艾司摩爾 (ASML) 2004 Q2 法說會逐字稿

Good morning and welcome to the ASML conference call. At this moment all participants are in listen-only mode. Later we will conduct a question and answer session. I would like now to hand over the conference to Frank D'Hoore for the introduction. Mr. D'Hoore go ahead please.

Thank you very much. This is Franki D'Hoore from ASML Investor Relations. First of all, thank you all for participating in the ASML second quarter results conference call.

Joining us today are Doug Dunn our Chief Executive Officer and Peter Wennink our Chief Financial Officer. This call will be one hour, and we will start with one question per person in the first round to give everybody an opportunity. When there is time left at the end, we may take your follow-on questions.

But, before we start, please allow me to read the Safe Harbor statement and it says that the matters discussed during this call include forward-looking statements that are subject to risk and uncertainties, including, but not limited to economic conditions, product and pricing, manufacturing efficiencies, new product developments, ability to enforce patents, availability of raw materials and critical manufacturing equipment, trade environment and other risks indicated in the filings with the US Securities and Exchange Commission.

At this point I'd like to hand over to Doug for a brief statement, followed by your questions.

Okay. Thank you Franki for reading that exciting Safe Harbor statement. I'm not going to give a long introduction, because I guess by now you've had a chance to listen to our webcast, see the presentation attached to it, read the press report we issued and the comments based upon that. And, possibly some of you even called in this morning to the press conference.

So, I just want to make a couple of comments and then throw it open to your questions.

I think the numbers themselves from our second quarter speak for themselves. It's a tribute again to my management team and the whole of ASML that we could produce such a comprehensively solid set of numbers out of Q2. And you may ask questions on those numbers as the call goes by.

I'd like to, however, comment more on the future rather than the recent past. We have built, as you can see, a very heavy backlog of orders during Q2. Our backlog increased from €1.35b to €1.8b, that's over a 30% increase in the quarter.

It's a very high quality backlog, quality in the sense of its -- a lot of its 300mm leading edge products. And, it's from a range of customers, all the -- all the guys you can think of I guess have got at least 1 order in there and some 10 and 12. So, it's a quality backlog in every sense.

It's quality also in that the prices we'll be getting for those products is increasing. You can see the average selling price for that backlog has increased quite substantially, as we forecast it would in January and also in April, as the mix swung from the early easy capacity of 200mm into the more kind of thoughtful capacity of 300mm. That swing now, as we suggested would happen in mid year is happening, and that shows in the value of the backlog, as the 300mm devices are more expensive, and also in the gross margin from that backlog.

We also recommended to you our gross margin will continue to increase through the year. That is again not news to anyone that's called in before. We've been on target now I think for 4 quarters on our gross margin ability to predict and I see no reason why we shouldn't be just as accurate for the next 2 quarters out, minimum.

Beyond 2004, which will be in itself be a very good year, we look forward to 2005, and I say look forward to it with pleasure. Every indication we have so far, it is going to be a pretty bumper year. The discussions with customers to fill out our order book in the first half of 2005 are proceeding well. A lot of interest; more interest in the first half of 2005 than the first half of 2004, that's for sure at this early date. Therefore, all the kind of piece of evidence that recommend and suggest how next year will be, look very positive.

So, just on that issue of the future, there's a lot of speculation around about when this will peak. I can say categorically from the facts that we have at our disposal today, today on July 14, there's not one fact at our disposal that indicates the peak is here, or even imminent. Therefore, we can say with the degree of confidence anyone can say about next year, it looks a very, very strong year.

And, in fact in our press release we implied, in fact we said, that the upturn will continue well into next year. That doesn't preclude it continuing all through next year by the way. We try to be factual here. We have visibility certainly in two or three quarters next year.

So, I want to give that kind of upfront information, maybe it helps you ask your questions. And, with that I will turn the floor over to the moderator and the questions. Just one little rule please, a lot of people on the line, about 150 people, therefore it is initially one question each. When we've exhausted those we'll come back for a second go round for second questions. So, moderator perhaps you can begin the question process.

And please introduce yourselves with your name and company you represent.

As said, ladies and gentlemen we'll start with the question and answer session now. [Operator Instructions] The first question is from Mr. Pelzer go ahead please.

Hello it's Ulrich Pelzer, Lehman Brothers. Just to clarify, the SG&A and R&D guidance that you've given. Does that extend to the end of 04 or to the end of the first half of 05? And, I am assuming that it's for the business as it stands today and it does not cover potential entry into other businesses.

So, it's to the end of 04 and there are no conditions on that. And, as regarding 05, just to pick up part of your question there, then providing 05 looks to be as good as we think it's going to be, we expect in 05 that those costs will inflate. I mean there is inflation and a lot of them are people related. Therefore, if you allow the inflationary effect, -- I think there'll be an impact of inflation in the first half of 05, we'll try and contain that, but it will be there.

And, also one other point that may help you understand this, because I'm not going to give a guarantee on this one. We are making significant inroads into Japan as you know. That requires from us a resource base to help grow that market place for us, it's important. And, therefore there may be some relatively minor additional expense in SG&A to take us into Japan. But it's going to be single-digit as opposed to double-digit growth there, in the SG&A and R&D numbers. Okay?

Thank you.

Okay, next question is from Mr. Berhaus(ph). Go ahead please.

Yes, a question on your gross margin. Also, in earlier questions there has been a -- in earlier quarters there has been some question about your gross margin 300mm versus 200mm. Considering your backlog being mostly exposed to 300mm, is it fair to say that your gross margin 300mm are very close to 200mm?

And, could you indicate what the relative gross margins are for 300mm versus 200mm at the moment?

I think what we said in the past was that -- and Peter can correct me if I'm wrong here. That they are converging, I mean obviously two years ago gross margin of 300mm was terrible, it was a new system etcetera.

