艾司摩爾 (ASML) 2009 Q2 法說會逐字稿

Welcome to the ASML 2009 second quarter results conference call on July 15, 2009.

Throughout today's introduction all participants will be in listen-only mode.

After ASML's introduction there will be an opportunity to ask questions.

(Operator Instructions) I would now like to turn the conference over to Mr.

Craig DeYoung.

Go ahead please, sir.

Thank you, operator, and good morning and good afternoon, ladies and gentlemen.

This is Craig DeYoung, Vice President of Investor Relations and Corporate Communications at ASML, and I'd like to welcome you to our investor call and webcast.

As the operator mentioned, the subject of today's call is, of course, ASML's 2009 second quarter results.

With me today at the SEMICON West venue in San Francisco, and co-hosting today's call is Mr.

Eric Meurice, ASML's CEO, and joining us from our headquarters in Veldhoven, the Netherlands is our CFO Peter Wennink.

At this time, I'd like to draw your attention to the Safe Harbor statement contained in today's press release and fourth quarter results presentation, both of which you can find on our webcast at www.asml.com, and which will apply to this call and all associated presentation materials.

I would remind you that the length of the call will be 60 minutes.

And now, I'd like to turn the call over to Mr.

Meurice for a brief introduction.

Thank you, Craig.

Good morning, good afternoon.

Thank you, everyone, for attending our second quarter results conference call.

Peter and I will, as usual, provide an overview and some commentary on the quarter.

Peter will start with a review of the Q2 financial performance, and will add some comment on the short term outlook.

And I will complete this introduction with some further detail on our current position, strategic, how we head towards the end of 2009.

Following this introduction we will open up for questions.

So, Peter, please if you will.

Thank you, Eric, and welcome to everyone.

As usual I would like to take a small amount of time to review some of our -- some of the events of the last quarter as well as some details on our second quarter results.

We closed the second quarter with higher than expected sales.

However, they were weak in comparison to this time last year, and also weak compared to what we consider for the coming quarters.

Our second quarter sales were above our guided range at EUR277m.

We shipped a total of 10 systems, four of which were new.

All new tools shipped were immersion systems, causing a new system ASP of EUR31.1m, which is very high.

And the ASP of all systems was around EUR18m, but that was helped by the high ASP of new systems of around EUR10m.

That was due to the sale of two used immersion tools in the quarter.

Our services revenues re-bounded a bit and we reported EUR94m service and field options sales.

The Company's second quarter gross margin was 12.5% versus 6.7% in Q1, and this reflects an increase in sales from the low first quarter levels.

We maintain our focus on R&D with a spend of EUR118m in the second quarter, which is unchanged from Q1.

And, as predicted, the very weak first half of the year did not impact our cash balance.

As a matter of fact we generated cash, and even including our dividend payment we managed to stay virtually even compared to our cash balance at the beginning of the year.

Net cash from operations was EUR71m in the quarter.

We ended the second quarter with around EUR1.1b in cash.

Like I said, cash flow in the first half was impacted by dividend payments of around EUR86m and capital expenditures of EUR87m (sic - see press release), mainly relating to the new manufacturing facilities, which were finalized in the second quarter.

As for the outlook Q2 net bookings of 15 systems were valued at close to EUR400m.

Net immersion bookings were 13 systems, and they demonstrated a focus on technology investments.

This bookings level pushed our backlog again over EUR1.1b as of the end of the quarter.

The backlog now contains a total of 43 systems, 32 of which are immersion systems.

And this includes new and used XT immersion systems as well as our newest product the NXT.

We believe investments in leading edge technologies has resumed in the second quarter, and is indicative of a pattern of technology spend that will result in a minimum quarterly net average sales level of between EUR400m and EUR500m for the next quarters.

This range could certainly increase as a result of a need for capacity, in particular for leading edge technologies, triggered by a potential recovery of semiconductor growth, even if this recovery is limited, as there is currently little to no excess capacity at leading edge.

We expect Q3 2009 sales to be around EUR450m with a gross margin of about 30%.

Expenses are expected to decrease to EUR115m for R&D and EUR39m for SG&A.

As previously indicated we manage the company towards breakeven at a quarterly sales level of around EUR450m by the end of this year.

The targeted cost structure will support all of our strategic R&D programs, and allows appropriate upside for sufficient production capacity.

Our cash balance in the second half of 2009 is expected to fall moderately below our EUR1b target as we prepare for shipments of the first NXT systems starting in Q3 of this year, and that will accelerate in 2010.

As well as we prepare to build EUV systems planned for delivery in the second half of 2010.

Now all of these products initially carry longer production lead times, which temporarily require more working capital.

And now I'd like to turn back it over to Eric, and, Eric, for you more on our strategy for the coming quarters.

Thank you, Peter.

So, as Peter said, we are indeed recovering to quarterly sales level which we consider somewhat safe, as these levels are linked to necessary technology transitions buys by our customers, which have to happen even during a tough recession.

