of time of sale because they can be more or less competitive against us. But for prices have no issues, again if it doesn’t do the over layer of the alignment that’s required, you can give it to a customer and he can use it. So without parity in performance and the price becomes a non-issue. But just to remind you, so we sell in euro’s and they sell in yen, so there is no – on our business generally there is no impact from currency.

Unidentified Analyst

No impact from currency.

Craig DeYoung

No impact from currency. I mean we have dollar bills buying from Cymer. We have a facility in – that’s a supplier we have in Connecticut, but generally there is not a very large impact on. It’s not like we sell in dollars and build in euros, like some of our – maybe the semiconductor (inaudible) guys.

Unidentified Analyst

Okay thank you very much

Craig DeYoung

Thanks. I appreciate it.

Copyright policy: All transcripts on this site are the copyright of Seeking Alpha. However, we view them as an important resource for bloggers and journalists, and are excited to contribute to the democratization of financial information on the Internet. (Until now investors have had to pay thousands of dollars in subscription fees for transcripts.) So our reproduction policy is as follows: You may quote up to 400 words of any transcript on the condition that you attribute the transcript to Seeking Alpha and either link to the original transcript or to www.SeekingAlpha.com. All other use is prohibited.

THE INFORMATION CONTAINED HERE IS A TEXTUAL REPRESENTATION OF THE APPLICABLE COMPANY'S CONFERENCE CALL, CONFERENCE PRESENTATION OR OTHER AUDIO PRESENTATION, AND WHILE EFFORTS ARE MADE TO PROVIDE AN ACCURATE TRANSCRIPTION, THERE MAY BE MATERIAL ERRORS, OMISSIONS, OR INACCURACIES IN THE REPORTING OF THE SUBSTANCE OF THE AUDIO PRESENTATIONS. IN NO WAY DOES SEEKING ALPHA ASSUME ANY RESPONSIBILITY FOR ANY INVESTMENT OR OTHER DECISIONS MADE BASED UPON THE INFORMATION PROVIDED ON THIS WEB SITE OR IN ANY TRANSCRIPT. USERS ARE ADVISED TO REVIEW THE APPLICABLE COMPANY'S AUDIO PRESENTATION ITSELF AND THE APPLICABLE COMPANY'S SEC FILINGS BEFORE MAKING ANY INVESTMENT OR OTHER DECISIONS.

If you have any additional questions about our online transcripts, please contact us at: transcripts@seekingalpha.com. Thank you!