ASML Holding N.V.

ASML Holding N.V.

$680
7.12 (1.06%)
NASDAQ Global Select
USD, NL
Semiconductors

ASML Holding N.V. (ASML) Q1 2011 Earnings Call Transcript

Published at 2011-04-13 17:00:00
Operator
Welcome to the ASML 2011 first quarter results conference call on April 13, 2011. (Operator Instructions) I would now like to turn the conference over to Mr. Craig DeYoung.
Craig DeYoung
Good afternoon and good morning, ladies and gentlemen. This is Craig DeYoung, Vice President of Investor Relations at ASML and I'd like to welcome you to our investor call and webcast. As the operator mentioned, the subject of today's call is ASML's 2011 first quarter results. Co-hosting today's call from our headquarters here in Veldhoven, in Netherlands are Mr. Eric Meurice, ASML's CEO; and Mr. Peter Wennink, ASML's CFO. At this time, I'd like to draw your attention to the Safe Harbor statement contained in today's press release and in our first quarter results presentation, both of which you can find on our website at www.asml.com. This Safe Harbor statement will apply to this call and to all associated presentation materials. The length of the call will be 60 minutes. And now, I'd like to turn the call over to Mr. Eric Meurice for a brief introduction.
Eric Meurice
Good afternoon, good morning. Thank you for attending our first quarter 2011 results conference call. Before we begin and as usual Peter and I would like to provide an overview. As usual, Peter will start with a review of our Q1 financial performance with added comments on the short-term outlook, and I will complete the introduction with some further details on the current position as we look forward to the rest of 2011. So please, Peter.
Peter Wennink
Thank you, Eric, and welcome to everyone. As Eric mentioned, I would like to take a moment to have some observations on the events of last quarter and we'll share also some details on the Q1 results. The first quarter sales as you will have read came in at €1.45 billion, which was just above our guidance and sales continue to be driven by Foundry and IDM customers, as they continue to add further wafer capacity at the leading edge. The number of dry systems, which were mostly KrF, again represented well over 50% of the total number of systems shipped. NXT shipments increased to almost 70% of our total immersion system sold, and our service and field option sales came in close to €170 million. The net income for the first quarter was €395 million or 27.2% of sales. The net bookings for the first quarter came in at 40 systems valued at €845 million. These included advanced immersion systems for critical levels as well as additional KrF systems for less critical layers, and those were mainly ordered by Foundry customers for capacity additions. The total average selling price of the Q1 bookings was €21.1 million. Looking from where the strength came in the first quarter bookings, the demand from NAND manufacturers represented almost half of the booked value 46%, followed by DRAM 23%, and Foundry 26%. Our order backlog existing the first quarter was €3.3 billion and included 134 systems with an average selling price of close to €25 million. For the second quarter of 2011, we expect net sales to be about €1.5 billion, with a gross margin for the second quarter expected to be about 45%, R&D expenses at €150 million, which is an increase, giving a growing focus of second generation EUV systems deliveries and third generation EUV system development. SG&A is guided at €55 million. Our near-term outlook, firmly indicate support for another strong year. We expect bookings in Q2 between €900 million and €1 billion, justifying again our expectation of net sales in 2011 clearly above €5 billion. As part of the share buyback program announced last quarter, ASML has purchased 4.6 million shares for a total consideration of €142 million, which includes purchases up to and including March 27th. ASML intends to cancel the repurchased shares and at the end of the first quarter we maintained just close to €2.7 billion in cash. With that short summary, I would like to turn it back to Eric for more on our view of the coming year. Eric?
