艾司摩爾 (ASML) 2007 Q1 法說會逐字稿

Ladies and gentlemen [technical difficulty] the first quarter results conference call on April 18, 2007.

[OPERATOR INSTRUCTIONS].

I would now like to turn the call over to Mr.

Craig DeYoung.

Go ahead please sir.

Thank you, operator.

Good afternoon and good morning ladies and gentlemen.

This is Craig DeYoung, Head of Investor Relations at ASML.

Welcome to our investor call and webcast.

The subject of today's call is ASML's 2007 first quarter results.

And with me here today and hosting this call are ASML's CEO, Eric Meurice, and CFO, Peter Wennink.

I'd like to draw your attention to the Safe Harbor statement contained in today's press release and presentation.

You can find both the press release and the presentation on our website at www.asml.com.

The length of this call will be 60 minutes.

And now at this point I'd like to turn over to Eric Meurice for a brief introduction.

Thank you, Craig.

Good afternoon and good morning.

Thank you for attending our conference call.

As it was the case in last quarter's conference call, we will split the introduction into two parts.

Peter will start with a review of our Q1 financial performance with added comments on the outlook for Q2, and I will complete the introduction with a longer-term view of our business opportunities.

Following the introduction, we will open up to the call -- the call to -- for questions.

Peter, please?

Thank you, Eric.

And welcome to everyone.

First of all, like Eric said, I would like to review our Q1 results highlights.

And in that regard I'm happy to report that our Q1 results were solid, as we have predicted.

In reviewing our results in detail, you've seen that we reported healthy revenues, operating margins and net profit for the quarter.

Q1 revenues were recorded at EUR960m, roughly 50% higher than the same period last year.

As indicated during our Q4 earnings call, the average selling price for the quarter was impacted by shipments of i-line tools as we shipped relatively large volumes of XT400s to newly constructed memory fabs.

This is a direct result of market share gains for this product line, resulting from our early efforts to significantly increase the varied ownership of these systems.

On the operating side, our operating cost grew 26% year on year.

It was mainly driven by increased R&D spend.

This resulted in an operating margin in the quarter of 20.5%.

And that compares to 18.2% the same time last year.

However, our first quarter operating results do reflect the impact of the Brion acquisition which had an impact on our operating income of around EUR25m, EUR23m of which was for one-time non-recurring R&D charges.

Without the impact of Brion, our operating income would have been EUR220m, or close to 23%.

With all this focus on the growth and market share gains, I must note that we have not lost sight of the level of our operational expenditure.

Although we have grown our operational cost base to support future growth, in both our R&D spend and SG&A cost, we do focus on creating cost variability, such that we can manage the profitability through the cycle.

It's about 30% of the R&D costs are variable within a quarter's time, and approximately 10% of SG&A costs are variable.

And it makes ASML pretty flexible and adaptive to any market environment.

And although we don't communicate our unit breakeven level since we believe that would really only be relevant in extreme business conditions, we do focus on it from an internal point of view.

And the internal regular reviews on it are pretty clear.

And it has remained at a breakeven level at a constant low level, despite our increase in the absolute operational cost levels.

IC unit growth forecasts and also the continued drive of the industry for leading-edge shrink capability will keep supporting the lithography equipment growth.

I think this was clearly evidenced by a good level of order intake in Q1, reaching a level of EUR911m, leading to another record backlog of close to EUR2.2b.

Although the order intake was lower in units as compared to previous quarter, the demand for our leading edge products continues to be high as clearly evidenced by Q1 order average selling price of EUR15.2m for our new systems.

Also in Q1 we continued to execute our share buyback program, with the purchase of an additional 8m shares, which is about 1.3% of outstanding shares.

And that was completed -- that actually completed the 10% share buyback as authorized by our shareholders in March of last year.

And since we feel confident in our ability to keep generating robust cash flows, we've asked for and were given the authorization to actually additionally buy back a significant number of ASML shares through September 2008.

Net cash from operations was EUR173m in Q1.

That allowed us to end the quarter with cash between EUR1.4 and EUR1.5b after having used in the first quarter about EUR350m for the share buyback and the finalization of the Brion acquisition.

Now, having finalized the Q1 review, that brings us to the outlook for Q2 -- the outlook for Q2.

With regard to bookings, as we mentioned, by the way, in our previous earnings call, the current market conditions make it difficult to call the exact timing of the order booking.

However, we do believe that technology and capacity needs of our customers in the second half will ensure an appropriate level of order intake to justify our growth expectations for the full year.

And I would also like to point out that when assessing our quarterly bookings levels, it's clear that focus on unit bookings is less relevant now.

It is due to the highly discrepant average selling prices for leading edge versus trading edge 300 millimeter systems.

And our order book at the end of Q1 is very healthy.

And we also expect that our order book, our Q2 backlog levels will remain as supportive to our expected 2007 growth scenario.

