艾司摩爾 (ASML) 2006 Q2 法說會逐字稿

Ladies and gentlemen, thank you for standing by and welcome to the ASML 2006 second quarter results conference call on July 19, 2006. [OPERATOR INSTRUCTIONS].

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

Go ahead please, sir.

Good afternoon, and good morning.

This is Tom McGuire, Head of Communications at ASML.

Welcome to our Investor and Media call and webcast.

The subject is ASML’s 2006 second quarter results, according to U.S. GAAP.

Here with me are Chief Executive Officer, Eric Meurice, and Chief Financial Officer, Peter Wennink.

ASML draws your attention to the message on today’s press release regarding forward-looking statements.

This call will be 60 minutes.

On our website, at asml.com, you can find our press release, a brief presentation and financial reporting with figures in U.S.

GAAP and in IFRS.

And now, ASML’s CEO Eric Meurice will speak.

Thank you, Tom.

Good afternoon and good morning, and thank you all for attending our conference call.

There are three main points which I would like to underline before answering your questions with Peter.

First of all, our capability to execute.

As you can see from our Q2 results, we have again executed very well on key measurements of performance -- growth, profitability and cash.

We increased sales by 50% compared to the previous quarter, and by 23% compared to a year ago.

We have been able to develop this growth by delivering to the needs of our traditional customers, but also through market share gains, in particular in technology development.

In terms of profitability, we delivered record operating profit at 25.2% of sales and record net profit at 17.7% of sales.

In terms of cash, we generated €330m from operations in Q2, while we used €253m for our share buyback program.

As of today, we have bought back 25.5m shares worth €400m, representing 5.25% of outstanding shares.

This completes the share buyback program which we announced in April.

So again, on the first point, we are satisfied to have been able to deliver solid and consistent results.

My second point relates to ASML’s technology leadership.

Q2 was an important quarter, confirming our market leading immersion technology as we brought eight XT:1700i immersion systems to market, bringing the total number of immersion systems shipped to date by ASML to 21.

While some of the XT:1700i systems installed during Q2 are already exposing wafers, customers are in parallel reporting continuous progress on immersion processing, and in particular reaching stable and low defect density.

This significant progress means that customers using ASML immersion technology will be able to begin volume production at the 45 or 55 nanometer node as early as Q4 this year, or Q1 2007.

To build on this lead, we also announced at SEMI West last week in San Francisco, the introduction of our TWINSCAN XT:1900i, the semiconductor industry’s most advanced lithography system.

We will start shipping in Q2 2007, and will enable volume production at 40 nanometer node, with the industry’s highest wafer throughput of 131 wafer [inaudible].

In view of our significant scale, and the favorable financial performance, ASML will continue to invest further in R&D, allowing us to accelerate the deliveries of leading edge technology, high throughput and low cost [ownership] systems.

To this end, we have decided to raise our R&D expenditure to €100m in the third quarter of 2006.

Finally, my third point will address the state of the semiconductor industry.

As we have articulated in the past, the industry is maturing, in particular as it is driven by the dynamics of five different demand segments, namely DRAM, Flash, IDM, Foundry and the R&D segment.

These five different demand segments each target multiple and product segments, and therefore work on different individual cycles, which results, on average in shorter and milder overall global industrial cycles.

At this point we see a build-up of Flash capacity, a relatively strong DRAM market, and a sustained IDM and Foundry business.

At the same time, the investment in [virgin] infrastructure boosts R&D equipment sales.

The different segment dynamics, just to summarize, are reflected in our Q2 billings and our Q2 bookings of 93 systems, as well as in Q3 bookings guidance, which is above 62 units.

Together, this confirms the strong outlook sustained by three drivers, three growth drivers.

First, the build-up of litho capacity resulting from the shipments of litho machines and future shipments, which are in the bookings, will cover adequately the semiconductor growth projected by most industry analysts, which is currently at about 10% in value, and 15% in units year to year, year 2005 versus year 2006.

The second driver is the targeted build-up for Flash memory, which as you know grows much faster than the average market, and requires significant capacity in place for the booming 2007 demand.

And the third driver is ASML’s continuous market share growth, particularly in the Taiwanese DRAM industry and in Japan, but also through a successful ramp of the XT:400 i-line non-critical layer product, and of course in the immersion segment.

So in summary, strong execution, strengthened technology leadership and a sustained growth, thanks to a significant and multiple source, market share gain.

Now, Peter and I will be able to take your questions.

Thank you.

When asking a question, please identify yourself and your organization.

Please ask one question and limit to a short, short follow-on question.

Can we have the first question please?

[OPERATOR INSTRUCTIONS].