Eric Meurice
Thank you, Peter. So when we convened here three months ago, we were able to confirm that this semiconductor recovery cycle was strong and sustained. Our backlog had grown to €3.8 billion a very, very large number and the business drivers justified our expectation for revenue growths well about €5 billion in 2011. Today the business sentiment is challenged by increasing oil prices, global conflicts, sovereign debt issues, rising interest rates. The Japanese disaster also adds up to uncertainty, although direct impact to the electronic industry worldwide appears limited at this point. The uncertainty is causing some of our customers to review the existing equipment delivery and order plans. However, none of this requires us to adjust our revenue expectations for 2011, which is still clearly, clearly above €5 billion. These limited adjustments that we have seen are in fact only affecting litho system demand, which was potential upside. And in any event, which was above the analyst revenue consensus for us. As said last quarter, the reason for the demand for lithography system to be so robust is that it is both, one, critical layer growth-driven, and two, a strategic capacity addition driven. On one hand, the NAND and DRAM memory lithography capacity built-up plans have been measured so far. And the backlog and plan bookings appear not to threaten oversupply to analyst expectation for 2011, as manufacturers drive aggressively for node transition, but not so much for wafer start growth. On the other hand, the Logic and the Processor capacity built-up is still driven by a hugely growing number of critical lithography layers from zero up to 18 critical layers per chip and by an overall process complexity. It is also driven by the customer business need to put minimum as we call it critical mass of strategic capacity in place for the 40 nanometer and the 30 nanometer nodes and to offer this capacity level to their end-customers long time before they have to ramp. So beyond 2011, we are preparing the industry's next significant imaging technology EUV, in which we lead a vigorous development effort as you know, and we are making significant progress. As of today, we have delivered three EUV pilot production system, the NXE:3100, which are now demonstrating imaging an overlay results, that meet the device production specifications requirement. And additional three systems are still to be delivered within a few months and the qualifications of these systems are currently underway. Point two, the aggressive customer process development effort are accelerating on our other tools and have begun on the 3100 tools we just shipped. The general EUV infrastructure, including mask writing, mask inspection, mask cleaning, photoresist etcetera, is now effectively transitioning into a phase of optimization. Point number three, the energy source is still the gageing factor. We are making progress, as we have identified the issues and have a roadmap to reach the planned 100 watt corresponding to about 60 wafer per hour by the end of this year. And point number four, we are starting the industry (utilization) phase at this time, with the completion of our new factory planned for this year and with the production start of the new machine, the NXE:3300, our new next generation tool for which we have now ten commitments. Revenues for the NXE:3100 shipped this year, we'll be recognizing 2012 at the same time we begin to ship the 3300. And often of this tool will provide ASML another engine of growth for 2012 and beyond. So in summary, we are very confident that our leadership position in leading edge is again enabling us to reach our business objective for 2011, as we articulated at the beginning of the year. With that, Peter and I would be pleased to take your questions.
Craig DeYoung
Thanks, Eric. Ladies and gentlemen, the operator will instruct you momentarily on the protocol for the Q&A session. But beforehand I'd like to ask that you kindly limit yourself to one question with one short follow-up, if necessary, and this will allow us to give as many callers in as possible. So we appreciate your attention to that. Operator could we have your instructions now, and then the first question, please.
Operator
(Operator Instructions) The first question comes from Mr. Gunnar Plagge.
Gunnar Plagge
You talked about the macro dynamics. I was wondering whether you could give us a bit more color. And are there any patterns in terms of end-user dynamics, any customers that are and particularly effected? And then, secondly, if you could also comment on the competitive dynamics, where a couple of customers and of all of your competitors that have over the quarter made some noises that the 620 is not really meeting the specifications. And I think, you have also alluded to the fact that you might have to take a slightly larger role, which includes obviously also that you have to do some supply things around lenses. And if you can comment around that, it would be very helpful?
Eric Meurice
So on the macroeconomic level, clearly at this moment nobody has real facts. All of us, ourselves of course, our suppliers and our costumers have looked into the potential first level impact of Japan, and there is no real second level of impact, which is the next supply level. Again they don't see much et cetera, et cetera. So most of the reaction at this moment is feeling a reaction that in view of such dramatic situation, "has to have a cooling effect of some sort." It's more of a feeling a sentiment that it is anything quantitative. My analysis, and this is a personal opinion, is in fact it allows most of our customers to digest, to think. So we are more in a thinking mode than in a very obvious reaction to a known problem. This is why, what we've seen is in fact very limited and if it were to even be a bit bigger than what we've seen, we are still prepared, because we have enough buffer demand in front of us to cover for any even potential more negative news. So this is why we are still bullish with the same numbers as we had three months ago, which is we will be well above the €5 billion target that we issued then. So macroeconomic-wise it's just a sentiment, no specific. On the competitive side, indeed we think that our NXT is doing very well. That slowly, but truly the customers is now going from one version of the NXT to a second version with better performance in overlay and in throughput, which then potentially put our competition in more difficulties to catch up if and when the product would be qualified. So yes, at this moment, we are still managing the company to a large market share and we see this has sustained as we go.
Operator
The next question is from Mr. Francois Meunier.
Francois Meunier
A question about the share buyback actually, because you've done 140 million in Q1, I was wondering, because the book-to-bill is below 1 for Q1 and Q2 and the shares are below (inaudible) euros, does it make sense to potentially accelerate the share buyback at the moment in Q1 and Q2?
Peter Wennink
Clearly we have said, the share buyback program is probably in total €1 billion. Whereby, the way that we do this is that at the day of the conference call we give an assignment or an order to a bank that will help us to buyback a certain number of shares and we give them a euro number for this quarter. That's what we have done today and we're not going to disclose that in a number. But clearly that number will be a reflection of how we see the market and what we feel. So during this quarter, the bank is going to execute what we tell them today and that's what they did last quarter. And actually last quarter what you will then see is that throughout the quarter we are buying back at the average price of the shares during that particular quarter. But any additional disclosure about the size of what we have instructed the bank today we are not going to give, but clearly we are looking at the current situation wisely.
