Seeking Alpha
We cover over 5K calls/quarter
Profile| Send Message|
( followers)  

ASML Holding N.V. (NASDAQ:ASML)

JPMorgan TMT Conference Call

May 15, 2012 08:40 am ET

Executives

Peter Wennink - EVP & CFO

Analysts

[Abrupt Start]

Peter Wennink

The third generation tablets are going to take between 1 and 2 gigabyte. DRAM, you see the smartphones going upto 2 gig meaning that more than 50% of all DRAM capacity leading edge capacity will go to the mobile space. That is something which is clearly driven by the end applications that are not on the market today but will be next year. So the general expectation is that bit growth next year in DRAM could be about 50%, up from the low 30s this year from a situation whereas nothing in the backlog means that we do believe that somewhere by the end of the year, beginning of next year somewhere in 2013 we will see an uptick in DRAM capacity.

On NAND, no memory, given the shipments we did in 2011 we anticipated through our simulation model that the first half of 2012 was going to see some level of ultra capacity which is actually happening. We are on top of that, we’re seeing an extension of that overcapacity throughout the rest of the year basically driven by the fact of that the three bids per cell introduction is more robust than originally anticipated. So it means that bit growth can be partially taken up by more bits per cell, that is more robust than we originally thought. That means that we could see that overcapacity extending into the remainder of this year. But also when we look at the new applications I just talked about it clearly in 2013, 70% bit growth seems to be very realistic and that means that we need to add also in 2013 additionally leading edge capacity.

Now, that means that memory I don’t expect it is a heck of a lot more this year which by the way would be good because we will have our hands full dealing with the logic sector means that the first half of 2013 still looks pretty promising for the logic sector and memory needs to kick in next year. So next year also will have 10 to 11 units revenue recognized for EUV. So all in all I think the future looks at least for the next 18 months pretty solid.

With that I think I’d like to hand over to you for questions.

Question-and-Answer Session

Unidentified Analyst

Yes, I’ll ask a few questions and then I'll open up the floor for questions from audience. Now you mention logic foundries basically are driving near terms CapEx spend. So do you think there is some level of overenthusiasm at the moment from the foundry CapEx side or you think as you mentioned you know there is dye sizes increasingly all as low so just for it.

Peter Wennink

I think what we said if you believe and I think it is not a crazy number that for the 28 nanometer node which is going to be a big node that the build up of capacity is going to end up at somewhere 280,000 to 300,000 wafer starts, then we are heading to that number but given what I just said I don’t think there is a major overcapacity situation looming.

Unidentified Analyst

And now from DM side you know probably over the last two years we've not had any capacity addition in DRAM space. So you mentioned you know next year probably we could see some capacity addition in DRAM because of higher DRAM content in mobile applications, but you do you think near term in this Elpida bankruptcy situation is also impacting the CapEx not only from DRAM Ram front but also from NAND companies who are waiting to see some resolution over there?

Peter Wennink

I think that's a good question I think if you think about DRAM situation and Japanese situation with Elpida depending on what the final outcome will be with Elpida and where the DRAM capacity you could also envisage that a part of the DRAM capacity will simply disappear because those tools or that particular production capacity will be used to start making other devices which we have seen at the bankruptcy situation with Qimonda, their 300 millimeter capacity in Virginia went to TI to make basically 300 millimeter analog product which basically took it out of the DRAM capacity space.

Now the fact that the Japanese DRAM capacity is in a situation where it is. It is a clear economical reason, I mean the technology there is I think not the most leading edge. Now you can argue where that capacity will then still be used by another DRAM maker who can make it economically viable. It is kind of a question mark there whether you can do that. So if you want to do that you at least need to invest in new leading edge tools which is good for us. So I think that whole situation is a bit blur.

We will just have to wait and see, but that there is a chance that some of that DRAM capacity might disappear and go to all the segments. I think that is a realistic one and at least a part of it. On top of that we have seen that some DRAM customers have actually changed and in any case in one major fab have changed from DRAM to NAND taking effectively DRAM capacity out. Now so this year 2012 uneconomical capacity is taken out. You've seen some transfer of DRAM capacity to NAND capacity with the fact that we believe that DRAM content in mobile applications will also grow, paint a picture of that for 2013 we must expect at least an uptick quite as reasonable uptick for our DRAM customers which means more tools.

Unidentified Analyst

Now on NAND side you said you know there is some low capacity, a lot of NAND companies have also said that, so when do you expect NAND orders to recover for next year CapEx, do you think it will be more end of the year story or maybe first half next year?