The first question is from Mr. Sheera.

Please state your company name, followed by your question, sir.

Thank you very much.

Good afternoon, gentlemen.

It’s Nav Sheera calling from Lehman Brothers.

I’ve just got a question just with regard to your lead times, because for a long time you have been focusing on reduction of your lead time program.

I just wanted to have any further details with regard to that, especially as the majority of your shipments are now TWINSCAN-based products?

Thank you.

Yes, Nav, lead time, as you know [inaudible].

We have now in place a shift program which allows us to work seven days a week, 24 hours a day, and we are working on the specific cycle [inaudible] time reduction in manufacturing.

That has allowed us to shrink our average lead time from nine months to six months.

And at this moment, I would say, standard lead time of six months for production of mature machines.

Thank you for that, and just a quick follow-up, in terms of [terms] opportunity, can you give us the order of magnitude or number of units that you believe you could have as terms business for the fourth quarter, which isn’t on your backlog as of today?

Thanks.

Terms business, as you know, and you’ve seen in Q4, Q1 and also in Q2, we had terms business a factor of five to ten per quarter.

The way to do terms, basically, which is getting an order and shipment at the same time in the same quarter, the way to do this, knowing that you have a lead time on average of six months, it means that you have already prepared certain numbers of modules that then we will rush into as final assembly later in the quarter.

So the ability to do terms is in fact the ability for us to forecast the appropriate sub-assemblies and, of course, having some customer demand.

Today our performance has been over the past nine months or year, the performance has been very good.

As you can see, nearly 10% of our business, in fact, was made on terms.

Thank you very much.

The next question is from Mr. Gaudois.

Please state your company name followed by your question, sir.

Yes, hi, Nicolas Gaudois of Deutsche Bank.

First question, Eric, on operating margins at 25.2%, from my recollection is the highest level for ASML post [SVG], and I’d have to go back to the 90s or 2000 to actually read that.

Would you quantify this is probably as much as a business model can generate in the current cycle?

Or should we actually revisit the business model going forward?

I will leave that one -- that difficult question to Peter.

Thanks.

Thanks for the difficult one.

The issue is that we have said in the past also what we want to do is leverage our operating costing base, which is R&D and SG&A.

As you have seen, we have kept SG&A flat in absolute terms, and yes, we are spending more on R&D as the percentage of sales is going down.

Also we have said that we are targeting at a 40% gross margin, but with the control of the operating expenses we’re currently seeing, basically leveraging our operating expense base to generate more sales.

That will have a positive effect on the operating margin as you have seen.

So the increase in top-line will lead at 40% to an absolute higher gross margin, given the statements I’ve just made on containment of the operating expenses.

That means that on the assumption that we can grow the top line, there is still room to improve on the operating margin line.

Okay, great, but 25%, Peter, for full, let’s say, tick-over cycle here is a relatively deep level, do you think or?

I’m really talking about the medium-term here of course, not --

It depends on the sales, Nicolas.

It depends on the sales.

If we could have done more sales this quarter, we would have had a higher operating margin.

Okay, fair enough.

And mostly a strategy question, you just hinted, Eric, I think, that you’re accelerating the release of a new platform for, I’m assuming, either double exposure or EUV, and working on making this a bit more affordable?

You just leapfrog, effectively, Nikon with the 1900i.

Is that then effectively to accelerate the rollout of a platform replacing the TWINSCAN altogether?

Certainly.

We think the industry is evolving in two directions.

One is an acceleration of the transitions to new technologies, so we expect, in fact, the rate of change to evolve from an 18-month transition period node-to-node to probably 12-month.

And then the second dimension is that we expect the multiple segments of this industry -- DRAM, Flash, IDM, Foundry -- to have different needs for products, in particular, when you overlay on these different segments’ needs with the new technologies that are being now developed like spacer technology in [inaudible], vertical gate transistors, three dimensional transistors, etc, etc.

So these things will require multiple different architectures to address most effectively each of those new ways of designing chips.

Therefore, putting more money in R&D will allow us to accelerate those platforms and to multiply the numbers of platforms to address most effectively each of those, I would say, sub-technologies.

So yes, this is for us an opportunity because only a very large investment allows a success in this strategy.

And as you know, at this moment, we are the ones able to invest the most in R&D.

Thank you very much.

The next question is from Mr. Brenier.

Please state your company name, followed by your question.

Yes, good afternoon.

It’s Thomas Brenier from Societe Generale.

I’d like to have a bit more detail on the immersion tools that you’ve shipped over the quarter, if you just could give us an indication of the average selling price of these tools?

I’d also like to know if you are still preparing for more than 20, 25 shipments this year?