Operator
Then our next question is from Mr. Andrew Gardiner.
Andrew Gardiner
I was just interested in a bit more detail in your response to Gunnar's question. I was wondering whether the caution you're seeing across the customer base, is this being so disproportionately in certain areas or certain customer groups. But is it having an impact on any major projects, or is it just these slight changes in delivery times that is your discussion you're seeing. And also what you compensate if this is more of an issue of just retiming of a limited number deliveries rather than a big (problem)?
Eric Meurice
The impact is mainly in memory sector, which leads us then to be fairly positive for the future, because if memory doesn't buy now, they will buy some time. So whatever has happened, is happening now is in fact a push out into 2012, which is good. And what is stable is what we call the strategic built-up of capacity in the Logic arena. And this as we told long time before is difficult, I mean would not be impacted by small macroeconomic changes, because these are strategic decision by investors to have the right capacity of the right node. And these things have not changed.
Operator
The next question is from Mr. Timothy Arcuri, Citi.
Wenge Yang
This is Wenge Yang for Tim. Just to follow-up on the previous question. You mentioned some of the delivery schedule change is in the memory sector. Could you specify, is it mostly in the DRAM or NAND?
Eric Meurice
No, it's bit of average, maybe a bit more in DRAM.
Peter Wennink
Let's remember these are not big numbers, so don't try to make sub-level or detailed analysis on these things.
Wenge Yang
For a Foundry spending, your bookings last quarter on a Foundry is about more than €0.5 billion and this quarter it's down almost 80% to about €200 million. We all know that Foundry is planning to spend a lot of CapEx this year. What makes that big dropping Foundry orders? Is there any change in a foundry spending pattern for this year?
Eric Meurice
Well, I think yourself and Tim have been trying to all the time discover short-term trend in short-term numbers. Well this is not how it works with strategic products. The Foundries have decided as we can take examples like, Global Foundries when they decide to make a new fab in New York, do not have a quarterly type approach to taking decisions and putting orders to us. So you will have some quarter, which were larger, when there is a phase of development. And you will have quarters, which are lower, because they adjust, I will say, digesting the decisions one-by-one. So these are clearly not quarterly trend and you should not even compare one quarter to another in terms of Logic bookings.
Wenge Yang
Just one last question. The tax rate for Q1 is much lower than previous forecast and could you provide some guidance on the tax looking forward?
Peter Wennink
Yes, sure. We have an agreement with the Dutch tax inspector under the new legislation, which benefits innovative companies and companies that are highly R&D intensive, and that is for the next four years. So clearly, the level at which we can attribute income that is related to R&D spending to let it be the taxable income, in alliance we have to make a separation between income that can be derived from. You could take patterns and the other way income, that particular division that is up for a discussion annually. But that has a marginal difference on the overall tax rates. So the tax rate would be anywhere between 12% and 14% over the next couple of years, because of this agreement and this legislation that we expect to stay in place. So depending, it already is 12%, it will be lower when we are at the high ends of our quarterly sales range, it will be at the 14% level where we are at the more low-end of the sales level. So higher the sales level the lower the tax rate that's basically how it works. But if you stay within the 12% to 14% benefit for the next couple of years, that will be good.
Operator
The next question is from Medhi Hosseini.
Medhi Hosseini
I want to go back to the EUV. It seems like there is maybe a few quarter push out, but it sill doesn't change how your customers are looking at the technology and the benefits that EUV offers. Is that a rough way of thinking about, because there is a perception out there that maybe a double patterning is going to be substitutive before EUV. And I want to get your opinion on what's out there and what you see the trend behind EUV?