As expected, the Q1 orders from our Flash and DRAM customers decreased in comparison with the last quarter as these customers have planned a front-loaded year.

Our logic customers continue the effort to increase capacity in line with what we think is sound IC demand, which resulted in overall in the strong order intake in Q1.

And ordering patterns of the different semiconductor market segments, they compensate each other providing robustness of the total market.

Now, we can also confirm that our optimism about the year is also based on the industry analyst average IC unit forecast by segment, as is regularly provided by Gartner Dataquest, VLSI and iSuppli, and this is currently supporting the customer plans of capacity growth for litho tools that we currently foresee.

And as a consequence of this analysis per segment which is confirmed by our Q1 results plus our strong Q1 bookings, our confidence for growth, like I said before, in 2007 is reassured and is also clearly supported by a record backlog of which 78% is shippable over the next six months and almost all of it before the end of the year.

Now, I think that closes my outlook on Q2 and I'd like to turn back to you, Eric, for the remaining.

Thank you, Peter.

So you did the easy things.

Good short-term and mid-term view thanks to a good backlog.

So let's talk about longer term.

Longer term, ASML continues to focus on technology leadership.

And at this moment we are working on four areas.

Immersion, value lithography with the integration of Brion which we just acquired.

Third area is improved TWINSCAN technology to address double patterning, and fourth area, EUV.

So let me spend a bit of time on each of those four areas.

First, immersion.

ASML continues to strengthen its lead in the area of immersion.

We have now shipped over 40 immersion tools, a significant number and on which a significant portion of those tools are already producing Flash memory device in volume.

With the last machine that we have shipped in immersion, the XT:1700, customers are able to do three things.

They are able to do production, immediate production at 65 nanometer node.

This one is not requiring immersion but it is enabling -- it enables better performance with immersion.

But they also, in parallel, ramp 55 nanometer, which is necessary, which requires basically immersion.

And they are also starting the development of their 45 nanometer half-pitch technology.

So this XT:1700 product is, in fact, the only system on the market which can do so -- which can stretch as far as 45 nanometer.

Leading edge Flash customers are even now reaching regularly over 1,000 wafers per day on these tools and are working now on ramping up to 2,000 wafers per day per tool.

We're talking production volume.

In fact, we are talking about mature production volume when you talk about 1,000 wafers per day to 2,000 wafer per day per machine.

In total, over 100,000 wafers at this moment are treated per month on our immersion machines.

As you know, there is significant risk in switching lithography platform.

And the fact that the customers have been defining their production processes on these machines does solidify our relationships for the future.

Furthermore, our new machine, the XT:1900i, which is due for shipment very early into Q3, has already imaged at 36.5 nanometer, which is a world first, unique performance.

And we have confirmed that this machine, the XT:1900i, has confirmed immersion and throughput performance on the current product, for the types that we have.

This was expected because the 1900 is in fact a derivative of the 1700, which means that all the experience accumulated on the 1700 makes introduction of the 1900 much easier.

So, again, we expect therefore a smooth production transition from the 1700 to the 1900 at our customers.

In parallel to our immersion development, with the Brion acquisition now complete, we are starting an aggressive development for ground-floor exploitation of the synergistic effects of Brion's computational capabilities and ASML's competence -- scanner competence.

As we have mentioned before, the objective of this acquisition is to reinforce Brion's business in IC design verification and in our ET or PC businesses market, while Brion will help, in return, ASML develop lithography platforms which have capabilities to optimize the scanner setting per specific application, so a value proposition, an additional value proposition for ASML.

The customer benefit is an ability to maximize yield.

That's the first good news.

But the second good news is to be able to define more critical design rules which will, in turn, increase the device performance.

These synergies between ASML and Brion will enable each to achieve greater value generation and improve competitive differentiation.

The third aspect, as we've discussed, is double patterning.

We know at this moment that EUV is confirming its potential as the most extendable and cost-effective long-term solution for technology nodes beyond 32 nanometer.

Our progress with customers and partners on our two EUV alpha tools being installed is allowing the industry to address the infrastructure needed.

But certainly, this will require a bit of maturity.

And we expect EUV to ramp into production in 2010 or 2011, although, as I repeat, it is fundamental that the work is started now to create the infrastructure necessary.

But, in the meantime, we believe that double patterning technology will be the necessary technology employed as a bridge technology up to EUV.

ASML is focusing significant resources to develop lithography solutions for this new application, double patterning, so as to support higher throughput required by this technology and more accurate overlay.

So, in summary, we confirm our robust execution in terms of market share progress, our robust financial results.

We confirm our short-term and mid-term growth forecasts for 2007, and that was Peter's point.

And we confirm our value generation proposition for the longer term, which was my point.

So, on this, Peter and I are ready to take your questions.

Operator, before I turn to you for the questions, I'd like to --couple of ground rules.

I'd ask that you please identify yourself and your organization when you're called on line.