And if you have a bit more idea than two months ago on the number of immersion units that you would be able to ship next year?

Okay, this is a tough question.

The first question is giving the average price of the current 1700 is about €25m.

The list price is €29.5m.

The forecast for this year is still between 20 and 25 units, so that we are fairly consistent.

We’ve announced this, I think, six months ago.

The development of the customers are going, I would say, ahead of plan, or very well.

We have, or you may have heard already, some customers publicly talking about better yield in immersion processes than in dry processing, and that is a very important milestone to get the whole industry to jump into volume production.

So yes, we are confirming those ramps of volume production to start in Q4 ’06 or Q1 ’07.

At this moment, we have clearly identified eight customers who will be ramping volume, and as I said last time, there is another seven, basically, to total the 15 which are still working on ramping some scaled immersion processing in 2007.

How many tools is a bit too early to say, but we are very optimistic, or comfortable, with our market share.

You think that this number will be higher than in 2006?

Immersion --

In ’07, I would say.

Again, we don’t want to give too much guidance on this, but obviously 2006 is an R&D year. 2007 is a production year.

Okay, fair enough.

Just a small one on your CapEx, which I think was quite small over the first half.

I just wanted to understand if there’s any change in your thinking in terms of the CapEx suggested for this year, and maybe your normalized CapEx projected for the years to come?

I will leave it to Peter.

Yes, CapEx this year will be around €90m, €90 to €95m, which has been around that level for the last few years.

We don’t anticipate to change that level in 2006.

Now, we’re constantly evaluating our capacity, the production capacity.

And depending on our forecast of future needs, we might decide, or might not decide, to have some additional capacity, which would mean some extra square meters for the factory here in Veldhoven.

Also, with the introduction of newer systems, which are more expensive systems, our CapEx debt will also consist of our own tools, which we’ll use for demo purposes and for prototyping.

That will increase when the average selling price, or the average cost of those tools goes up going forward.

So going forward over the next couple of years, we could see an increase of that €90m to €95m within 10% to 20%, but that does not include any, I’d say, major square meter capacity additions that we could add to the factory if we would need it.

Thanks a lot, Peter.

The next question is from Mr. Deahna.

Please state your company name, followed by your question.

Mr. Deahna, are you there?

Yes, can you hear me?

Yes.

Okay.

Nice quarter, guys.

It’s Jay Deahna from JP Morgan.

Just a couple of questions.

First of all, it looks like you’re guiding orders to be down about 10 units or so in 3Q, but given the mix shift to immersion and whatnot, is there a probability that orders will be up sequentially in euros?

And then the second question is, it looks like you’re guiding for revenues to be down in 3Q.

Is that correct?

And is that seasonal or is that due to some sort of push-out activity?

Regarding the bookings, we are always very cautious as to forecasting a mix.

As you know, we are pretty good at calling the number itself, because we are working on certain numbers of projects.

And yes, we are working on more projects than 62, but it always takes a bit of time to negotiate and finish, but you never know which one will be easier than the others.

Clearly, we expect Q3 to be the first quarter where we will book the 1900, and the 1900 is a high price, of course, higher than the 1700, already at €25m.

So there is a fair probability that the ASP of these Q3 bookings will be higher than the average ASP of Q2 bookings.

The guidance that we have taken for Q3 is a guidance, again which we could exceed in case terms were to happen, but again, we have not included this into the forecast because these are things, which as you may imagine, are pretty difficult to call.

This, however, whole picture, is still a picture of a strong Q2, Q3 and you can imagine, Q4 trend, if you look at these current, I mean the backlog that we have, which is 88% shippable within the next six months.

Was my question on revenue correct?

Are you guiding -- you’re guiding revenue-- It looks like, based on the slides in your presentation, you’re giving ASPs and you’re giving a unit outlook for shipments.

If you add that up, it looks like revenues are down sequentially in 3Q.

It sounds like you’re saying that you could upside that if you get some terms.

But generally speaking, given that your backlog is up coming out of Q2 and your revenue guidance looks like it’s slowed down a little bit, is that just conservative or is that impacted by summer seasonality, and possibly some push-outs?

Have you seen any push-outs?

Yes, I think you have a point.

We do not -- I didn’t catch exactly your – the nature of your questions.

We expect, in fact, Q2 to be on the same strong trend -- sorry, Q3 to be on the same strong trend as Q2.

The seasonality is just the production schedule during the famous European tradition times which makes it a bit more difficult for us to plan more units.

So clearly, zero push-outs from customers, and for us, it’s a very solid and booked quarter.

So it is in the same vein as Q2 and for us, a major growth compared to last year, and showing sustained business for the year.