They are converging, and I think that that convergence point will be reached early in 05. So, as we go through this year the gap between the two will narrow. 200mm still at a slight advantage because of the -- the tremendous experience we have with the product. But that gap is narrowing, there is a convergence point, it looks like early 05, maybe Q1, Q2 of 05 we will meet that convergence point, possibly a cross over point there.

And I think the biggest driver there to add onto that is, the cost of goods reduction program on the 300mm platform, which will conclude, not entirely, but in the shape of the program as we -- currently execute it, will end in Q1 of 05 and that will be part of the general cost of goods reduction we always have on our products. But the focused efforts will be there in the Q1 of 05.

Thank you.

And then the gross margin we have forecast obviously is the blended mix that we see right now. So, when we see 40% in Q4 this year, that's taking into account the mix shift between 200mm and 300mm.

Right, right. Okay, thank you.

The next question is from Mr. Gayle(ph). Go ahead please. Mr. Gayle are you there?

We hear you. We can hear you.

Go ahead Mr. Gayle.

Yes hello, can you guys hear me?

Yes.

We hear you Matt.

Yes, I just have a question on your ability to do churns orders in the second half. I think when you gave the breakdown of the way the backlog shakes out it's about 100 units are in the backlog for new systems in the second half. I'm just interested in what type of capacity you have to take in an order and ship it before the end of the year. Either based on where your component supply is and have you done any pre-ordering to be able to do a ramp quicker than 6 months?

Okay Matt. So, what we gave you was the backlog that was -- As you know, we issue a very crisp and clean backlog. What we don't tell you, because it's not yet on the books, is the orders we have in our pending file. The orders that came in the last few days, they require trimming on terms and conditions and some final adjustments. There are several handfuls of those.

And, all those orders will get inputted into the order book, some are in by now I should imagine. And, they will get turned, or churned, whichever -- we use the word turn here - turned this year. So, we have the strong ability to turn significant number of orders within the 6 month period.

In fact some of the hint of that is in our statement, when we said Q3 would be flattish on new systems, with acceleration taking off again in Q4. If you do some estimates on the backlog we've given you, you may think it's going to be flatter 2 quarters. That's because as yet we haven't got all the orders in for Q4, they're coming in as we speak. And there'll be a -- without hinting here there'll be a good lift off again in Q4.

Yes, and I would like to add there, that all the material that we need for those orders has been -- that has been secured. Those were pending orders, orders we received from customers where we're still in the final negotiations of terms and conditions. And we only include it in the order backlog when we have confirmed it back to the customer. So, some of that is still pending, but from a production point of view we've already secured the material.

Okay. Thank you.

We would like to take the question, our next question from the US, and Mr. Matt Gable go ahead please.

Maybe you said this earlier, what is the ASP for new units shipped in Q2? What was the ASP?

For new units shipped in Q2? A flurry of activity with paperwork here while people try and -- €9m I think is the -- €8.9m I think we're finding that number -- €8.9m I believe.

Okay.

Euros I'm talking.

Okay, thank you.

My pleasure.

The next question is from Mr. Jay Daner(ph) from the US as well. Go ahead please.

Okay, thank you very much, Jay Daner from JP Morgan. Doug the question is what was the unit mix of new system orders between 300mm and 200mm in the second quarter? And, how does that compare to what it was in the first quarter? And, what I'm trying to get at, is this the sort of signal the shift, or the second wave from 200mm to a bias towards 300mm?

Okay Jay, let me comment -- kind of not quantify it, but - one of my colleagues is just trying to find that precise number. I don't know the precise number, we'll try and give it to you in a few seconds if you hang on there. But, absolutely you're right -- without knowing the number I can tell you there was a mix shift in Q2. And, if you recall Jay, because I know you're regular on this line with us every quarter. We forecast that even in January of this year and again in April of this year.

And, we said there'd be 2 phases, the first phase filling out 200mm shelves and the second phase beginning to expand into the new 300mm shelves. We're now going through the transition onto that second phase. I think it'll -- it's a way to go yet obviously.

So, I can't give you a number just on the phone just now. Maybe my colleagues are still trying to find it. But you're right in principle, that shift has happened, and it reflects itself, most obviously, in our average selling price on the backlog. You see how much it's gone up in the past 3 months.

If you can get the data in the next several minutes--

They're trying right now, so there's--

Why don't you just move onto the next question and come back if you can find it?

Thanks Jay, we'll do that.

Yeah we have this only on the backlog [indiscernible] not on the orders.

We can find it out.

Jay I don't think in this room we have that information available actually. We've got most things, but not that. We can get back to you on that one though if you would like that?

Do you have the unit backlog split between the first and second quarter, 300mm/200mm?

Yes we do.

Could we -- While we're looking could we move onto the next question rather than have [inaudible] -- We'll come back to you Jay in a few seconds.

Okay next question from Mr. Donoghue(ph). Go ahead please.

Hi, good afternoon gentlemen. I had a question on the ASP for orders in Q2. If I look at what you report in terms of backlog for new systems, and if I try to construct orders every quarter and the ASP. Basically in Q2 I found an ASP for orders for new systems of €13.7m, which is a very strong increase.

I was wondering, besides obviously the shift towards 300mm, was there also a shift towards high-end 300mm an Immersion?

Yes. So, the shift was both ways, both the 300mm and towards 193mm and we took some Immersion orders don't forget in Q2 as well.

And would you expect that to be a trend or well it can fluctuate on -- maybe even quarter?

It's going to fluctuate Jean but, you know, over a long time, I wouldn't speak for a month or a quarter, over a long time it's a trend. And it's a demonstrable trend. When I joined this Company 5 years ago the ASP was around -- I don't know €4.5m/€5m. Now we're looking at €10m and that's trend going to continue.

Each system we produce has more features, it is more complex, has more throughput, more chip generating capacity, and therefore it's got a higher value, and customers pay us for that.

Thank you.