We'd also like to kindly ask that you limit your question to one, with one short follow-up if necessary.

This will allow us to get as many callers on line as possible.

So, operator, could you have the first question please.

[OPERATOR INSTRUCTIONS].

The first question is from Mr.

Harm Luttikhedde.

Go ahead please sir.

Please state your company name.

Betten Financial News.

My question is about the dividend.

You mention in your press release the dividend is certainly an option in slower growth model.

Could you explain a bit more about that?

Yes.

We think that the best way to return cash to our shareholders momentarily is buying back shares because we do believe in the growth profile of this Company, and we think that attributing that growth profile to a lower number of outstanding shares will create bigger value.

Now, when that growth profile somewhere in the future is going to tail off and then, I think, that argument does not hold any more and then dividend is a very good alternative for returning cash.

Could you be more specific about the date when you're thinking of paying a dividend?

No.

No, we think we are going to grow for a long, long time, Harm.

All right.

Thank you.

The next question is from Mr.

Simon Schafer.

Please state your company name followed by your question.

Hi.

It's Simon Schafer with Goldman Sachs.

I was wondering whether you could share some more detail with respect to the Brion acquisition now that it's closed.

Any type of more updated guidance or targets that you may have -- you may want to share in terms of the operating profile or the type of revenue growth that we may be looking for?

No.

As you know, Brion is a private company, had or executed a certain numbers of products in different segments.

Now that they are integrated since a month we are reviewing the business plan of each of those segments and trying to take decisions on the appropriate place to put the resources and accelerate the market segments with the highest potential.

So this activity is taking place at this very moment.

And we will not be able, I guess, to present anything of material value before the end of the year.

So, in the meantime, trust the fact that we have acquired the company at $275m so -- because we believe there is potential for more than that.

We are not disappointed about the quality of the people that we have acquired, basically.

And we, however, need some time because this is a long-term business decision that we have taken.

Like Eric said, it's not going to have a significant effect on our -- the results for this year.

Thanks.

And then a quick follow-up on the immersion side, how many units do you have in your backlog and how does that break up between 1700s and 1900 side?

We will try to avoid making the differentiation between the 1700/1900, not for any other reason that when we talk about this for the [reason] in the quarterly results, the questions, it's a yo-yo with customers sometimes preferring 1700s and then converting to 1900 and vice versa because, in fact, both of those products will survive.

Both offer a different function to the customers.

So this ratio is a bit irrelevant at this moment.

It changes all the time and we don't even know where it will go.

But in terms of backlog, at this moment, we have got 24 combined units and we shipped four in Q1.

So, so far, in three months of activity basically, we have accumulated 28 potential for the year and, of course, we expect to book a bit more than that.

So we are well on the way to continue leadership into immersion during the year.

Great.

Thanks very much.

The next question is from Mr.

Nicolas Gaudois.

Please state your company name followed by your question.

Yes.

Hi there.

Nicolas Gaudois, Deutsche Bank.

First question is specifically on some of the backlog movements.

We've seen memory backlog coming down, and Korea seems to make all of the delta effectively.

Taiwan backlog is flat and, within that, foundries is flat as well.

So does this imply that you saw an increase in your Taiwan DRAM backlog in the quarter, while actually rest of memory U.S.

and Europe was down?

And if that is the case, what could we see as a potential offset if, by Q3, some of these Taiwanese DRAM customers decide to actually delay a bit some of the delivery dates looking obviously at what your prices have done recently?

Yes, Nicolas, good question.

In fact, what is happening now is exactly what we had modeled and communicated in the past, I think, two conference calls, where we said for the year 2007 we will see front-loaded Flash which is a lot of revenue or billings into Q1 and Q2, less in Q3 and Q4.

But then we said DRAM was spread Q1, Q2, Q3, less in Q4.

And then we said foundry is going to help us out into, in fact, Q3 and Q4 as well as other logic customers being spread.

So nicely spread, was a good number.

So this is basically developing.

And this is what you see with those percentage backlog where Flash reduces, DRAM goes up, foundry goes up and you should see also IC going up in percentage.

Is there a risk?

So this is basically written on the wall since six, nine months.

We have said that and we are executing to this.

And we don't see change to that.

Your question then, your hidden question is, is there something that can happen in Q3 in the market that would make some customers decide to delay their receipt of our machine?

Well, first of all, legally they can't because our orders are non-cancelable.

But obviously when they have good reasons, and there is no need for capacity, we are always willing to spread because there is no interest for anybody to force capacity that is not used.

On this case, we are fairly comfortable because you are talking about DRAM at the end of the year which is usually a place where nobody can call the need for the Christmas season, etc., very well, and therefore most of the capacity builds are strategic.

And this one is even bigger as a strategic move because most of those Taiwanese are basically building new fabs.

So, you've got two reasons to move strategically which are less impacted by some kind of a fear of a short-term price question.

So I would be highly surprised that we would receive phone calls for pleading for push out of our DRAM plans for Q3.