Eric Meurice
So as we said, before there are three segments NAND, Logic and DRAM. When you are in DRAM, you do not know how to do sub a 20 nanometer without EUV. So those guys, even if we delay day-by-day, knowing that they have no alternative technical alternative we would in fact delay the node if we continue delaying machine. So these ones would be, I would say, not having choices. In the NAND arena, they do have choices. They can develop complicated triple or quadruple patterning NAND up to 15 nanometer, after 15 nanometer they will be in the situation of DRAM where they have no choice. However they would gladly go to EUV, if EUV is cheaper. So NAND is for us to win. If you quote the price and you give the machine with a right throughput, which hit the right level of cost, then they go. If you don't, then they wait. So NAND is in that sense, what you call, is progressive. Logic is in between; it's between DRAM and NAND. On one hand, Logic would absolutely need EUV on some layers like DRAM does; but on the other hand, you can design in Logic less aggressive layers which makes your chip bigger and not as good in power consumption. But you can go around it, which is what you hear some players saying, well, if push comes to shove, they can go beyond 15 nanometer node. What they really mean by this is, they can compromise the performance lower if EUV is not available. So NAND is economics; Logic is performance; DRAM is, you've no choice. So at this moment we are delaying without delaying, and we are, as you see, delivering our machines as we plan to ship them. But we haven't yet proven the performance that the machine will run at. We were planning to prove that performance in Q4 timeframe, I think we made it clear. Now we still have more work to do to prove that performance. And as I said, proving the performance is important, because then you will determine which node you take. And then your customer will align to that node. So we are in this situation where customers are developing the nodes anyway, but there is a bit of uncertainty as to exactly how many layers we are going to catch and when.
Medhi Hosseini
What is the critical factor here? Is that for you and your power source provider to hit 60 wafers per hour, or does that have to be 100 wafers per hour?
Eric Meurice
The target of the production machine, the 3300 is about 120 wafers per hour. So if we are at 120 wafers per hour, we catch all the layers and everybody fight to have the machine. If we are between 60 and 100, we are in good business where we catch probably 80% of the layers. If you are under 60, we're starting to only catch the ones that have no choice.
Operator
The next question is from Mr. Sandeep Deshpande with JPMorgan.
Sandeep Deshpande
Firstly, regarding your visibility, I mean based on what has changed at the customers' at this point, would you say now that you have full visibility into 2011? I mean, the last quarter you had talked about third quarter visibility. So now, does a visibility extend to end of the year? And given that you have seen some pushouts the next year, do you have any visibility into 2012? And I have one follow up.
Eric Meurice
Indeed, we are getting more visibility in 2011, and therefore we confirm this number with more certainty than we had three months ago. But reading 2012 now with the current database that anybody has is impossible. The only way at this moment to read 2012 is through our, what we call simulation system where you have a theoretical set of models of how many machines you need. But the data does not allow you to read 2012. The only thing that we learn today is, well, it's not bad news that the amount of machines are slipping from 11 to 12. I think this is good news for us, who would prefer to manage a company with growth potential year-to-year rather than a very, very, high year and then a lower year the next year. That usually is a bit more difficult to manage.
Sandeep Deshpande
And on the EUV question, following on, on that, is the laser now the only gating item? And if you do get this to this nirvana of 120 wafers per hour by the end of next year potentially, then would 2013 be a year when the volumes of EUV tools shipped would suddenly jump up because that would be a production tool in that year?
Eric Meurice
This is a very good question. So EUV target is still the same; we should be able to guarantee or improve to 60 wafer per hour by the end of this year, and 120 wafer per hour by the end of 2012. If we do that, I would guess that a good standard ramp would be about one tool per month in 2012, 2013. And then 2014 would be a multiplier. This is not just normal time; it takes a bit of time to do R&D with such a new technology.
Sandeep Deshpande
Sorry, I didn't understand 2012. So if you are able to do 60 wafers per hour at the end of this year, you will have additional ramp of one tool per month in 2012 itself?
Eric Meurice
No I do a total market; so if we have the right performance, the 60 at the end of this year, 100 at the end of next year, we could maximize the introduction of the machines, and I would estimate this to be about one machine a month in 2012 and one machine a month in 2013, and 2014 being the big one.
Peter Wennink
But in 2012 our product is the same. This does not start in January 2012; it just starts somewhere in the middle of the year, that's my point.
Operator
The next question is from Mr. Janardan Menon from Liberum Capital.
Janardan Menon
Just a follow-up on the EUV, on the 3300 side. Since you expect to start shipping 3300s in the second half of next year, would you be including any of them in your backlog from the second half of this year, and if so, roughly how much of the ten commitments you have can we expect you'd include in the second half of this year? And a second question if I may is, on your Q2 order visibility, your backlog right now is quite heavily memory intense; quite a lot of both DRAM and NAND is in the backlog. And given your earlier comments about some of the moving of the time scale coming from the memory side, would it be fair to assume that your Q2 orders will be predominantly on the logic side? And to what extent would the Q2 orders be shipped in Q4 of this year? And what percentage of them do you expect to be into Q1 and Q2 of next year?