The next question is from Mr. Bilton(ph). Go ahead please.

Yes, hello its [indiscernible] Bilton here, West LB. A few quarters ago you forecast that the trend in used systems would slow down and you would see a limited availability of used systems in the market.

Yet, until today in the order backlog and the shipments you stay very strong. Do you have any explanations for that? Or what do you see now? Do you think use will go down from this point onwards?

Forecasting the used market is the most difficult task we have actually. I don't say that to excuse myself, but it's a fact. We know this -- it's difficult to tell what a customer is going to want to do with his old equipment sometimes and, so, it's difficult for us to forecast that.

I think the trend will be down however. I think that the people with used equipment -- are using it fairly fully right now in their existing fabs, and therefore there'll be a decreasing availability of those systems on the market over the next -- maybe not the next 3 months, the next 3 to 12 months, but time will tell.

And, if we're wrong on that one, forgive us. It's a small thing to forecast and it's very easy to get it wrong.

But in the near future it won't change?

Well you see we have a used system backlog of 27 units for the next 6 months or so. 12 months actually, but most of it's in the next 6 months. It's skewed very heavily in the short term. And that's how it tends to be. So, between conference calls, every quarter we -- you know there's a lot of churn of used systems.

So, I think that you're going to see maybe Q3 and Q4 flattish, not growing, as the rest of the market grows around it. And, probably a slight reduction maybe in Q2 of next year as the market finally dries up for those systems for a while.

Okay. Thank you.

The next question is from Mr. Van de Breggen(ph). Go ahead please.

Yes -- Referring to your slides 6 and 7 of your presentation, very interesting. I was trying to wondering how do you match -- I mean currently the -- some kind of going along utilization capacity and demand. With the current growth rates, you have -- I mean in revenues and also in units?

I guess the question was how do we correlate the semi-conductor demand against the capacity and therefore derive the utilization factor for it? Is that correct?

Exactly, yes.

Well, there are all kinds of ways to get to the answer, and I won't bore you with the long tweakers on this one. Suffice it to say we use 5 techniques and average them out. The world has probably hundreds of techniques that do the same thing.

So, the utilization is a published number historically, so you can -- you know by type of end user what utilization is. Don't forget all of our systems are on line, so we also know per system the utilization of our tools in the factory. So, that's one factor.

The growth of semi-conductors in dollars is well forecasted, and let me say largely accurately in the 1 to 2 year category, beyond that who knows. We can convert that into units and the millions of square inch of silicone are also well forecast, and that is a kind of a correlation point there.

And then we can look at the sales of steppers that we believe is going to be the worldwide sale of steppers this year and next year, based on some assumptions. Look at the size mix between 200mm and 300mm. And, work out from that what capacity we're going to install along with our two competitors over the years.

And it works out that that capacity that is going to be installed from the 3 major companies who produce steppers and scanners, and the calculated demand based on forecast on semi-conductor revenue sales and ASP's and millions of square inch of silicone, line up to give some tension still through this year and well into next year. And, that's kind of supported by the high utilization factor. So, we do it by that technique.

You can see some of the assumption we use on page 5 of the presentation.

Pretty atypical cycle then?

Sorry?

Pretty atypical cycle then?

I'm sorry again I missed it. You've broken up.

Your assumption is then, it's a pretty atypical not typical cycle?

I think it's slightly atypical just now. I mean, don't forget our customers have been in a growth for 2.5 years or so. They were leading off previous install capacity in 2000. That shows on slide 7 where there was a distinct lag between the sales of semi-conductors and the bookings, if you like of equipment.

So, now that's caught up with itself, and now cautious customers are placing orders in line with demand. There's no exuberance out there yet. So there's no Blenheim light rush to go -- you know, over populate their fabs. And, when there is, we'll see that coming through in different ways.

So, it is atypical because of the huge growth in 2000 and a huge fall in 2001 left a lot of under utilized capacity.

I would like to add that for the slide number 6 we use the semi-conductor revenue growth on the low end of that range that we give on page 5. So, for this year, 2004, 25% growth and for 2005, 14% growth.

Whereas if you ask me personally, I'm a 30% this year and near 30% for next year as well. So that's towards the top end of the ranges. For that we chose the conservative numbers from your perspective to make sure that we don't over exaggerate this tightness of supply.

Thanks very much.

One answer to a question that Jay Daner asked about the backlog split in unit terms in 300mm and 200mm. In the first quarter, the first quarter of 2004 we had a backlog of 163 units of which 40% was relating to 300mm tools and 60% to 200mm tools. That reversed in Q2 where we had a backlog of 174 units, where 60% was 300mm units and 40% was 200mm units. So, that's for Jay -- That's the answer to Jay's question.

Next question please.

We'd like to go back to the US. First Mr. Irani(ph). Go ahead please.

Thank you. One of the trends in your bookings this quarter gentlemen was the units book came down, but the bookings growth in fact was strong, and probably better than the industry average.

Could you talk about booking -- unit booking trends ahead and, you know, where you see the count of tools per fab, are we still at about 40 lithography tools per line? And, do you see the unit bookings reaccelerating in addition to the favorable mix vibrations that you're seeing right now?

Okay. So, I mean guess what everybody realized, I'll just repeat it, although I'm sure it's common knowledge.

Every 300mm piece of equipment outputs in the same unit of time if it's the same speed; 2.5 times as much silicone as a 200mm, I think that's the background to your question in a way Ali.

So, we will see possibly a slow down in the growth of unit sales, because each unit is worth 2.5 times as much silicone as -- you know as the previous 200mm unit. But, of course, the value of those units is considerably higher, the value to me and to the customer by the way.

And, therefore, as you saw in Q2, we increased our backlog of units by only 10 systems net number, but we increased the value of the backlog by over 30%, from €1.35b to €1.8b. That trend will continue -- Will continue until the 200mm sales reduce to some kind of base level, replacement base level, and then it may flatten out. We're years away from that point yet.