Fair enough.

And just as a very quick follow-up, if we look at your Japanese backlog, it stepped up basically in Q1 '06 and then it's -- and Q2 again.

And then it's been flat since then essentially three quarters in a row, at a decent level, 6% of backlog.

But when and how -- what will be the drivers to actually start to see this backlog portion coming from Japan effectively take another step up in the next 12 months or so?

We are -- we were very, very happy with our progression in Japan.

The key measure at this moment is probably more market share than it is backlog percentage.

In the percentage, the market share is, as you know, and has been published, starting to get above 10% in the past six months or so and is heading towards 20% [as I see it] and above.

So that, yes, of course, translates into some machines coming in.

These machines coming in are the high-level machines, the high ASP.

And it is therefore now difficult for us to assess which quarter is going to see them happening and which quarter we ship and which quarter we take them in.

And they make big difference, lumpy differences.

And this is not -- I am not hiding anything by saying we are not shipping a lot of i-line in Japan at this moment, and not too many KrF.

And we are in the ArF Flash immersion business.

So I can't really tell you where you will be starting to see the color of Japan much bigger in the backlog.

But I can tell you that I see market share above 20% or so soon.

And above that, Nic, 6% of backlog, that definitely has been rising in absolute terms.

So in absolute levels, our business in Japan has actually grown.

And just for your indication, of the order intake, the bookings in Q1, 11% of the bookings value was Japan, so that's more than 6%.

Okay.

I was looking at the actual value, to be honest, in the backlog.

But that's fair enough.

Thank you very much guys for these questions.

The next question is from Mr.

Jay Deahna.

Please state your company name followed by your question.

Thanks.

Jay Deahna with JPMorgan.

Congratulations on great execution.

More than one question, but quick ones.

The first one is given the mix shift towards higher value products in that you've got some i-line and KrF through the pipeline recently, does the order value in 1Q feel like a drop?

That's the first question.

The second question, you're seemingly slightly incrementally more confident in this year being a growth year.

Are you assuming that this is a growth year driven just by ASPs for ASML or does that include some unit growth as well?

And then the third one is, are you still looking at 35 to 40 immersion tools this year?

And then the last one is what are you thinking in terms of the possibility for turns business for Flash later in the year given that Samsung Semiconductor's [CEO] said recently that Flash could be in severe shortage in the second half?

Well, Peter is going to try to handle your first two questions and I am preparing myself for the last ones.

Go ahead please.

I'm just writing -- making some notes here.

Jay, first, the question on mix and high value, whether that would indicate, that Q1 would indicate an order drop, I don't think they're correlated, to be honest.

It's just when we look at where the demand for the year comes from, basically you're leading also to your second question, is that we do believe that in the segments IDM, DRAM and also Flash, we have very -- we have significant unit growth numbers that force our customers to put capacity orders in.

But also when we look at their shrink roadmaps that they need this technology that we will be introducing in the second half of the year.

I think that is driving really when customers or what customers will place in terms of an order and when that order comes in, and that's also why we are not guiding because of those market conditions and there's a level of uncertainty there, that we don't care that much when they come in.

And timing of the order book is not that important.

For us it is more important that we have a strong view on what we think our customers need in terms of capacity against the background of their unit growth, our market share gains and the technology introductions.

And that also comes to the answer to the question two, we think it's going to be a growth year, driven, I could say, in the second half of the year clearly by ASP because there is the 1900 introduction.

And you will also see there the majority of the immersion shipments.

I think market share will account for some unit numbers.

But I would say generally it is the quality in ASP that is driving the top line this year.

Okay.

And then the -- just one quick clarification on the first response, I wasn't clear exactly what you were saying.

The question was does 1Q look like possibly a value order drop for the year?

And you're -- am I correct in you're saying maybe, but it depends on timing of orders and you don't care when the orders are made?

Yes.

Correct.

Correct.

And also when you look at the remainder of the year, we do expect an immersion ramp up in 2008.

Clearly, going forward, the ASP trend of this Company is, of course, up.

So yes, that is, I think, what we care about.

We care about market share in that segment.

We care about the ASP trends.

And we're pretty confident that there's going to be top-line growth as a result of that.

And when those order timing comes in doesn't really matter that much.

Safe to assume that the downside risk from 1Q is relatively limited given that you've got a lot of immersion in front of you?

We would again, we say there, Jay, we don't know.

We are pretty comfortable on the capacity terms.

And that's why we are calling in the billings.

We [really] done this, now we've done this and said the year is going to be growth, the year is going to be growth, because we know which those projects are.

But it is true that knowing which customer -- sorry, which product each customer will take and when they take this, it's very difficult to call.

Today, as I said to one of the callers, we don't even know if they will want the 1700 or the 1900.

This is based on them decided whether they want to drive more 55 nano or more 45 nano, then it is also based on us.