Eric Meurice
So on the EUV booking, I think Peter and I will probably say that we want to be conservative, and we probably will not recognize bookings this year. We will recognize bookings and the billings probably (vary) at the same time in 2012 as a way to have all the data with us when we do that. Because again, we will recognize revenue only when we are sure the machine has real performance. So we will wait until the last moment where we recognize bookings and billings. But the target is probably to ship. And this is why we have I would say, when I say "wonder month" in 2012, starting in June or so, 2012, we could, on paper ship 6 or 7. I think we have going to see our target of 10 in 2012. This is pretty hard to do, but we owe that to the customers; in fact all the customers wanted us to do so. So we are going to try to do the full 10 in 2012. Regarding the orders for Q2, as I said before, the bookings at this moment and the bookings in Q2 will probably not give any information on 2012. I mean, I would be highly surprised if there would be appetite for any of our customers to already call some capacity in Q1 2012. We don't see this; it would be not natural. So therefore, the whole Q2 bookings, and we have given a range of €900 to €1 billion would be for shipment in 2011.
Janardan Menon
And would it be fair to assume that it will be more logic than memory in Q2 bookings?
Peter Wennink
Yes, that's a correct assumption.
Operator
The next question is from Mr. Gareth Jenkins of UBS.
Gareth Jenkins
Just a quick one on new fab plans. I think, historically, last quarter you totaled out 13 potential new fab plans between '11 and '12. I just wondered if there's been any change in that, either up or down. And then I have a quick follow up as well.
Eric Meurice
The plans are the same. The timing and intensity of the numbers of machine going into it is the stuff that is being discussed because of sentiment.
Gareth Jenkins
And then, the follow up is somewhat different. Just on directed self-assembly, whether you feel it's a technology that is complementary, or something that poses a threat long term to (easy)?
Eric Meurice
No, I think there will be certain numbers of layers which could potentially be done with self-assembly. But at this moment, I think if ever it is successful, it would be one very specific layer, maximum two. So I would see this as complementary.
Operator
The next question is from Mr. Simon Schafer of Goldman Sachs.
Simon Schafer
I am sorry to dwell on this topic, but again is on EUV. Did I understand correctly? You said 10 EUV systems in total; would be a real challenge, or were you referring to the 3300?
Eric Meurice
The ten 3300 in 2012 is a possibility, but of course it is challenging. Before the ten 3300, we are going to ship six 3100. We have shipped already three.
Simon Schafer
Understood; got it.
Peter Wennink
And the ten systems is what our customers want; and that's why I had said it's our challenge to get them out. But it's going to be a bit of a fight.
Eric Meurice
So to be honest, on your plan for 2012, you may want to plan the best case of us shipping six NXE:3100 and ten NXE:3300. This is the maximum, best case we could do.
Simon Schafer
And then my second question actually was, just when you look at your own forecast and your base is still very comfortable with where I can sense the expectations are in the mid 5 to 5.5 billion or so, does that incorporate a specific thinking on wafer fab equipment spend in 2011 or overall CapEx budgets? Just trying to get a sense as to what your own assumptions are based on. I know you have your own very specific modeling as it relates to the memory industry and so on. But any color as to what your assumption is for overall wafer fab equipment? Would be very helpful.
Eric Meurice
So this assumption is based on the fact that memory DRAM would be a bit on the low side of what Gartner says. So in DRAM they would not have enough capacity to do 55%, bit gross; just be sure I said the right thing. So they would not be able to do the 57 from Gartner. At this moment our forecast, which is well above five, would not guarantee 57; it would be guaranteeing under 50. For Flash it would not guarantee the Gartner 83% NAND; it will guarantee nearer to 70, 75. But the logic will be higher than the poor Gartner unit growth expectation, which I think is 70%. So in the logic arena, we would expect to build more than 7, but as we said 7% unit growth equivalent. But as we said, this is natural because their new fabs strategically have to be ready before they can garner some new customers. So in other terms, it is very clear that TSMC, Samsung logic, Global Foundries and now UMC is in and SMIC is starting building up capacity to try to attack certain numbers of similar businesses. And as we said before, there will be duplication in this business. This is not to say that there is overheating, but there is duplication in this business because there is more competition.
Peter Wennink
I think Eric, Simon was probably referring to the total wafer Fab equipment for 2011. And when we have a view on that, which I think we don't. We stick to our simulation model, and we have said it more often on calls like this, the wafer fab equipment CapEx will be what it's going to be. It will go up when customers see that they have a business opportunity; it will go down when they don't see it. We actually don't follow it.
Simon Schafer
I understand. That's very clear. Thank you for sharing those assumptions.
Peter Wennink
I apologize. And if I feel that anytime somebody asks me about CapEx and the fab I only think about Litho. I didn't even know there was any other equipment.
Operator
The next question is from Mr. Jagadish Iyer of Arete Research.