So, we may find that the number of systems that we -- and previously have talked about quite freely with you people, because it was a constant, all 200mm systems for the past 10 years. We'll now focus more on the value of the backlog, because frankly the number of units, whilst it's important from an approximation, the value plan is what drives our bottom line.

So, we'll be increasingly focusing on that very fact. That as the units begin to slow down in growth possibly, not reduce but slow down, the value will continue to accelerate, and that was demonstrated precisely in Q2.

I would like to add to that, are the -- Is that if you read through our presentation on the website and you take our expectations for the year 2005, which is on page 5, which is basically the worldwide litho forecast. You take our market share that we actually have, that means that 2005 for us is going to be a strong year. I think that most of the analysts would have that in their models, that 2005 is a better year for us than the year 2004.

In order to support that good year, we would need to see sequential growth in the order backlog, at least in the value, but very likely also in the number of the units. So, that's why we think that throughout this year we will experience that.

You're preaching to the choir Peter. I'm actually wondering if the unit growth reaccelerates only because we're moving to some of these higher metallization processes. You know, some part of your customers, the OEM guys are now planning 60,000/100,000 wafer start, 300mm a month fabs. And the other part of your customers, the foundry and the larger customers are migrating towards 10 to 12 metal layers.

And, I'm just trying to identify when -- If you have visibility as to when the unit acceleration would come to add to the favorable mix migrations you're seeing right now?

So, you're taking it to the next degree now. So, we have two things working in our favor from a unit point of view. One thing that works against us, if you like, from a unit point of view, is the wafer size increase obviously. The one thing that works for us from a unit point of view is the fact that processes have become more complex. So, a 30 layer process 5 years ago is now a 40 layer process. Therefore it requires more, not less, litho production hours to produce the same product.

And, the fab size, which I remember doing fabs, big fabs of 10,000 starts a month and then 20,000 starts a month and now its 30,000. It's coming nearer to 50,000, 60,000 and 80,000 I hear talked about as well.

So, that all mitigates in favor of higher volume. So, you've got to work out the -- the kind of simultaneous equation yourself here. But I think there will be a second kick to 300mm growth, probably -- I think it could come next year personally, where some of these more advanced processes begin to really move through into production, consuming a lot more stepper production hours, and therefore requiring many more units to support them.

Great. And Doug one final question. I'm wondering if the better yields at Intel on 90 nanometer have to do with your 193 nanometer tools and production there.

You'll have to ask Intel that. But I would like to think so.

Okay. Thank you.

The next question is also from the US. Mr. Stuart Adrian. Go ahead please.

Yes hi there. Just looking at the gross margin in the second quarter, I think you gave relatively specific guidance of an improvement of about 200 basis points and you managed to upside that by quite a considerable margin. Could you just say what changed in the quarter to allow you to get that? Was that you just being conservative? And, if so, can we look at the numbers you've provided for Q3 and Q4 in a similar sort of light.

The guilty man for that one - speak now my CFO Peter Wennink.

It's -- The main reason is that the outlook for the second quarter was indeed higher than we originally thought. We shipped more systems than we originally thought.

Okay.

So that gave a -- gave a much better loading in -- the faster you get the much better coverage of the fixed costs. And, also some of the cost of goods reductions that we were planning came in somewhat earlier.

And, I think that was a favorable mix. But since the extra systems basically were pull-ins from the third quarter, that means that we should look at the third quarter as a quarter that is in unit terms is about equal to the second quarter, with an acceleration in the fourth. So, that's -- we will not see that same kicker effect in Q3, because Q3 generally we have a couple of weeks where we are at 70% production capacity because of the vacation period. So, that's why we think and we expect that the outlook in terms of new systems will be about a similar rise in the second quarter.

Okay. And, just a clarification. On the R&D guidance you've given, is that net of your credit?

Yes.

It is. Okay, thanks very much.

The next question is also from the US. Mr. Mark Fitzgerald. Go ahead please.

Right, I was wondering on the backlog here on systems, if you can give us a split on memory and micro-processor and foundry?

Yes, so memory is 36% by value of the €1.8b. Foundry is 43% by value of the €1.8b and the IDM which everyone else, 21%. That should add up to 100% if our colleagues have got their sums right.

Okay. Thank you.

My pleasure. By the way, it's acceptable to give us congratulations on our results you know.

Okay. The next question is from Mr. Power from [indiscernible]. Go ahead please.

Hi guys, it's Mark Power here from Redburn(ph) & Partners. A question for Peter. Based on your assessment of the [S&E] market, what would you estimate would be the effect on future operating expenses? Supposing ES&L have entered the USA market.

And, secondly, does it look like entry into this market would require significant upfront capital?

I think those are all relevant questions, at the moment we have communicated an affirmative position on whether to enter that market or not. I'm not going -- on this call going to give any speculative indications of level of operating costs or capital that we would need. [Be great] to have this question first, whether it's sensible for us not to go into that market. And, I think that -- you might want to spend one or two minutes on this.

But we are looking into this market, that's generally known. We are looking into what the customers in that market need. And, we are currently investigating it. Hopefully this year we can finalize that -- some of the technical issues that we think need to be resolved before we can take a decision.

So, first things first, that means that the second [indiscernible] product aspect of it needs to be resolved. We need to take a position on it. And when we have taken that position we will go out and we will talk about it in length, I am absolutely certain about capital that we need and about the impact on the operating costs. So, it's too early Mark.

Okay.

Next question from Mr. Shearer(ph). Go ahead please.

Good afternoon gents. First unprompted, I'll congratulate you on your great quarter there. Gents, I just wanted to ask what would persuade you to go into the LCD stepper market. And, are you anywhere closer in making that decision?