Just a quick follow-on.

In light of the fact that you’re seeing no push-outs, what’s your general feeling on this cycle?

Is this a sustainable cycle that will drive higher CapEx next year?

And given your immersion ramp, can ASML grow its revenues in a flat CapEx environment next year?

No, this is a very important question.

This is the 2007 question, how do we see this evolving?

Clearly, we usually do not [course] so long because we are trying to give you the factual data on the future.

And as you have seen, our backlog is the most forward looking data that we have, which basically covers until the end of the year.

However, what we can quote you as factual data to understand a bit better 2007 is as follows.

First, yes, we understand that some analysts see 2007 as a stable year, in terms of capital equipment.

And I understand that most analysts would plan the 5% unit growth of semiconductors, but again, these are not our numbers.

This is what we hear from the industry.

On those numbers, we feel optimistic going into 2007, because of our significant market share increase, which as you have seen in the press release, is multi-source.

There is a market share increase in the customers that we have already acquired some time ago, but which now ramp production and volume.

We have a market share increase in Japan.

We have a market share increase in the lower end of our business, in i-line.

And we have serious market share increase in the high end of the business, which is immersion.

You add to this fairly large optimism about the introduction of the 1900i, a new machine which will probably also increase the value of this R&D segment as customers will want to develop their 40 nanometer load, and you add to this on a picture where we close 2006 with a good growth, but where we think the customers will have a good utilization rate from our current simulation.

So yes, we feel optimistic that 2007 could be a very good year for ASML, in view of this market share strengthening and the technology leadership strengthening.

Does that answer your question, sir?

Next question, Operator, please.

The next question is from Mr. Menzies.

Please state your company name, followed by your question.

Good afternoon, gentlemen.

This is Titus Menzies from Jefferies & Company.

Can I ask you just two quick questions?

Firstly, on the gross margin guidance, if we’re modeling in a capacity demand for i-line tools, XT:400 tool, end third quarter, shouldn’t Q3 gross margin be slightly higher than Q2, assuming that these machines have a better gross margin and no price cuts in the quarter?

Peter?

Yes, I think there is no major difference between the gross margins on the different tools.

We have worked very hard on the cost of goods reductions over the past few years, and has led to the situation that there is not a great deal of variance between the different tools that we are selling.

So additional capacity income -- sorry, incremental shipments -- of any kind will not have a major impact on the gross margin guidance.

Okay, thank you.

Secondly, there was a significant increase in the number of employees in the quarter.

Could you just give some guidance of what regional or what part of business were they allocated towards?

Yes, that was two-fold.

Service, because we are gaining market share, and we’re having new install bases at new places around the globe.

And also to support the shift pattern in the factory that we have started with in February of this year.

We need additional people to do the specific work in the shift patterns, so that means some additional people also.

Okay, so it’s in one point in the production side and for services as well?

Yes.

And should remodeling for the services more closely to the €100m mark which we had in the quarter?

Or is that where we should be looking at going forward?

No, that’s pretty lumpy.

We’ve always guided on that, and I don’t want to change that.

Our contract business is around €50m a quarter, and our options business, our field option business, can be between €20m and €30m a quarter.

It was a bit higher than €30m in the second quarter.

And also our contract business, we have some specific assignments from customers that led us to bill additionally in Q2.

So the €75m to €80m a quarter is, you could say, average, which could peak to levels that we have seen in Q2.

Thank you very much.

You’re welcome.

The next question is from Mr. Arcuri.

Please state your company name, followed by your question.

Hi, it’s Citigroup.

Two things.

First of all, do you guys think that it’s a valid analysis to look at 200 millimeter equivalent systems that are booked by you relative to the number of chip units being sold?

Because if you do that, we’re right back up to where we’ve been at a peak the last five years, five or seven years here.

So do you think that that’s a valid analysis?

I think for this one, you may want to have discussions with the IR team about simulating different scenarios, but it’s very difficult to take what happened five years ago and try to replicate it on now.

In fact, I must say, I may not have understood exactly your reasoning, but 200 millimeter for us at this moment now is a kind of legacy fab, and they have their [last] on their own.

You have a lot of customers trying to pull them in and remove them into China, etc, etc.

You’ve got better productivity and when you do this then you need some additional 200 millimeter type machines to complement, basically, and everybody is moving on 300.

So again, I’m a bit unsure, and I don’t think we want to take the time but I’m sure we’ll be interested to discuss the new model with you, because that helps us to understand future trends.

Okay, and then I guess last thing quickly, in your comments you seemed to have a little bit more, maybe, cautious tone on memory than what we’ve heard from other vendors, and I’m wondering, was that a misread?