Jagadish Iyer
First, Eric, you have always talked about a little capital intensity. You have given a slide which talks about increasing process steps for DRAM, NAND, Foundry and all of them. So I was just wondering, what is your early read on this going forward for '012 versus '011? I know it's kind of theoretical, but can you give us some kind of quantitative numbers in terms of how the Litho capital intensity is going to change? And I have a follow up.
Eric Meurice
Again, this is really more a team actually doing these things. CapEx intensity for us is a dangerous thing in particular that we can't really correlate to the past, because the past doesn't make any more sense. But the two or three numbers we kind of follow, because in the foundry environment the possibility of being a total CapEx intensity is 50% of the total revenue of foundry, is not a ridiculous number. And then on the 50%, we own, of this, 20% to 25%. So you could say that the term spend to 15% of the revenue of Foundry could be given to lithography. And if you block those numbers into what Global Foundries and TSMC and SMIC and UMC and Samsung Logic would tell you, they are expecting sales in 2012, 2013 and 2014. These are good numbers; in fact these are higher numbers than what we already ended. In the memory business, we usually in fact do not talk about capital intensity because it's so widely dependent on the ASP of the memory chip. Sometimes you have a chip that goes down in price by 70% in a year. So the capital intensity basically doubles, and therefore we can't follow that one. So we really follow mostly our own model per unit and per bit for more (inaudible). Sorry if I did not answer your question.
Peter Wennink
I think the secular trend, I think as also said already, the secular trend is that we believe that the Litho CapEx diffusion rate will go up. I think, historically, ten years ago was around 12%. We are currently now between 18% and 20%. And giving the trends that we are seeing, and also like to refer to Slide 17 and Slide 18 of Q1 that is on the website where we basically show the number of Litho (immersionally) as that is going up, which is just a function of the complexity that our customers are trying to deal with. If you add also the introduction of EUV in 2013 and beyond, it is clear that the long term secular litho diffusion rate will go up. And that will be clearly above 20%. And when it's going to be 24%, 25%, 26%, we'll just have to see. But it's very difficult to make this particular comment from one year to the other. We just look at the minimum long term trend, and that clearly shows that Litho CapEx, in intensity will go up.
Jagadish Iyer
I have a follow up. If I take your Q4 bookings and your Q1 bookings, you are probably running close to say €1.5 billion run-rate in bookings. Do you think your bookings can recover to such levels sometime in '011 or is it thought likely probably an '012 event at this time?
Eric Meurice
No, it will be natural to go back to €1.5 billion sometimes. Remember that if you do also calculation you see the current bookings are €3.3 billion. Is a very large number as to justify a run-rate of €5.5 billion sales. If you look at our historical backlog to billings, sometimes you could multiply by two. So €3.3 would be a run rate of €6.6 yearly billings. This is high; this is on the high side of the equation at this moment. So you could expect a natural reduction of that backlog to the normal level, and then a pick-up to $1.5 per quarter is pretty natural. So yes, definitively, we should expect this. Now what we don't know is, would it happen in a continuous fashion; would it happen a bit bigger on one quarter and not on the other; or, would the customer remain at the historically high backlog, because we are now so strategic for them and the products are so strategic that we maybe now in situations where we manage a higher backlog than we ever did before. So, sorry, at this moment we can't really gage this. And I don't think any analyst would not try to read anything again in this backlog trend; they do not really correspond to a business trend; they correspond to another way of doing business with us.
Operator
The next question is from Weston Twigg from Pacific Crest Securities.
Weston Twigg
Just a couple of questions real quick; first, you said that your cash balance contains €863 million for EUV and other programs. Can you give us an idea of what other programs may be included in that?
Peter Wennink
That's the regular NXE and NXT programs, so that's the regular business.
Weston Twigg
Okay, not new development programs. And then just to clarify, on EUV for revenue, it sounds like you are still hoping on the upside to maybe ship all ten of the 3300s in 2012. But it sounded like you said revenue recognition will be based on performance, so not necessarily any revenue on delivery. Do you think that that €863 million would get recognized on delivery?
Peter Wennink
Well, in effect the €860 million will only be recognized when we formally recognize the revenue on the shipments. So every repayment Euro that we receive today will be booked in income statement when we finally have revenue recognized under accounting principles in 2012. So that will stay in the accrued liabilities until we ship and we recognize the revenue. Now, on the 3100, that revenue recognition will take place on final acceptance. On the 3300, we have agreements with our customers that the minimum requirement that the 3300 has to fulfill are of such nature that we believe that we can have revenue recognition upon shipment because those are really minimum requirements that we have set as minimum for the 3100s, which we think we will definitely achieve when we ship the 3300. And even (it's) 125 wafers per hour, we have an agreement with the customers that they will take the full risks and rewards of ownership at that moment in time. So that's also the moment in time where we can book the revenue.