All the persuaders [indiscernible] colleagues, said ultimately a good business case, okay? I know that sounds a bit trite I know. But, behind a good business case you've got to have a robust and growing market. I think there's no doubt it is a robust and growing market. There's an 80, 100, 120, 150 system sales you know in that market as we speak. We enjoy none of that.

You've got to have the right kind of customer and world coverage. I think we have that. We're connected to all those customers as it happens through our primary business. And you've got to have a product which sells on its merit, not just based on some cost competitive -- you know, let's go and compete on price position.

And that is the point that Peter alluded to, we're doing some careful investigations internally and with potential customers, to see if we do have such a concept. So, that's what would convince us.

I mean, it's quite simple and you know, you can see there's no magic in that. And, we will come to the right conclusion when we have all those facts, and I won't even predict when that will be.

That's great Doug. Thank you.

The next question is from Mr. Orji. Go ahead please.

I just have one question on the customer win in Japan. Can you clarify for us what type of tools this customer win is? Is it a full line [indiscernible] or just Immersion related? Also, can you tell us what other products you are making in Japan? With this win in the bag, do you think there's a lot more success to come?

Yes, what I'm prepared to say, and I'd recognize that customers have the right of veto, quite properly, on information about them. What I'm prepared to say right now is what I can say, and I won't go beyond that in confidence to our customers. It's a 200 millimeter tool, and therefore it's advanced. It's the beginning of a process, and I'm going to say no more about that one specific example of the customer that we've indicated we'd won in the press release today. So that's as far as I'm going to go on that one.

Regarding our general progress in Japan, you know, it's good to look back and see what happens, so in the past three years we've gone from zero customers to now four customers. That, I think, is good progress. I'm never satisfied, none of us are, but it's good progress. It is a difficult market for us. Geographically, it's a long way away, as you can imagine, and of course, we have some competitors there who will do anything, quite bluntly, to keep us out of their market. They are ably supported by other issues.

So we're making good progress, and in fact, we have now, having made the progress we have made, we have now decided to accelerate our presence in Japan, our involvement there, our progress there, by putting resources in place in anticipation of more orders to come in the future, and I'm talking medium and long-term future, not tomorrow afternoon. That was, in fact, in the press release recently, another press release, but that's the message recently from our colleagues.

If you remember earlier, I alluded to the fact that SG&A next year may go up a little bit, due to inflation, and possibly one or two specific investments we want to make for the future. Japan is one of them.

It's, by the way, not the sort of the people that we're [indiscernible].

No, it's nothing like that, no. So in general terms, a very valued customer, a new one, with an advanced 200 millimeter tool, the beginning of, hopefully, a healthy relationship, and we certainly have been quite open about the fact that we deserve more market share in Japan, but we have to earn that. We're going to earn that by supplying great tools, and putting excellent, qualified people in place in Japan to represent us, and that's happening, and we've announced that already.

Okay, thank you very much.

The next question is from Mr. Ramo (ph). Go ahead, please.

Hi, good afternoon. Would it be possible to have an update on your option programs, please?

Yes, that's for the customers, and we're in very intense discussions about terms and conditions. That's the only thing there that I can say. We're a bit disappointed that it isn't going faster, but we're moving, basically ahead on this new proposal, inch by inch, you could say, but we're getting there. It is new for our customers. It is new for us, and what the customers want is the guarantee on the option delivery, which we can guarantee, so it does have a new smell to them, so that's why it takes a little longer than we thought, as we're moving in the right direction.

Okay, thank you.

The next question is from Mr. Minon (ph). Go ahead, please.

Yes, hi, good afternoon. Congratulations on a great quarter. Just a question on margins into 2005 - given that your ASP is going to be clearly higher, even from Q1 2005 and continuing to be possibly sequentially increasing through the rest of the year, and given that you're saying that you're peaking on cost reduction plans around Q1 2005, does your margin have much more leverage beyond 40% into 2005?

Well, you said much more, as the term suggests. Yes, it has more leverage, certainly. We never pretended that 40% is satisfactory for us, and in fact, in the last peak year in 2000, we were over 40%, around 42%, I think, from memory. We said also, probably too many times to colleagues such as yourself, that we can get back to those margins in the next peak, and I think the next peak is next year, so we are putting a number on it. We expect to go through that 40% value, which we'll achieve in Q4, providing all the market conditions are as we anticipate, and there's no untoward happening.

The mix shift, by the way, the ASP movement, is a result of more exotic products, if you like. It is not necessarily a direct reflection of gross margins. I must caution you slightly on that, but your primary point is, can we do better than 40%? The answer is, we have done in the past, and we can do in the future.

I would like to add, there is a lot of focus on the gross margin issue, but I think you should look at the gross margin, and look at the operating expense level. The OpEx level, as we call it, control the OpEx level which we clearly intend to do, and Doug made quite clear statements about that, that you can only think about single digit growth next year, because of reasons that he mentioned. You should focus on the operating income level, and when you compare that operating income line of what we're seeing in this cycle, as compared to the previous peak, then certainly we will break the operating income margin. That's compared to the peak year 2000. I mean, you can do the math.

We'll break it high, wide and handsome.

Thank you.

Okay, the next question again from the US, Mr. Bombani (ph). Go ahead, please.

Yes, I have a question on your backlog. I think you mentioned earlier that you are expecting your backlog, in revenue terms, to be up sequentially. Could you confirm that?

Yes, we can confirm that.

And would you expect your net new orders in revenue terms, to be up sequentially?

In revenue terms, did you say?

In euro terms, yes.

In revenue terms, yes, I think we can say that as well, actually. It's quite a trick little question, actually, I can give you some math here, but I think we can also confirm that, yes.

Thank you.

The next question again from Europe, Mr. Tasson (ph), go ahead, please. Mr. Tasson, are you there?

Yes, it's Joe Dutton from UBS. I just wondered if you could give us some detail on the TWINSCAN, the percentage of the backlog, and also the percentage of the order intake during the second quarter.