If the 1900 runs faster than we think, then they are going to jump on 1900 because they get more headroom.

But -- so in other terms, imagine that we are successful and introduce our 1900 a bit earlier than necessary, that will freeze orders of the 1700.

You understand my point?

So you've got complexity about the timing which is why we're not wanting to call bookings.

But we call billings and you show fairly significant, we show fairly significant body language on that.

Yes, I think we're going to have to move on.

Operator, can we have the next call.

Immersion.

The immersion question, yes, we -- in fact, we've read the literature.

Gartner says about 45 immersion tools in the year.

You know that our competitor thinks -- says it may be much bigger than that.

At this moment, on the total 45 -- total market, total market, we've got, as I said, 24 in backlog, four shipped, that's 28.

You can imagine that the 35-ish to 40-ish range is absolutely within reach.

Regarding the Flash turns, absolutely.

We are confirming that the current capacity plan that we are building for our customers from now until Q4 will get to a near full utilization by the end of the year if, of course, the statistics that have been discussed by the Dataquest and so -- are correct on the units.

If that's the case, there is opportunity -- there are opportunities for [returns].

If this is the case, there is an upside potential, yes.

Thank you.

The next question is from Mr.

Mehdi Hosseini.

Please state your company name followed by your question.

Thank you.

Mehdi Hosseini, Friedman Billings Ramsey.

My take of it so far from the conference call is that there's definitely a downtick in the unit shipments of immersion for this year.

And I want to ask a question that would help me understand what's going on and maybe answer the question about the immersion, how it's tracking.

When you look at the business opportunities between Flash and DRAM and logic, is there any way you could quantify the units of immersion used for like a 30K per month Flash DRAM capacity versus logic?

In other words, if you were to compare double patterning with immersion, is there any way we could think about the difference in the number of units used in a typical fab with a 30K per month capacity?

Okay.

This is, of course, a complex question.

Let me try my best with different facts.

First of all, I don't think we are saying that immersion is lower than before.

I think we -- there was a number running around of 35 to 40 units for us.

And we just confirmed that this is within reach.

The -- if you remember, a long time ago I was asked what is the typical pattern for a new technology like immersion to grow, and I said, well, if you were to look at what we did with the 1400, which is a former generation, I said about 70 units per 18 months was what we achieved.

And at the time I said I don't know if that could be replicated.

To immersion, we will see, but that's another [inaudible].

If you look now at 35 to 40 units in the 12 months and then you add the 25 units or so that we shipped six months ago in the last six months of the year, whatever, then you're talking about already nearly 65 units, which is not far from my 70, which happened on a non -- on a simpler technology.

So you -- we are confirming that, in fact, everything is happening as planned again at this moment.

Regarding probably the background of your question which was what's happening between immersion and double patterning, and who is going to win, and how does that impact lithography business?

Honestly we don't care.

If we -- if you do double patterning you will use more of our 1400 and more of our i-line tools.

And if you buy -- well if you don't do double patterning, you will pay more for immersion at a lower throughput than the 1400.

What is the exact equation that says how much of which is which?

Honestly I don't know.

It's highly complicated.

It depends on how many layers you do in immersion.

It depends how many -- what tool -- type of double patterning you are using.

You've got some people doing double patterning in logic that has significant passes.

And there are some other people doing double patterning in memory.

All this is highly different.

So, honestly we feel very, very happy to be very well positioned in double patterning and in immersion.

We have a 1450 machine perform faster with better overlay for double patterning.

I think we are the only one in the market with this.

And we have here, of course, our leadership in immersion.

So, be a customer, be our guest, we love it.

If I may just one follow up, your days of inventory went up significantly, and you only handled four immersion booking in the Q1.

With the days of inventory going up, and assuming six to nine months lead time for immersion tools, does that imply that you are going to see sort of a hockey stick in immersion shipment in the latter part of the year?

Well first of all I correct you, I said four shipments in Q1 and I should have said eight bookings in Q1.

Sorry about that yes.

So -- and we expect, yes, a good set of bookings of immersion in Q2.

But as I said the last week of Q2 somebody is going to tell me, 'where is the 1900, and I won't tell you, I won't give you your bookings before I know if I want you to switch my potential 1700 to a 1900'.

So you are going to have some orders into Q2, yes.

But it is possible that we will get more orders of immersion in the July/August timeframe, because they will want to see how good the 1900 is.

But yes, we are going to have billings significant into Q3 and Q4 because this is when we are ramping the 1900.

Thank you.

Next question comes from Mr.

Timothy Arcuri.

Please state your company name followed by your question.

Citigroup, thank you.

I want to talk a little bit about the 200mm/300mm upgrade in the memory world.

I wanted to ask you two things, number one what's your view on the obsolescence of the existing 200mm capacity in the ground?

Is there a fundamental technical or cost issue from a litho perspective that is going to force the industry to go and replace all this big 200mm capacity in the ground?

And then I had a follow up, thanks.