Weston Twigg
So just to clarify, for the 3300s you expect to be able to recognize revenue upon shipment in 2012?
Peter Wennink
Yes; and the 3100 is upon acceptance.
Operator
The next question is from Mr. (Richard Lee from Stanford).
Unidentified Analyst
My question revolves around what is happening here in Connecticut at the Wilton facility. Just was wondering if you can just give me some idea as to how the slow U.S. economy has impacted operations at the Wilton facility and what you see going into the rest of 2011 here in Veldhoven?
Peter Wennink
I think the Wilton facility is an integral part of the ASML worldwide operations. They make a significant part of the EUV systems and of our current systems. So you could say that what happens at the Wilton facility is more a reflection of what happens at ASML, and 2011 is the best year ever. So that should be very good news for our people in the Wilton facility and has not much relevance to the U.S. economy.
Operator
The next question is from Mr. Adrien Bommelaer.
Adrien Bommelaer
I was wondering if you could give us some color on the disruptions you see to NAND and DRAM production as a result of the events in Japan. Have any of these machines been turned off? And I've got a follow-up.
Eric Meurice
Are you specific on our machines in Japan?
Adrien Bommelaer
Yeah exactly; I mean, you must have a pretty good view as to utilization rates and what's basically running and what's not running.
Eric Meurice
So no impact during the quake. And the quakes, the machine stops and then it restarts. And it takes some time. So on the lithography part of it there is no significant impact. And our utilization levels and are at the same levels as before.
Adrien Bommelaer
And just a follow-up. Coming back to Japan. And again, I know you talked about this already, but you seem to be saying that the delta basically in spending would be linked to Memory. And now if I look at the split between DRAM and NAND this year, DRAM is not really spending, so it's essentially NAND. When you look at NAND, its actually fairly tight because of the success of the iPad and so forth. So how does the Japan problem actually impact NAND spending this year? And I am trying to understand what is the thought process of most of the spenders, and if you can maybe quantify if that's possible.
Eric Meurice
As I say, the sentiment is more of thinking that maybe things will not be roses, so you have to be cautious. So everybody in the NAND and the DRAM business is going to push out some to level though their enthusiasm, but there is nothing quantitative into this. It's one, two machines in fab, there and there until they get more data. So I think your questions would be more relevant I think if and when people get more specifics at the whole chain.
Peter Wennink
And I think customers were quite specific on this in telling us, okay, this is what we feel today, but please be prepared to turn around in a month's time. So this just shows you the sentiments right now. On a certain day they ask us to think with them and to be with them, but also to be with them when they are wrong and they want those (tools). So they say, this would change every month, so please be prepared.
Adrien Bommelaer
So this could turn around exactly in the next quarter in a positive manner.
Peter Wennink
Yes. That could happen. At least that is what our customers told us that be prepared that that could happen. We simply are too uncertain at this moment im fact to give you exact guidance, but this is what we feel today. We might feel different in four weeks time.
Operator
The next question is from Mr. Jérôme Ramel from Exane. Jérôme Ramel: Quick question on EUV. Is the situation from your laser supplier the same for each of them? And just from a (inaudible) point, would it make sense for you to only focus on one supplier instead of three?
Eric Meurice
Let me answer the second question. First is, there is a chance that we have three suppliers with three different ways of doing some things. So there is just risk of the total industry. And as I said before, EUV is so important to this industry, they need it, everybody needs it, that it is the duty of the industry to have multiple ways of trying to find solutions. And we are part of it. Of course it costs us more money because we support all the compatibility and its not a small amount of money. But we need to do this to reduce the risk of our customers. All of them are within their, I would say original roadmap plus one year. So they are all guilty of the same crime, all for a delay of about the same amount of time. But for a while now they have been stabilizing a roadmap of improvement with Cymer, Ushio and Gigaphoton showing us to their own speed certain numbers of progress. Jérôme Ramel: And maybe just a follow-up. Would you prefer triple or quadruple or do you prefer EUV from the P&L standpoint for ASML?
Peter Wennink
Peter here would kill EUV for the first two years. So the financial impact for a new product - EUV is one of them, and of course it's a big one, is always negative in the first two years. So the margin is not going to be as good because the way we price the machine is, we price a machine at the target price. So at the beginning, the machine has a warranty impact. And also on our side has a multiple obsolescence in terms of parts, where we improve parts, so we cannot use what we had planned before etcetera, etcetera. So the first two years is bad news. But of course the next years we enjoy what we have been enjoying now for a while, which is leadership in high technology, and therefore we will be more happy financial wise if we stayed in the commodity business. So we do this because it's necessary for the industry. But we also do this because it is going to be in due point a profitable business of course.