The percentage of backlog, by value again, it's 82%.

Yes. In units, it's usual that you get that number. There's a backlog of 60% in unit terms of TWINSCAN products, and 40% 200 millimeter. That's in units. In value, it's in the presentation. It's 82% in value for TWINSCAN and 18% for steppers and scanners, 200 millimeter. We don't have that data available exactly on the bookings, but this is what we currently have here. It's probably separate [with the ART].

Right, thanks very much.

The next question is from Mr. Rees. Go ahead, please.

Yes, good afternoon, maybe a more general question, but we have heard from [Inteleco] that they have been more successful, featuring higher yields for the 300 millimeter production capacity. Do you think that maybe that there may be concerns at the overall [indiscernible], and where that could be ramped faster to higher yields, and therefore maybe capacity could come up faster?

I think not, actually, you know, and I don't account for Inteleco. They give their own advice, not usually with me. But in general terms, there is now a clear breakthrough on yield between 200 and 200 millimeter. You would expect that. It's higher class equipment, and it's designed with modern techniques, and it's therefore able to yield higher. That was built in to the original equations and business plans, and I think the differences we see, mostly it's slightly positive.

It's not hugely positive, and in this, you're talking in yields that go from, you know, estimated - someone estimates 90% of yield and you get 92%. It's not exactly a doubling of yield, it's in the margin there, so the good news is, it's positive now, not negative. So I don't think it creates a great fear, if that's your concern, that suddenly we're going to find quite as many chips as we thought coming out of the equipment, and therefore, you know, it's going to limit the growth. Not at all. It's a relatively modest increase compared to what their original plans of our customers indicated.

I think that probably is a tribute to the equipment itself, and I'm sure that many a good consumer is going to reap the benefits of that in a low-cost phone or a low-cost DVD rewriteable, or whatever.

Fine. Okay, thanks.

Just one addition, not to Doug's answer, but the percentage bookings of 300 millimeter in the second quarter was 76% in units, for this quarter, if you're interested in that. That was Jay's question, also the question from Mr. Dutton from UBS.

The next question is from Mr. Pelzer. Go ahead, please.

Yes guys, just a bit of housekeeping stuff, actually. Peter, could you remind me what the CapEx budget and the DNA is, that you're planning for 2004, and then whether the working capital will need to rise? Inventories seem to be in check for the time being, but just as accounts receivables, basically?

Well, Peter is getting the numbers there. I do like it when you guys refer to my CFO as a housekeeper! I appreciate that very much!

I was referring to myself as a housekeeper!

Well, as you know, this year, we will see an acceleration of some investments in probably flat equipment, and largely in the equipments side. We are going to have demo versions of the Immersion tools available, and also, of our new 1400 tools. So that will give rise to a CapEx rise which was net for the year to date, actually €50m only, but that included €17m of disposals, so we had €32m of the investments, and the estimate is currently that it will be about €54m for the remainder of the year, so that will bring gross purchases of BPNE to about €80m to €82m. The other question was, the depreciation in amortization?

Yes, and how much - is there any need to grow working capital via accounts receivable for the second half?

Yes, I think depreciation and amortization from accounts receivables was about €15m, and that will stay stable for the second half of the year. When you look at the working capital, accounts receivable went up, but what you cannot see from the information that we gave, is the actual day sales outstanding, calculated on an invoice-by-invoice basis. That actually stayed stable, and was just below 40 days, so AR went up because of the timing of the shipments. On timing, we had a lot of shipments in the month of June.

I think that on the inventory level, the inventory levels stayed, in absolute terms, as I said, at the same level. We are at inventory turns on work in progress and in raw materials of over 3 now. We're over 3.5 as a matter of fact, and our target for the end of the year is to go to four, so that means that, depending on what our Q1 outlook will be, it could be at absolute terms, the inventory level stayed higher, but in terms of inventory turns, it is going to be better.

So if we look at where the first half of 2005 could be, or especially the first quarter, the inventory could grow at between €80m and €100m in total, and we are, of course, trying to keep that as low as possible, but clearly, where we have a lot more to ship in higher value, even when terms go up, in absolute terms, inventories could, you know, grow and the quality of that inventory will be better than it is now.

That's an incredibly clear statement. Thank you.

The next question is from Mr. Crossbill (ph). Go ahead, please.

Hi, there, it's actually Andrew Griffin at Merrill Lynch. I have a question on cash-flow and congratulations on a great quarter, and particularly, as well, I think, that the balance sheet isn't as well noted is some of the other items, and since your Analysts' Day back in 2002 when Peter, you talked about improving the balance sheet, it's really coming through a lot now. How far through that process of improving the supply chain do you think there are, and is there further to go with improving how this whole side of the business works?

Yes, Andrew, I think this is something that is now engraved in everybody's way of working here, that, like I said, we have a target of inventory turn of width and raw materials of 4, which effectively means you have a quarter worth of inventories of the raw materials and width, which, in that quarter, you will always shift. The whole supply chain is now tuned to actually support that.

So I think 4 is our internal target. That will give us a higher level of cash generation, because we also have our day sales outstanding now, with about 40 days. So if you put it all to get a [indiscernible] improve, yes, perhaps a few days, but our standard payment terms is 30, so I think we're doing a pretty good job there. I think that the focus is now, if you want to generate cash, we are going to generate cash through the operating income line.

As we did in Q2, Andrew, as you noticed, so the cash will continue coming in, but through operating performance.

But increasingly through the operating performance, and of course, we are going to squeeze out everything that we can on the inventory and the width level, but since we are at the level of 3.5 in terms of turns, there is still somewhat to go, but where you do your math, for 2004 and 2005 based on the expectations, then cash-flow will come from earnings.

Great. Thanks very much.

The next question is from Mr. Gaudwa. Go ahead, please.