You sounded like maybe some of the memory makers are adding capacity at a rate that might not be sustainable, in your view.

Is that the wrong read?

No, in fact I said -- well, yes, it is the wrong read.

Last quarter I guess I was conservative in saying that we expect the DRAM business to be strong and the numbers of tools that we ship to the DRAM business is appropriate, and parallel to that.

And I said that the Flash business is clearly booming as a lot of new entrants are attracted by this significant growth of this [business].

I guess this year, the Flash business is probably going to grow by 180 or 200%.

So these are significant segments, attracting a lot of new entrants, and new players.

And any time you have new entrants and new players, you will have some duplicated capacity, I would say, which first we judge to be reasonable.

For us, it represents about, say, 10% of our total business, and second, that we consider as not so dangerous because even if there is a bit of over-capacity on Flash, Flash is growing so fast that we’ll be caught very quickly.

So in other terms for us, this little over-heating of Flash is 10% of our business which is reasonable into the whole picture, and as I say, that may trigger, some time in ’07, a reduction of run-rate of Flash players, but not before time.

So are you saying that Flash orders are 10% of total, or that the over-ordering is 10%?

The over-ordering is -- well, when we say over-ordering, the ordering which is ahead of the curve.

Great, thanks a lot.

The next question is from Mr. Schafer.

Please state your company name, followed by your question.

Hi, Simon Schafer from Goldman Sachs.

I was wondering whether you could give an indication on the mix as far as order intake is concerned in the second half?

It seems like the incremental upside this quarter is very much driven by some of your Asian customers, specifically memory, but also a mixture much more towards capacity buyers.

Do you expect that to continue, and what should we deduct in terms of ASP progression on your order intake?

Yes, as I try all the time to avoid giving guidance on order intake, average ASP because the timing of receiving orders is very granulous from time cap in September/October and the difference of price between an i-line system at €35m and 1900 at more than €25m is going to make my guess difficult.

Regarding your first question, are the bookings that we’ve got, the famous 93 units, in fact is fairly spread.

Flash, DRAM, the foundries, IDMs, difficult to give you a feeling of which segment would be a bit on the weak side of its seasonality, and which segment would be on the high side.

They are all sustained, which is why you have seen our outlook is strong and our sales will be strong for the next six months or so.

Maybe I can ask in a different way.

Would the run-rate of order intake for immersion continue at the same pace, and would that include 1900i’s in the second half?

The immersion order entry will accelerate.

This will start partly in Q3 and Q4 order entry for immersion will be probably the highest.

Thank you, and then just a follow-up on gross margin, I was a little surprised to see the 41 number creep into your press release.

Historically you’ve sometimes mentioned that maybe at 40% you were maybe capped in order to maybe capture some share.

Does this higher number, or this higher range signal that you’re becoming a little bit less aggressive on trying to capture more share?

Well, Simon, on that question, gross margin, what that actually indicates is that we’re pretty confident about the volume of our business, because the volume gives more leverage on the fixed cost in the factory, and that has a positive effect on the gross margin, and that’s what you see reflected.

Okay, thanks very much.

The next question is from Mr. Menon.

Please state your company name, followed by your question.

Yes, it’s Janardan Menon from Dresdner Kleinwort.

Just a clarification on the previous answer, what is your total percentage of NAND Flash in the backlog?

You’re saying 49% is memory.

How much would be NAND Flash?

I think it’s between 25% and 30%.

25% and 35%?

So --

It is a bit more than half.

It is a bit more than half of the memory.

Okay.

So but then you seem to be suggesting that if there is any over-heating or anything like that, it is in NAND Flash, because you’re saying there could be a 10% of orders are more than they should be, or to that extent, but given that your NAND Flash orders are roughly the same as Foundry, IDM, etc, I’m just wondering why you think that there is too much of NAND Flash and not too much of Foundry or IDM or DRAM?

Well, we have a fairly sophisticated simulation model which models how many tools will be necessary per segment, so the NAND Flash segment grows about 80% or 100% in units, and so we’re trying to model the numbers as of the square meters of silicon, etc, etc.

And of course, you are talking about a high growth with a much smaller business.

So the numbers of tools required in DRAM to do the smaller growth, and the numbers are still required in Flash to do a much higher percentage growth, but sometimes is about the same, so I don’t have the model in front of me, but I can tell you we’re looking at, at the end, the square meters of silicon which are being built on the different technologies.

And just a small follow-up, and this is on your ARF ASP, if I just use your presentation, I think it’s on Slide 12, to calculate the ARF ASP on your shipments, both Q1 and Q2 after backing out the services revenue, I find that actually, it’s gone down a little bit, which is, it’s gone down from about 19.7m to 19.5m.