I -- in fact, Peter you may -- it's difficult to know the economics of a fab.

Sometimes fabs are limited by square meters, in which case we have some customers who are driving us to move from 200mm to 300mm.

So, we basically -- square meters and people available means that some of those guys say output per square meter is a fundamental, in which case there is no question.

So we are doing some of that.

And some of the customers are saying, look I am happy with my square meters throughput, I can buy -- I can build the 300mm fab outside, so I want both.

And you do have some people still asking us to do, in fact, today -- well, today, six months ago we had one or two customers asking us to allow a 1900 200mm machine.

So, in other terms I think you can probably justify both economics.

It depends on the customer also.

It very much depends on the customer.

And like Eric said we have seen both on the 1400 and the 1700, but also with a request for the 1900 to make a 200mm compatible.

And the only reason why they do that is to extend the life of the 200m fabs.

But in principle if you would have a fully depreciated 200mm fab and a fully depreciated 300mm fab, it is clear that the cost advantages go in the direction of the 300mm fab.

Now, when we get more 300mm fabs in a depreciation phase where it's more comparable with the 200mm fabs, then yes you will get more shifts over.

And what we are seeing with some customers that also the 200mm tools are being used for specialty DRAM products.

They are being remarketed and refurbished onto the second-hand foundry market.

So there is some 200mm business in the refurbishment area that is a result of some customers going to 300mm, but it's not significant at this moment.

Okay, great.

And then last thing, have you done any analysis that might try to pinpoint what the total, or what your total dollar opportunity is related to the obsolescence of this 200mm capacity?

So, as the memory companies go out and replace this capacity in the next few years, have you done any work on what your opportunity is there?

And do you think any of that money has been spent yet?

Thanks.

We have not because even if that doesn't happen we have simulated significant growth anyway.

So there is a definite upside potential if there is an obsolescence wave that happens, you are right.

We -- but we are showing growth anyway just by the build up of the -- to the 300mm capacity.

And again Tim we haven't seen a lot of activity in that area yet.

Okay, so there's not much of that money that's been spent yet from your view?

Correct.

Okay thanks.

Next question comes from Mr.

Thomas Brenier.

Please state your company name followed by your question.

Yes, good afternoon to you.

This is Thomas Brenier from Societe Generale.

Got a very general question if you could help us on that, if you could compare the investment you did for an equivalent fab in terms of lithography between moving from 90 to 65 nanometer and moving from 65 to 45 nanometers.

How much would be the increase or the difference?

Well given -- the only thing I can do -- well we can, yes you can probably call Franki and Craig to get more.

But I can tell you that a 1400, no 1600 -- okay, fine, you'd better call the guys.

But obviously the ASP has grown as you know from EUR15m per machine at 90 to EUR20m at 65 to now at EUR30m or so.

And the throughput originally, where 150 machines for the cheap machine, and now 120.

So you can make some calculations about the fact that, yes, 45 nanometer cost significantly more in lithography, which is by the way one of our opportunities.

But again, I will push you back to --

We have the data by the way, we have the data.

So if you could call either Craig DeYoung or Franki they will, they can give you the answer.

Okay thanks.

And then could you maybe update us, I am not sure anything has changed, but you are on your way to go to a size where you can address between EUR4 to EUR5b in sales.

And I think last time we discussed that you were mentioning more EUR5b.

Do you have more visibility as far as when you think you can get to this level, in terms of your ability to serve the market, and the ability of the market to take that?

No we gave a direction of EUR5b sales target by the year 2010, because we believe this is what the market requires in view of the new technologies, and the famous 8 to 10% semiconductor unit growth that people would consider is appropriate.

So, I am just saying this market share of about [17%] market share plus this growth will generate EUR5b.

But we didn't want to say that we knew anything about cycles.

And, therefore, it could be plus or minus two years depending on where the industry is.

And at this point we are not calling cycles long term; we don't know better.

But we are running on an intrinsic curve that should get us at EUR5b by 2010 if there is basically no cycle any longer and everything is loose at 10% unit growth of semiconductor.

Okay thanks.

Your next question comes from Mr.

Jonathan Crossfield.

Please state your company name followed by your question.

Yes it's Merrill Lynch.

My question was just on the change from a EUR1b net cash target to a gross cash of EUR1 to EUR1.5b.

First of all could you just give us a bit of background on why you've changed from net to gross?

And secondly what's the basis -- or what will determine what level of cash you target within that range?

Yes, well the -- to answer that last question first, the EUR1 to EUR1.5b is basically depending on what we need for the cycle.

Well, we generally don't need a lot of cash for a down cycle, where we might need more cash for the strength of an up cycle.

And if you talk about a EUR5b plus company it could be that in a ramp-up to EUR5b, the working capital needs of the Company, because we are still largely pre-financing everything that we ship to our customers in the work in process and in accounts receivable, that that will take a lot of the cash.