Peter Wennink

Yeah, we don't expect that will be significant this year. So it will be probably end of the year, beginning of next year. So nothing imminent.

Unidentified Analyst

So how do you think in terms of this foundry CapEx trend going in second half, do you see remaining as strong as in first half or do you think, you could give some motivation there in second half.

Peter Wennink

Well, I think we see a continuation of the strength that we are seeing today over the next couple of quarters in any case and we think that the build up will last like I said till the middle of next year. Also you have to take into account that at the leading edge we have seen more competition today than we have seen over the last 10 years and what does that mean, for us it is relevant because we have tracked, let's say the utilization rates of our machines over the last seven to eight years and made a distinction between first tier and second tier foundries over that same period. And in up and downturns the competition at second tier foundry has always been much more intense and we saw utilization rates 10 to 15 percentage points lower throughout the cycles and down.

Simply because they need extra capacity to attract new customers. I think that same situation you will see now at the leading edge. So I think there's going to be some level of structural capacity needed, a buffer capacity of leading edge foundries to make sure that the leading edge customers when they come in have enough buffer capacity to expand their business. I am pretty sure that some fabless customers out there that are requiring quite significant rampup capabilities at the leading edge and that will be true for every leading edge foundry and we have three. So that means that we will have a structural you could say a higher capital intensity at the leading edge in the foundry market because that's what's needed, that's the buffer that they need, capital intensity will go up, capital efficiency in that sense will go down.

Unidentified Analyst

Okay, maybe I will open up to the audience now. Can I repeat the question? The question was how is 20 nanometers rolling out in terms of speed at the moment?

Peter Wennink

Yeah we are in quite deep contact with our customers now for the last two years to make sure that we can support a 20 or 22 nanometer node. Everything that we are seeing now is that is accelerating at a pace which is let's say faster, higher than what we saw in previous nodes. That would also, that's also a competitive edge to that you know and you are sooner ready with the 20 nanometer node, you have a competitive edge. So but we are seeing any customers making inquires about building up 20 nanometer capacity in the course of next year. So that’s rather fast.

Unidentified Analyst

In terms of orders to the (inaudible).

Peter Wennink

That will be end of the beginning of next year. I mean the order patterns for those machines will be -- we need six months -- six to seven months.

Unidentified Analyst

(Question Inaudible)

Peter Wennink

So who you know said this, I missed the part of your sentence? Intel? Yeah I think every customer that we speak to I think we have actually asked our customer base and say what is the level of productivity that you need to actually get it working and actually it's pretty low, some say you know let's start at 30 and 40 wafers at the end of the year is okay, but we need to be at 70 wafers by the first quarter of 2014 and we want to have 125 wafers by the second quarter of 2015, but that’s a kind of minimum. And it actually means that they want EUV because especially in the foundry area, EUV has a lot of issues which currently phasing with double and triple patterning. So the most important thing is not so much to throughput to be honest. I mean on the EUV we’ve made steady progress on the wafer, let's say on the voltage output. So the output in terms of watts is sufficient to do that let's say 40 wafers per hour, that’s not the issue, the issue is how can we make the source more reliable.

So can you put it into production then the tool needs to work 75% of the time and then you can plan on it. And so it so more what we are focusing on today is where the EUV is making sure that we can get the reliability up not so much deep, we will get to power, I mean we have now shown that. So it is more about the cultural mechanisms to go (inaudible) the laser beam and the plasma, the (inaudible) those are issues that are impacting their liability and that’s what we are working on today and we will integrate those solutions in December of this year. So when you then look at the performance of EUV, they definitely see a combination of lets say 40 wafers per hour machine to start with, grow into 70 grow into 25 with let's say the double patterning solutions that we have today and actually this is what is all about with EUV.

It is not like we went from dry to emergent technology where it was complete node to node transition you know from 65 to 45 you had to go from dry to wet and it was all or nothing and that’s not the case with EUV. EUV will be an economic equation per layer. So when they start with most critical [expose], they cannot make them in any other manner to start with EUV and all the other critical edge with double patterning. When EUV productivity and the reliability goes up it will eat in on the double patterning layers and basically you see a layer to layer transition. So basically what they say we are going to use an EUV whenever it is available. We could use for one layer or two and we will grow those layers over time and basically eating in on double patterning layers when EUV becomes more economically viable for those layers and that’s what will happen. So it's quite a different transition that we had from let's say dry to immersion or from 248 to 193, those were you could say node by node or not in transition. This is much more gradual.