Given that you shipped eight 1700i’s in the quarter, which you didn’t do in Q1, I was just wondering why the ASP is flat, effectively, it’s slightly down, and not rising in the quarter?

How can you come up with this -- how do you calculate this?

Can you read for me the page you are looking at?

Okay.

Oh, I’m sorry, you’re doing it.

Sorry, it’s on Page 10 of your presentation.

Well, in the -- when you compare to Q1?

Q1 and Q2.

On page 10 of the presentation you’ve given the breakup by revenue and the number of units.

We’re struggling to try to reconcile your number because you’re right to say that in addition in Q2 we had those 1700, but a lot of 1400.

So we will have to come back to you.

There is no reason why the price should go down and -- there could be --? [inaudible].

So we need to go back to you.

But one, I can assure you of a certain number of things.

First, the ASP has not gone down per machine.

And two, you do have the impact of the 1700, which is a higher price than the 1400.

So you may have either a mix of 200mm, 300mm in there, or -- and you may have, I wouldn’t be surprised, a mix effect of customers.

You may have customers buying high volume and customers buying one at a time.

So you would have the ones in high volume calling for higher discounts than the one with low volume, which could explain the very slight difference that you have seen.

But again, [Frankie] would be happy to follow up with you and give you the data.

There’s nothing like a trend that you can see there.

Okay.

Thank you very much.

The next question is from Mr. Adrian.

Please state your company name followed by your question, sir.

Yes, hi there.

It’s Stuart from Morgan Stanley.

So, a question on the R&D.

Last couple of quarters I think it’s been edging up and you’ve provided a good explanation of the projects that resource is being put into.

Do you think R&D, net at €100m a quarter, is pretty much where it’s going to top out for the time being?

And secondly, could you just talk a bit about any flexibility you’ve got within R&D, if there is a slow down at some point out there, in terms of managing that cost base?

Yes.

So to your first two questions.

Clearly we invest R&D funds to do the right things.

And today we think that €100m is a result of what we can ramp, because it’s important to find people, and [giving us] with the right skill set and the numbers of projects we have.

And you have noticed at the beginning of the call, I hinted about how many new architectures we are now ramping in parallel and accelerating.

So, at this moment, yes, indeed.

I could see the need for more money, in view of the opportunities which we have in front of us.

So there is potential to achieve more high speed and to achieve more penetration in some segments if we put a bit more money.

So yes, there is going to be a part of our brains looking at investing more.

The part of our brain which looks at investing less is the one that says we need to anywhere deliver to our profitability target, no matter what.

So if the cycle is positive, if our performance allows it, you will see us moving north of 100m, even in a very short timeframe.

At the moment, or if the cycle weakens, we will take a decision, depending on the length of the cycle, to scale down.

All our OpEx -- sorry, most of our OpEx -- are significantly scalable down by a factor of 20 to 30%.

This is done by the smart usage of subcontracted work on subcontracted workers.

So we have switches everywhere and, in particular in R&D, which allows a quick change of run rate of R&D.

But I repeat, this would be based on our reading of the cycle, as the cycle is very long down, in which case we may trigger, and if the cycle is only a quarter of etcetera, we may just plead for your patience and keep the R&D afloat for that cycle to continue because it is going to create value -- so much value in the years to come.

Okay.

Understood.

And just the follow-up is just back to gross margins, I’m afraid.

As we think about immersion becoming a more mature technology and moving out of the R&D into a volume product for you, I’m still quite surprised that you’re essentially saying the gross margins you’re going to earn on those products are essentially just in line with Group gross margins, given that we know immersion is at least double the ASP.

Is that going to be the case as we move through 1700 into 1900, and the 1900 becomes a true volume tool for you, that you don’t expect any upside pressure on the gross margins?

Yes, I agree with you that Peter did not answer your question very well last time.

So we are going to ask him again.

Peter, could you please, this time, answer the question?

I think your question, Stuart, specifically, just to be sure you are asking whether there is an upside opportunity on the 1900 because it’s going to be more mature, or immersion in general?

Well, I think just generally as immersion becomes a volume production technology.

Normally, I think, when you’ve introduced new tools we’ve gone through this learning curve type of gross margin.

But actually, I’m surprised that we’re not going to see any of that that would maybe provide upside risk to your 40% gross margin over the next year or two.

Yes, so it’s upside opportunity you mean, or upside risk.

I was stood on the wrong leg, you were talking about risk.

I thought you were talking about the downside effect.

I think what the difference between the gross margin -- sorry, the gross margin is determined as the difference between sales price and the cost of goods.