And that will then determine let's say the upper side of that range.

And in areas where a -- and in times of a cycle where the business is not at its peak then we could do with, in principle, less cash and that is the lower side of that particular range.

The reason why we changed from net cash to gross cash is because we feel that even in a -- if you would look at our current cash balance, which is about EUR1.4b and that is about a EUR1b net cash.

In cycles when we are looking not to invest that much money in working capital we could actually do with a lower cash balance.

So it's less conservative and we feel it is appropriate under the circumstances.

We are a growing Company.

We take care of our working capital needs.

So I think also from a cash target point of view we can do with some less conservative targets.

And that's why we communicated basically like that.

Great, thanks very much Peter.

The next question comes from Mr.

Adrian.

Go ahead, sir.

Please state your company name followed by your question.

[technical difficulty] for Peter, just can I ask -- following on from the last question actually, the rather open ended comment relating to buybacks that you have in your press release.

Yes.

Is there anything that you need to do internally other than the AGM authorization that you already have to actually action a buyback?

I am just trying to work out what the milestones from here, but --?

I think its -- it's in principle, it's not that.

We have to go through the normal statutory and regulatory approvals in our Company.

And we just need to make sure that everything is put in place from, let's say, a legal and a tax point of view, and that we can execute.

So that is -- in principle nothing that would hinder us.

Okay, and just in terms of mix of business if we look at the backlog, backlog that memory -- the proportion of backlog has come down from 64% end of Q4 to 54% now.

When do you think that normalizes, and what do you think a normal level of memory as a proportion of total backlog is, given the current environment?

I think we have answered that also, a quarter ago where we said it's around 50%.

That's what we said, about half of our business -- that's where we currently are.

I think that is what we would regard as, let's say, a more stable division between the three segments.

Okay that's great, thank you.

Next question comes from Mr.

Robert Maire.

Please state your company name followed by your question.

Yes, Needham & Company.

Similar to the question you just answered, in terms of share of the foundry versus IDM, versus memory going forward over the year, would you expect that strength in foundry offsets perhaps the little weakness or softness in the memory market, while IDM remains flat?

How do you see that playing out over the year?

Do you see foundry picking up more aggressively in the second half, or fist half loaded?

And do you see any weakness in memory being more weaker in the second half?

And I know you don't want to give order guidance, but in terms of one offsetting the other.

And the second question, at a recent supplier's conference for a major logic manufacturer there seemed to be some aggressive talk about 450mm wafer size being pushed.

And I'd like to get your comments on that, and your R&D plans and development plans for potential other wafer sizes.

Well to answer your first question how we see the year panning out, I think it's also a bit like we said in the previous call.

Clearly from a shipment point of view we've seen -- it's evident by the way of our shipments in Q1 where close to 70% were shipped to memory customers, where that's where we said it, first half of the year the shipment to memory customers.

We are seeing, from the utilization that we are seeing from our tools, we are seeing that clear indication that we have seen a [fluff] in the foundry business in Q1.

Now that means that they are still not at utilization levels where they feel a lot of capacity strain.

But with that inflection point in Q1 it is also not unreasonable to suspect that by the end of the year in Q4 they will need extra capacity, and that means that they will need to start ordering that over Q2 and Q3.

That, in fact, would offset indeed the weakness in terms of shipments from the memory business.

Because if you look at our bookings, yes of course, the bookings for memory customers came down in Q1 as we would have expected, since the shipments are frontloaded and they don't need a lot in the second half, except of course the new technology.

So yes there is clearly -- that is an offsetting trend.

The IDM customers were pretty healthy, pretty healthy in the first quarter.

And that is basically one of the most predictable segments that we have.

When we talk to our customers at the beginning of the year and go through their CapEx budgets, and CapEx needs then it -- then that is a segment of the business that is probably the most predictable to us.

So, yes, if you look at the year how it will pan out, first half dominated by memory shipments, second half will see a return of the foundries towards the end of the year.

We see the shipment of new technology the 1900 with an underlying base of our IDM customers making sure that there is a good healthy shipment pattern.

If we are just looking on an order basis, does that suggest that throughout the year we see a sequential slowing of orders from memory, while we see a sequential increase of orders from foundry, is that the correct assumption to make?

That is basically what I am saying.

And also what one of the previous people that asked questions brought forward is that, if it turns out that there is going to be a Flash shortage towards the end of the year, you could also see Flash customers coming back towards the end of the year.

And that's what Eric basically told also.

Regarding 450 --.

450mm yes.

450, we are I would say passive on this question, I would say positive, not positively passive, but passive, here to help as usual.

But there are multiple decisions to be made in the industry.

First of all is what's the cost of a fab, and it's not only lithography into this.

Second set of question is can we do the same thing with a faster machine?

Either it's a large size that you process, or is a smaller size 300mm that your process much faster.

And third set of question is what is the industry for 450mm?