Unidentified Analyst

(Question Inaudible)

Peter Wennink

Yeah, real time performance.

Unidentified Analyst

(Question Inaudible)

Peter Wennink

Well, like I said this is adding to the simulation that we used could be that the over capacity that we thought we’re going to have somewhere in the middle of this year could be extended over three to four months simply because of its more bit capacity available. So its there; it’s not huge, but it does have an effect which basically brings the over capacity that we have calculated more towards the end of the year than to the middle of the year.

Unidentified Analyst

Now in terms of you know you said EUV, so do you think the progress has been good over there, because I think management seem more confident, you and Eric seem more confident in the first quarter conference call on EUV progress?

Peter Wennink

Yeah, and that's largely driven by the fact that we have seen let’s say for an extended period of time, the right wattage, the right power output, which is okay; like I said a couple of minutes ago, it is more about now it is about making it work for a prolonged period of time that you can go to customers and say this will work in your factory 75% to 80% of the time and that's what we need to go for which actually means that we need to focus now more on the control systems to feedback and the feed forward in the loop systems that we need to build into in order to – they are under development today and they will be integrated into the final source in the summer of this year.

So people were really worried about the wattage, about the power, because power drives the productivity; we’ll get there; we’re now focusing on the second layer of those issues and that probably showed to our body language and our confidence; because those are engineering issues that we can master and I know we can.

Unidentified Analyst

So 60 wafers per hour is the target for summer?

Peter Wennink

Our proof of 60 wafers per hour is by the end of the year; we could ship like customers say, even if it’s 40 wafers, our ship has the tools because we actually needed and we can upgrade in the field to 60, 70 90, [125] all those tools that we will ship, will be upgradable.

Unidentified Analyst

(Question Inaudible)

Peter Wennink

Hurting in prices or what are you…

Unidentified Analyst

(Question Inaudible)

Peter Wennink

Yeah, just to give you an indication the price of the EUV tool, the bare bulk price to 65 million, but on top of that you have option packages that can easily bring it to 70 million to 75 million. Now when you go to 125 million or over 125 million you see EUV prices really from an economical calculation of Moore’s Law to 80 million to 90 million.

So you see few units, but you see much higher prices. When we run our simulation you see that also our topline growing; our topline will grow because not every layer will be taken over by EUV. Immersion will be very important production technology till the end of the decade and even going into the next decade.

So it will be a mix of the ASP, the average selling price will therefore grow and basically if we run the simulations, I mean we clearly see a growth for the company over the next five years which is quite significant driven by the fact that the value is captured by higher priced EUV tools, lower in units, but very much higher in the average selling price.

Unidentified Analyst

Peter you mentioned integrated feedback a couple of questions ago, with the source integrated feedback you just mentioned, so should I infer that you are aligning with one source provider to do that and that maybe the first two or three years EUV will be just one source?

Peter Wennink

That’s obvious for this year; we have two suppliers; Ushio is the other supplier supplied tools for the first six R&D machines that we shipped last year. Their architecture doesn’t fit the 3300, so that was the 3100; it was the R&D tool, now we have the 3300. Their architecture doesn’t fit that tool. The Cymer architecture fitted both the 3100 and 3300, so for this year lets say till the middle of next year there is only source supply that we can use and that Cymer and after that the architecture work being done by Ushio now to make sure that they can come up with the 3300 source which could be supplied to us towards the end of 2013.

Unidentified Analyst

(Question Inaudible)

Peter Wennink

No, not really; I mean they are more in, I think in a research state. But the supplier that we are working most closely with is of course Cymer today, because they are the only viable source today; literally the only viable source.

Unidentified Analyst

(Question Inaudible)

Peter Wennink

As what you call background IP which basically you had when you started to work together; this is for both companies, we own what we have. And when you work together you own what you share together, so you basically own it together. So we have good IP arrangements with them on this point.

Unidentified Analyst

(Question Inaudible)

Peter Wennink

Well, the Ushio architecture is quite different, so its not that EUV would have more difficulties if you would say well, there's a kind of similar architecture. The Cymer EUV source works through a laser induced plasma, while the Ushio source is a discharge which comprises different architecture. So the chance of let's say IP evolution is not very high, because it’s really different.

Unidentified Analyst

And let’s say you know Peter, first helping DPP; would you think the same fab would mix and match those two if DPP ever work?

Peter Wennink

Yeah, I mean to say impolite, it’s a light bulb, so whether we use a Philips or an OSRAM, it doesn't matter in a sense that is the case. So as long as we can generate sufficient power through the EUV source, it really doesn't matter.