The sales price is really driven by the value that we’re giving to our customers.

So our customers actually value our tools quite well by looking at what those tools deliver in terms of the shrink and in terms of productivity, thereby determining the [laboratory] cost per layer.

And that’s determining the value of the two [manufacturers].

There’s not much more we can do than follow our customers in that particular sense.

So how do we control the gross margin and driving costs down?

Now, are there opportunities for driving cost down further?

There always are.

But with every new node, and also we have seen it with the 1900, the optical cost will go up because the complexity of the optical system goes up.

The series are not necessarily being larger.

So that means that R&D costs need to be earned back over lower volumes.

And that paints a pretty complex picture whereby it is our task to just keep the cost of goods at 60% of the sales price.

Now, we’ve been able to do that.

And given the complexity of what we’re doing, it’s a bit difficult to predict whether that complexity will be going down going forward so that we can decrease the 60% and increase the gross margin.

So currently what we are seeing is to keep that balance that we have seen over the last few years, of a 40% target, to keep the balance also going forward, largely given the uncertainties with respect to the cost of goods development.

Okay.

Understood.

Thanks.

But I would be a bit more optimistic.

Peter is the CFO, I am the CEO, so we are working very hard on the cost of goods.

One part of the 100m is in this.

So -- you should -- you could be optimistic, yes.

Understood.

The next question is from Robert Maire.

Please state your company name, sir, followed by your question.

Needham and Company.

Just some clarification on the future guidance, and a follow-up question on double exposure.

You’re guiding above 62 systems.

I’m assuming that’s new systems forward guidance for bookings, as compared to 78 that were booked in the current quarter, or is that versus 93?

It is total systems.

We are comparing total systems to total systems. 93 total. 62 Total.

Above 62 total.

Okay.

That’s obviously a significant drop.

And I realize you’re saying above 62.

So there’s room for --

No, again -- sorry, we had the conversation in Q1 where we said we think a strong business, and you can make the calculation with the ASP, etc., etc., is around 60 to 70 per quarter.

And the bookings -- sorry, billings.

And the bookings may happen at different quarters because of assumptions of our customers on lead time and assumptions, or risk, on slotting.

So you will see the bookings curve having some life of its own which does not reflect so well the need for customer -- of tools.

If I make myself clear, in Q4 we have bookings going well into nine-month lead time.

In Q1 we had bookings which, as well, about six months’ lead time.

In Q2, we are now getting back to nine months.

So you see an accordion happening.

So what we called for Q3, at 62, is a -- well, if I ask, probably [inaudible], but it is compatible to a nicely growing billings business.

And, by the way, I said above 62.

I didn’t say 62.

So it’s, again, some discussions you can have with the IR team to try to understand how the booking trends will translate into customer billings requirements or delivery requirements.

If we’re trying to derive out of that what the customer’s looking at, if we’re talking about a nine-month lead time, that would assume that most of the order -- the 93 orders in the current quarter are aimed at first quarter ’07 delivery.

Is that correct?

No.

Very roughly, the current backlog we have of 1.7b, we said -- €1.7b, we say is 88% shippable in the second half.

That’s the first piece of data you should have to try to understand 2006 and the customer needs because, at the end, bookings is not customer need.

Billings is customer need.

So then you have, in these six months, you already have [terms].

And I think I answered one of the questions on terms, and I said, as we were having before, 10% terms on our regular invoice.

And obviously related to that, the revenue for bookings isn’t going to be down as much since the ASP should be up substantially, given the --

Exactly.

You’ve raised another question which, if you get a lot of immersion in there, you would get a jump of ASP.

And the reason why we guided you on 62 bookings plus is if we get more i-Line, it’s in fact going to be more than 62.

And if we don’t get more i-Lines and we get all immersion instead, it’s going to be 62, but the ASP will be higher.

But I repeat, very complicated for you to call a business trend on the bookings.

There’s so much discussions on the lead times that the bookings trend is different.

The transition we make is we are today seeing growing sales.

That’s what we are trying to hint through these different numbers.

And if you make the math, you will come up with those numbers.

Okay.

Just related to that question, we touched on it before, double exposure.

It seems over the last maybe one or two quarters in the very recent past that double exposure had become a very real probability in the industry.

Is any of the order activity that you’ve seen in the near term related to customers perhaps being more willing to order tools based on this new-found methodology for using non-wet or non-immersion tools, sort of a bridge to get to immersion?

And related to that, I’m assuming you don’t have to significantly alter your tools to do double exposure.

It’s really a relatively simplistic thing, as I understand it.

Well, it’s not so simple, in fact, because it makes your design pretty complicated, especially if you do logic.