Obviously I think I guess who this large logic manufacturer is, but if you ask the same question to a smaller guy you may get an answer that says the 450mm is already in my production for two years, and that would not work out as an efficient way of producing.

So this is an industry question.

ASML clearly thinks that they can do.

But we -- there is a lot of economics to be thought through before that were to be done.

And I even I forgot to tell you, and by the way if you do this and we introduce EUV, which is already complicated, so if you put this on a 450mm type base you are also adding difficulties.

So, long term I think it's an interesting concept.

We should always continue to document how to do this.

But I would not think that it has any short term [feasibility].

I can take it from that answer that you haven't committed to build or are spending significant amounts of R&D money on that project, is that correct?

In any event we would not before the industry would say so.

We are not -- we are going to be humble today.

We can't drive the whole industry here, so we will -- we are glad to participate to any industry scale project yes.

But at this moment we haven't yet decided to invest, because it's not -- there is no project starting.

Thank you.

Congratulations again.

Thank you.

The next question comes from Mr.

Doug Reid.

Please state your company name followed by your question.

Thomas Weisel Partners, and thank you.

You mentioned early in the call that much of your outlook for '07 is based on views you've obtained from third party data vendors.

I wonder if you could give just a little bit more color into the actual quality of your own visibility based on discussions with customers.

How has the tone of discussions with memory customers in particular changed over the past quarter?

And generally speaking how would you describe the quality of your visibility into memory?

Well this is an important question, because it hasn't come.

We all had the question whether memory Flash and DRAM would be impacted, in terms of CapEx plan, in view of the price pressure that these two segments are getting into.

What we see, and this is a fundamental message I give you, is we've seen a lot of discussions about scheduling and about mix and about switch from DRAM to Flash, which has been making our life a little bit difficult, and has stretched our capability for lead time and rescheduling and etc.

And this has happened in Q1.

This is happening in Q2 where we have this type of activity.

So, in other terms the volume is still there, the unit volume of DRAM and Flash are still sustained.

These strategic projects, as we discussed at the beginning of the call with Taiwan, which are still there.

But the price pressure obviously put responsibility on those manufacturers to do the right thing, and to be smart about which machine they want and when and etc.

So, honestly we are living through hell to give them satisfaction in terms of mix changes and etc.

But at the end, if we succeed to do that we still have the same volume, and this is what basically the message you got today about a difficult environment price wise -- sorry at the customer level.

But in terms of unit wise we are -- this is well sustained.

Thank you.

Next question is from Mr.

Mark Bachman.

Please state your company name followed by your question.

Pacific Crest Securities.

Peter, could you give us some more color on your guidance here, specifically with the 148 units in backlog, and your comments suggesting that 78% are shippable in the next two quarters.

This seems to me that you are going to ship roughly [152] of that backlog in both Q2 and Q3.

And then given the guide for the 69 units in Q2 this just leaves me with 46 units being shipped for backlog in Q3.

So my questions are, do you expect any additional shipments in Q3 from the orders received in Q2?

And then also can you describe how the shippable backlog in each of these quarters will change with regard to ASPs per units?

Yes, on your first question, yes we will book orders in Q2 that will be shippable in Q3.

That is a direct result of our continued push on lead time, [a reduction in] original average six months, which doesn't mean that our customers honor that, some of them are somewhat shorter, and we have some fights over that from time to time.

And clearly I think the order intake will also reflect the medium term ASP trend of the Company.

So you will see going forward, in 2008 there will be more immersion ramps.

There will be definitely -- we will be right in the middle of the Flash volume production ramp, which is using immersion.

You will see that DRAM will start to use immersion in volume ramp.

Foundries will be starting using it.

So, I think when you look at let's say the medium term order trend, this clearly has to go up as it follows the ASP trend of the Company.

And yes, also in Q2 we will be booking orders for all Q3.

Okay, and then just as a quick follow up here, can you talk about -- you said that normalized backlog should be around 50%.

Yes.

But why isn't the value of your backlog more weighted to memory, if these are the actual customers that are taking the leading edge tools with much higher ASPs?

Are you suggesting to us that logic and foundry put together are going to equal memory as we go forward in terms of immersion units?

If you talk about the total backlog value, total backlog value, it's about 50%.

So in units, it's of course less because you are right.

But when you look at the average selling price per segment, the highest average selling price is for the memory customers, followed by the IDMs and are followed by foundries.

So in that sense we don't make that split per segment for the immersion part of the backlog.

So my general answer would be the 50% that I talked about, so roughly 50% is based on the value, which is not in terms of units -- 50% is indeed lower, because yes the higher ASP tools go to the memory makers.

Okay thank you so much.

I think we are out of time, and I think -- I'd like to thank everybody for joining us on the call.

And operator if you'd be so kind as to close the call formally, I'd appreciate it.

Of course.

Ladies and gentlemen this concludes the ASML 2007 first quarter results conference call.

Thank you for participating, you may now disconnect.