Unidentified Analyst

Peter one question is on long-term targets, I think last year I mean 7 billion is what’s talked about in terms of mid-term potential revenue for ASML. So do you still think that is possible and by what time do you think we should be looking at 2014 or is it more 2015 by which EUV ramps up to full?

Peter Wennink

Yeah, I think when you would look at increase of the topline definitely when we see a full blown adoption of EUV which is in 2015 timeframe and that 7 billion is a very realistic target. So when we run our simulations that is definitely a number that should be achievable, but probably more interested in what the most realistic number is which is going to be what we are working on so very likely in the second half of this year we will come out with some more details, because the reason is that we need to collaborate with our customers, how their roadmaps look like and how our technology, our (inaudible) technology is going to intersect with that roadmap and what does that mean for the adoption rate of EUV double patterning, also taking into account the different architectural choices that our customers make on their own designs. So that's why the complex work that we are doing today, but I would say that number that you had just thrown out I think that is a minimum; I think its…..

Unidentified Analyst

So realistically it could be more than 7 billion as well…

Peter Wennink

Yeah sure, I mean if we look at certain scenarios that we currently have looked at that would imply that it could be higher.

Unidentified Analyst

Okay and one question on potential buyback, you have probably 75% you had used around that level by the current buyback by the end of first quarter. You’re still generating a lot of cash, so what is your view on next potential buyback, because our view is you can possibly done as much as 1 billion more?

Peter Wennink

Yeah, I think you know that we are limited with [subsidy], with holding tax free ceiling that we have on buybacks it’s a 1.130 billion. We can add on top of that about 70 million to cover for ESOP plans, so its 1.2 billion we did 700 million last year, so we conceal the 500 million this year which we’ll do. On top of that we paid 200 million dividend, so as about 700 million as we return to our shareholders.

If we keep generating more cash we have the possibility of what we call a synthetic share buyback which is in fact a capital remittance; it’s a repayment of capital with a combined reverse stock split which we also did in 2007 which is a synthetic buyback structure and we have the capacity to do 1.2 billion, the only things is that is a kind of logistical, I wouldn’t call it as nightmare, it is a lot of work, because it’s capital reduction so you need a special meeting of shareholders in which you have to announce that it needs to be approved. You have a two month waiting period which is relative to protection than you have an execution of about a week.

So all-in-all, it can take easily three, four months. So you don’t do this for 50 million, so what you want to do, you want just save up like we did in 2007, we saved up a sizeable amount with 1.2 billion at that time, because there is a lot of logistics and work involved that you don’t want to do for small numbers, so you could probably see as saving up some money and then do a bigger one.

Unidentified Analyst

(Question Inaudible)

Peter Wennink

Well, I think you can have -- if you would say you have an EUV situation where we have full blown EUV, 125 wafers per hour or even higher then the economic advantage for our customers are such that definitely they are going to go for more doubled patterning layers to a single patterning EUV. Which means that the EUV lithography content of the CapEx spent will go up. Total CapEx spent might of course go down, because you don’t need the extra step that you need for double and triple patterning.

Having said that, once you can do double patterning you have mastered that technology and once you have EUV running reliably why don’t you do double patterning EUV; I mean that is what all the customers are also looking at.

So I think it will more or less cause an acceleration of the shrink than that you say and which also would benefit our peers that are currently supporting double patterning in the area. They will need to start supporting double patterning in EUV, but you will have an intermediate period in which EUV will be only single patterning and that means that the relative percentage of the EUV part, sorry of the litho part in the total CapEx spend will go up and that might be quite significant

Unidentified Analyst

(Question Inaudible)

Peter Wennink

I don’t see any reason why; when we look at the potential growth of the industry and then look at the percentage of the CapEx spent historically, why did we go down, so if the industry grows, I think the CapEx will also grow.

Unidentified Analyst

Maybe Peter you can summarize?

Peter Wennink

I can summarize, all-in-all I would say 2012 a healthy year. It looks currently better than we thought at the beginning of the year driven by foundry; which we think will be extended into 2013. By next year memory will be very likely come also back and combined with the EUV revenue recognition of at least 10 units say about 700 million Euros worth of sales; it just gives us lets say a good feeling about the next 18 months and so it’s all-in-all pretty solid.

Unidentified Analyst

Thanks Peter and thanks everybody for joining us. Thank you.

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!

Source: ASML Holding's Management Present at JPMorgan TMT Conference (Transcript)
This Transcript
All Transcripts