Operator
The next question is from Mr. Niels de Zwart from ING.
Niels de Zwart
My first question would be on the increase of your R&D spending going forward. You mentioned, it relates to strategic projects. Could you maybe elaborate a bit more if it just relates to your current, so to speak little business or if there are also any additional investments going into your incubator activities?
Peter Wennink
It is particularly just to do with EUV, like I said in the introductory statement. It has to do with brining our second generation tool to the market, which is 3100, and the development of the 3300. And also in a large part has to do with also the latest situation that was discussed before. The investments, lets say the money that we put aside for incubator projects has basically stayed at the same level, and that's not the reason why we increased the R&D.
Niels de Zwart
A kind of follow-up, you just mentioned that of course in the first few years of shipments the EUV tools will come with a lower gross margin. Is there any way you can quantify the impact of that on the overall gross margin you would expect, assuming that the revenue level otherwise would have stayed the same?
Peter Wennink
Yes, you can look at what happened in the past, although that is not a proxy for what's going to happen now always. But you would have to think of new systems that we ship with gross margins in the mid to the high 20s. And that in a three-year period, generally we have a three-year learning curve that brings us to the corporate gross margin. Now, luckily, at the introduction of a new tool, as a lot of those tools is going to be (a few) tools, and the bulk of our shipments will of course be our bread and butter product that we know how to make and where we are constantly looking at improving the gross margins on it. So really, if you want a number, think about this. Although we are still a year-and-a-half away from the first shipment, so still a lot can be done to focus on this gross margin issue. But like I said, there's not going to be a lot of tools in the first 12 to 18 months, and that gives us an ample opportunity to start working on the gross margin.
Niels de Zwart
A final question; if I look back at the last quarter's earnings presentation, your slide presented that you would ship all your NXE:300 systems during 2012. Now I get the feeling that you mentioned that would be challenging. Is it just a different way of saying the same? Arguably, something has changed in your expectations on the base in which you can ship the 3300 systems.
Eric Meurice
What has not changed is the customers' appetite. So we can still do it. What has changed it, but this is not abnormal is the complexity of doing all this at the same time is significant. I would not change my excess (inaudible) at this moment, we don't know enough. Some of the improvement of the 3300 to the 3100 is very straightforward. In fact, a very large part of the improvement of throughput between 60 and 100 is really within our own responsibility, it's not even a laser boost. So this, for instance, we will do pretty easily. For instance, imagine that we get the laser people to stabilize a number in three months; then we can still do 2012 easily. But if we don't, then we are going to have some delays. So again, at this moment I would still keep the target of ten in 2012.
Peter Wennink
Ladies and gentlemen, I think we have time for one last call. If you were unable to get through and still have a question, you can feel free to contact your friendly IR contact. We would be glad to follow up with you. And operator, now if we can have the last question please.
Operator
The next question is from (inaudible).
Unidentified Analyst
First, Eric basically you have said that you saw some revenues that got pushed out from 2011 into 2012. And I think you also said that this was quite minimal. I was just wondering. I mean, if I think about it, it basically means that from a bookings perspective it lowers Q2 bookings and will like lead to a comeback in Q3 or Q4 this year. Is there any way you can put a number kind of to that revenue that has been delayed? And I have a quick follow up as well.
Eric Meurice
Yes and no. Yes, it is exactly as you said; the revenue that we have been discussing being delayed is not going to materialize into Q2 booking which is why we have announced only €900 billion to €1 billion. But no, we don't want to quantity it because the key question is about how is 2012 going to look? And I think it's a bit too early to guide on it.
Unidentified Analyst
And a follow up. Coming back to EUV, there are obviously, for the EUV ecosystem there are a few things that are not directly in your hands like the inspection tools for the masks cleaning and photoresist. You have mentioned a few things earlier. I just wondered where you feel that the most work still has to be done and which of those other products are well on track in order to make EUV ramp as you expected at the moment?
Eric Meurice
I am sure they are not on track and I am sure they are terrible suppliers like we are. But to be honest, it is an irrelevant question. The customers are now putting momentum on it. The problems to resolve are smaller scale than our own problems. So we are the critical pass, they are not. But if you talk to a customer, they will still be pissed off at mask inspection not getting there. There is not enough interest to buy the (inaudible) people who are making (blanks). But again, this is irrelevant; this has no business impact really.
Craig DeYoung
Okay, on behalf of ASML's Board and management I'd like to thank you for joining us on the call today. And operator, if you could help us by formally concluding the call, we would appreciate it.
Operator
Ladies and gentlemen, this concludes the ASML 2011 first quarter results conference call. Thank you for participating. You may now disconnect your line. Have a nice day.