Yes, hi there. I think, during a press conference, Doug, you made comments that actually, you saw interest from customers billing up in recent weeks. Could you give us a bit more color in terms of where it is coming from, by either end market region or whether it's broad based? Thank you.

As you would expect, these are the big traditional ones, but also, going to our Sales Manager, Dave Chavoustie, VP, last week, actually, he was here for a couple of days, and he's now seeing some of the smaller customers, the smaller IDMs, coming back to life, these guys mainly by one or two each, but there's a lot of them. So it's kind of, as we said earlier, it's a broadening of the customer base, so when I say it's continuing to eke up from the demand point of view, we're getting pressure to ship earlier for the customers we've committed to already, to try and bring our schedules in, and we're getting now, a high level of interest from customers who, frankly, have not bought a system for probably three years, because they haven't needed to, coming back and saying, if we want one or two 200s later this year, and the next year, what's possible, etcetera.

So it's a broad based recovery still, which is good news, and there are still peaks of pressure from the big guys who are, you know, out there, trying to satisfy their customers.

Thank you.

If there are no additional questions, I would like to start with the US again. Mr. Irani (ph), go ahead, please.

Yes, thank you. Peter, a question on headcount, if you could give us an update on that, and also, I'm just wondering how you categorize split companies like IBM? Is it a [indiscernible] camp or an IBM camp?

Well, to add to the last question in the IBM camp.

Right now, we're still in the IDM mode with them, but it is a description we have occasionally, but right now, they're IBM as we're concerned.

With respect to the manpower, we have 5,000 people. We have got 5,000 people right now. That is, like somebody asked the question, I don't know whether it was now or this morning, we had a target of 4,650. We are still executing - or we have, as a matter of fact, executed the restructuring of everything about the factory and the logistics of our organization. About 20% of our people are in the factory and the logistics organization. The other 80%, the restructuring has actually gone through.

With the factory, given the fact that the market has picked up so substantially since then, and we have changed our move rate, that is, the number of system starts per week from four, to five, to six, we have added people in the production area with the vast majority on a contract basis. So I expect that by the end of the year, that we could add another 100 people, but that's probably it.

And there to repeat, Peter, they are flexible people in that they're on contracts, and therefore we can take action where necessary much more quickly.

Okay, and of the 5,000, how does the breakdown between your US operations where you're doing the sub-assembly work now internally for the TWINSCAN platforms and your European operations?

Yes, we have 2,800 people, roughly, in Europe, and 1,600 in the US, and 600 in Asia.

And of the 1,600 in the US, how many of those are presently in manufacturing?

It is about 500.

400 to 500, I guess, for the modules you mentioned, in that order.

Yes, between 450 and 500, yes.

Are you exploring, at this point, for [indiscernible] platforms a move to module assembly in Asia? This is obviously a very popular theme among many of your equipment peers, and in terms of seeking low cost supply and also some more flexible labor markets. When could we expect ASM Lithography taking steps towards Asia?

What we are effectively doing is what you see in our supply chain. We don't have a one tier supply chain. We have a two, three or sometimes even four-tier supply chain, and that is because the goods reductions came through cost reductions we forced upon our suppliers that actually went for their suppliers to Asia, to Eastern Europe, for instance. So that has already happened. We are constantly looking into the suppliers that can deliver us the right technology, at the right cost, at the right time, but although you might feel it's not visible right now, many of what you just described did actually happen in the second and the third tier in this supply chain, and that drove to lower the costs of goods reduction.

Great. No, on a very different topic, I know you discussed the SG&A at length in terms of expansion, some of the investments you have to make in Japan. I'm wondering if you could talk similarly about R&D. You do have some programs in LCD. It's a big market, but at the same time, your Semiconductor platforms are hitting a degree of maturity in terms of the bulk of the investments, and we now have a pretty good idea of the technology road-map with Immersion. Would it be fair to assume that you could hold the R&D line flat, even with the top-line expansions for 2005?

Well, first of all, it's a good question. R&D will increase. It will increase inflationary, and we do have still, some major programs to finish off. We've got the hyper-NA lens to install. That's for tools and equipment, which was the [L Power] 1700. We have an UV to, well not finish off, but to continue, and bring to a conclusion over the next three or four or five years. We have other innovative ideas for existing products, so don't underestimate R&D spend.

So for new things like LCD, I think you must expect, at that point, we will come to you with a release that says, we're going to enter this market, or not, and if we do, these are the indications. So I'll take LCD out of this discussion, because as, Peter said earlier, we're not yet in a decision making mode. I will say to you that this year, there's a little bit of expenditure on LCD from an evaluation process, just to evaluate whether the techniques work or not, and it's already contained in our R&D budget.

And I would like to confirm, and you already alluded to that, the fact that Immersion is on the road-map now. That does have an impact on our ability to contain the Research and Development spend. I mean, Doug was talking about single digit growth, and I think that is what you could expect in 2005 for all the operational expenses in R&D and SG&A, and it is not going to be 9.8% single digit, by the way, so we are containing those costs, and it is true, too, that the choices that have been made on the technology road-map are helping us in containing those costs. Yes, there is an amount, which is in the incubation group for LCD spending this year already in R&D.

Okay, and one final, technical question - we've had two years of calcium fluoride lenses down in the field, and you know, with a varying mix depending on yourselves or the Japanese OEMs. With the number of exposures going through these lenses, we get some of the darkening, or the reduction in the useful life. Are you seeing some of your early customers now come back for service replacement of lens units in your earlier platforms, and is that an opportunity to grow your service revenues in the next year ahead? Thank you.

The answer is no, no we don't see a lot of our units coming back. We should explain that although the 157 program was wholly calcium fluoride, even the advanced 193 program uses only relatively small amounts of calcium fluoride, and therefore, we haven't yet seen, and I hope we never will, serious degradation that warrants lens swaps and lens changes. So the answer is no to your question, and therefore we won't see increased service revenues because [indiscernible].

Okay, thank you.