And second point is the overlay, which is the capability of the machine to replicate the other edge of the design on top of the one you have already exposed.

It’s fairly complicated.

So no, it’s not a no-brainer.

However, double patterning is going to exist.

It’s a good breakthrough of the industry.

And it will help a lot of customers go to the next step node.

Double patterning will exist in different forms.

It will exist, in fact, at different nodes and it will exist at different technology.

Some people will use double patterning dry.

Some people will use double patterning wet.

Some people will use double patterning for -- until there is an economic solution in the [UV].

And some people will have to go in the UV, even if the UV is more expensive for a period of time because they can’t duplicate their design.

So you are going to have, from 2007 or 2008, on a mixture of approach to [some of] nodes.

But very clearly, if you do have double patterning dry today, and imagine you can do for the niche 45 nano with double patterning dry, guess what, you can do double patterning immersion giving you 40 nano or less.

So dry -- immersion will always be there.

Double patterning does not replace immersion.

Immersion is also an enabler to double patterning.

I understand that.

The simple question is that the emergence of double patterning or double exposure enhanced the sales or the ordering in the near term of dry systems which may have helped the orders that we saw in the current quarter?

Today it looks clear to us that immersion will be cheaper to do at 45 nano than double patterning dry.

But that doesn’t mean some customers may not want to have one layer or two done with double patterning because they have an installed [path].

[One weight].

[inaudible], sorry, immersion is done.

Everybody is now ramping production with the orders, etc.

So you won’t get anybody in double patterning production faster than somebody in immersion.

So you really have the choice now.

And it’s really how much experience you have in both and how many wafers you really want to process.

If you want to process a lot of wafers, etc., immersion is much cheaper.

There’s no question there.

But if there is not so many wafers, you’ve got to do it in the first half of ’07, it’s going to be some full structure of design, then you can obviously use double patterning dry.

Thank you.

That helps quite a bit.

Thank you.

Operator?

Yes, sir.

We have time for one more question.

One final question please.

Okay.

The final question is from Mr. Hosseini.

Please state your company name followed by your question.

Sure.

Mehdi Hosseini, Freidman Billings.

I have two questions.

First, if I heard it correctly, you expect one unit booking for 1900 in Q3.

Can you tell me what type of customer you expect to do this booking?

And then regarding your other comments regarding share gain in Ireland, is that in dry ArF or that particular customer has decided to include immersion at 55 nanometer?

Okay.

So I think I confused you.

I don’t think I said one 1900 in Q3.

I said we will start the bookings of 1900i.

So we do --

So it could be more than one?

Yes.

Yes.

Absolutely.

In fact, at this very moment, every customer of ours is negotiating his special specifications and, of course, condition of sales and schedule.

So, again, we don’t exactly know which timeframe we are going to get all these results, but this is -- the decision has started.

Regarding the i-Line market share that is the non-critical layer technology, what was your question exactly?

Okay.

Just on a 1900i, what type of customer would you expect to start ordering?

Would that be NAND or DRAM?

And then your comment regarding share gain in Ireland, I assume that’s for 55 nanometer.

And my question was is that for dry ArF or immersion.

And if it is immersion, is that going to be used in combination of double patterning?

Okay.

I think you guessed I said Ireland, the country, that means [intel].

I would never do this.

No, I said i-Line.

I’m sorry.

Sorry.

It’s my fault.

And the reason why we are successful in there is because our i-Line is non-critical layer, but with a very, very good overlay.

So lots of success there.

Regarding your questions on the 1900i, very important question.

The 1900i is using the 1700i platform.

So it’s a fairly easy switch from 1700 to 1900.

So that means 1900i will serve very, very quickly in production.

So lots of customers doing 1700i will move with minimum R&D to a 1900 production.

So you’re going to have a whole bunch of customers like this, and these are the [facts].

And the other part of the customers, like mainly logic and Foundry, will use the 1900i to develop 40 nanometer technology.

And they will take R&D tools and they will take a bit more time to do production.

Sure.

If I may go back to my second question, you talked about share gain in Japan and Taiwan.

How come you’re not bringing up the U.S.?

U.S. are -- I don’t even know where to gain market share.

It’s one large customer.

Yes, the big microprocessor customer.

[Inaudible].

Yes, we, as you know, we do not comment on single customers.

It’s very difficult for me to tell you.

But we feel optimistic obviously.

That concludes the Q&A session.

Thank you for attending.

See you next time.

Goodbye everybody.

Thank you.

Ladies and gentlemen, this concludes the ASML 2006 second quarter conference call.

Thank you for participating.

You may now disconnect.