I have a new question from Mr. Foonagogh (ph). Go ahead, please.

Oh, thanks. What kind of bottlenecks are you experiencing with your suppliers, and what is the assembly, or cycle time, in average for TWINSCAN in your targets?

I come at this on a different way. I mean, we have a bottleneck today here, and there, and everywhere. Don't we all? We've told you what we've shipped. We've told you what the backlog is, that we're going to ship, and therefore, what we're going to ship reflects what we can do. It's what the customers want. It's what we can do, so the bottlenecks that exist, today, it might be a cable, tomorrow it might a screw. I don't think that it's an issue that we want to go into with you guys, because it's all over the place, so there are no structural issues there. So just be content in the fact that we give you upfront information. We can ship those products, as indicated, and that's where it's going to be.

Right, thanks, and the assembly time?

The assembly time, depends what it is we're assembling, you could take our assembly time from all the peach parts available to making, say, a TWINSCAN -

We're down to, I guess, about ten weeks right now.

Howard's saying the time, and Peter is saying yes, so yes, it's probably eight or nine weeks. By the way, when I first joined ASML, that was the assembly time for the 200 millimeter, single scan products, so it just shows how we've made progress with TWINSCAN. But if you include the assembly time, including all the supply base along these items before it, and it goes up, obviously to, it can be six months, it can be nine months, depending on the product type.

So there's no easy answer to your question there. Internally, it's eight to ten weeks for an advanced product, with a total stack fleet time which can be six, nine, and occasionally twelve months if it's a really advanced product. But the thing is, the customers know that, and therefore, they give us orders accordingly.

Right, thanks.

The next question, a new question from the US, Mr. Flannery (ph). Go ahead, please.

Yes, good afternoon, congratulations on a good quarter. Basically I have a two-part question. What are the ARF unit numbers that are coming in, as a percentage of your total units, or your new units for 2004 and how that would look for 2005, and similarly, on the Immersion units for 2004 and 2005? Thank you.

So 193 ARF, as you referred to it, in our backlog by value is 39%, if that helps you, and it's [writtening], it's getting, if you like, thicker on 193. What was the second part of your question? Oh, Immersion, sorry, Immersion yes.

Immersion is high on price, but low in percentage of the total, yet it's a new technology. You know, the first Immersion tool hasn't yet shipped. We will ship it, in September of this year to TSMC, which are a well-known [indiscernible]. Therefore, you know, it's early days. Customers have to take one or two, get used to it, understand how to use it, in their FAB, with their [indiscernible], in their process, and I think Immersion should not figure strongly into any unit shipment kind of factor. It won't change our lifestyle on unit shipments this year, or probably next. It's important from a dollar point of view. It's a high price item, $20m-ish, is the Immersion price so far. It becomes significant, I think, in the year 2006.

I think on ARF - I mean, what's the talk in 2005, and will the units be 30% plus, of your -

Well, it's 40% of our order book, so you can extrapolate 40%, to reflect that.

In value.

Value? Oh, he said units. I'm sorry. Yes. So it could become 50% of the value, and it could therefore, become 30% of units. That's a wild guess, okay, it's speculation. If I can just calibrate it for you, does that help?

Thank you.

The next additional question is from Mr. Gill. Go ahead, please.

Yes, I have an accounting question for Peter. I was wondering, could you give us a little guidance in the share count, the upgraded share count this quarter, assuming conversion of the 2010 notes. Was there an add-back of interest expense in the quarter, and what should we be looking for, going forward, as a result of that?

Yes, in calculating the diluted number of shares, there is an add-back of the interest cost, and the interest cost add-back was - I will need to look at that - it was about €3m a quarter.

Okay, and that's the run-rate going forward?

Yes.

And would you happen to have, at your fingertips, what the trigger point in net income would be for the 2006 convertible, to be taken into the diluted share count.

Not at my fingertips. I will need to have a look at that.

Is it a substantial degree above where we are now?

Oh, no, it isn't. I'll make a note of that, and I'll get back to you.

Okay, thank you.

By the way, you win the prize. We were betting that no one would ask a question about the anti-diluted nature of that, because it's too complicated a question, too complicated an answer. You've fooled - you've won the prize!

Any word on what I get here?

Yes, you get to answer a question next time, in three months time, an extra free question in three months time!

So I get two next time? Oh, very exciting!

The next question is from Mr. Phil Griffin. Go ahead, please.

Last question, please.

Yes, hello. Going back on your best guess on Slide 5 of what the market could be in 2004, and I'm sure you've also looked at what your competitors are guiding, and if I add all the guidance together, I end up with 30 more [tools] than what's on the Slide 5. So roughly, 585, 580. What would be the mix match between this assumption and the means, showing together all the guidances?

It could be [indiscernible]. We just took the Dataquest number. We tested it ourselves, to be honest with you, and we felt it was about right, but you know, the more we modify third party independent data, the more it's suspect, so we try to stay neat and clean on that. It could be new systems, but the 554 is what I believe Dataquest were forecasting. They may have changed that number, by the way, in the recent weeks or days, I don't know. It's a new number is it, sorry? This is current, and it's only new systems, so this could be some aspect of second-hand systems, refurbished tools, and so on. But we don't use the information from any competitors added to ours in this kind of data analysis for you people. We may do it internally, but we like to stay with an objective third party number so that you can see that we're not using the numbers to suit our purpose. We're using published data here.

And it's 30 units this, the swing factors definitely, the new systems it's 30 units out of 550, which is 5%, so that's the kind of band width I think that is wanted in the estimates that we give out.

Okay, thanks.

Thanks. Gentlemen and ladies - I guess there are some. Thank you for your questions. As usual, it was exciting and interesting for us, hopefully for you as well. Thank you for those who took the time to congratulate us. The management team work hard to achieve these results, and we will continue doing so. I look forward to hopefully speaking to you again in three months time. Over and out.