The question is when the capacity buys will add up, will add up to the current sales, to bring us back to our secular growth trajectory.

At this point there are no sustained indications of the world moving out of recession as you know, or of semiconductor units recovering to at least the level of 2008 historical peak.

We will have to wait for further signs before being able to forecast such a recovery, which will translate in lithography sales significantly above the current EUR450m level, quarterly level that we are guiding at this moment.

We can, however, note a certain number of positive factors.

First, the semiconductor industry has significantly shrunk total wafer out capacity, in particular by shutting down all 200mm lines which will not come back, and by transitioning to new technologies on the 300mm, which are much more capital intensive.

So this reduction of capacity makes a return to unit growth translate quickly to capacity buys, in particular, in the leading edge segments.

Second point is that the new technologies I am talking about, which comes from this technology transition buy, are necessary for the DRAM, NAND logic sectors and are clearly lithography hungry in numbers of layers, for instance, when using double patterning.

And they are also lithographic critical in terms of overlay, lens resolution, CD control.

All this allows us to project a very rich mix, product mix, and for high product ASPs.

This is the case for the current push by leading DRAM manufacturers to, for instance, DDR3 memory devices, as they provide higher performance, speed and power consumption at lower cost.

This -- and the need for the -- and the need to differentiate by driving technology even faster to 40 nanometers will support a very healthy high end immersion market in the DRAM business in 2010.

The NAND flash memory segment, currently ramping 30 nanometer technology will begin preparation of the 20 or 2x node in 2010 to hit the very important $1 per gig target price, a little bit, target price in order to start significant penetration in the laptop computer market, with solid state drive to replace hard disk drive.

In the third segment, logic and foundry, growth will be sustained by the current aggressive chip transition to 40 or 45 nanometer nodes, which now represents a significant share of the total business.

ASML's strategy and ability to invest heavily in multiple technologies in parallel, namely the XT platform for cost effective solution.

The -- number two, the NXT platform for critical application with no equivalent performance on the market today announced.

Number three, a suite of litho enablers called holistic lithography.

And number four, the next generation EUV technology.

All these -- the four efforts that we are doing in R&D is starting to show its strategic value, as the current technology transition have to rely on the most aggressive technologies and can no longer be compromised.

In summary, we believe that this recession is allowing us to focus on our technology execution for significant market credibility gain, and will allow us enough sales to manage around the breakeven level until our semiconductor customers' own unit demand fires up.

So with this Peter and I would be pleased to take your questions thank you.

Thanks, Eric.

Ladies and gentlemen, the operator will instruct you momentarily on the protocol for the Q&A session.

But before hand I'd like to ask that you kindly limit yourself to one question with one short follow-up if necessary.

This will allow us to get as many callers in as possible.

Now, operator, if we could have your instructions, and then the first question please.

Thank you.

(Operator Instructions) The first question comes from Mr.

Nicolas Gaudois.

Please state your company name followed by your question.

Proceed please.

Hi there, it is UBS.

First question is on NAND flash.

You sound a little bit more positive there on, I guess, some of the discussions you've had with at least two key customers on that.

And if we look, I guess, at substitution rate in NAND flash being pretty much close to saturation as of now, Intel commenting on solid state drive acceptance yesterday also improving, I guess one could construct that beyond shrink investment there may be requirements for capacity as well coming back next year.

So if you could give us your view on that as a starting point it would be great.

I would then have a follow-up for Peter on margins.

Thank you.

Yes, Nick, thank you for the question.

Where we -- if we -- we had to prioritize the segments which have more growth potential at this moment, we would still position DRAM first at this moment.

There is much more activity in the DRAM sector.

But closely comment in a sense that there has been a significant over-capacity built up in 2008 or so, and you can see that the market demand has been catching up on this overcapacity, and therefore the main customers are at this moment starting to plan for the restart of their business.

But I would guess at this moment most of the shipments that we will do in the next six months will be going to DRAM and logic, and that the flash business will start late into the year, and into 2010.

Added to this you are starting to see a conversion discussion of NAND from 30 nanometer which is the current new technology to 20 nanometer which again will start mid-2010.

Great.

And on the cost side your gross margin guidance for Q3 was probably a touch light versus expectations, if you could clarify the reason for that.

And how should we think about gross margins going into the back end of the year towards your breakeven target, in particular as far as variable costs are concerned.

Thank you.

Yes, with respect to the third quarter there are two reasons why the gross margin is a touch light, and it has to do with the fact that one, we are still shipping tools out of old inventory, inventory we purchased in 2008.

We are only going to replenish that in Q3 for shipment in Q4.

So cost reduction of modules and parts will only enter the income statements towards the end of the year.

That's one.

And number two is the cost reduction initiatives that we have started are going to see effect in Q3, but increasingly in Q4.

So those are the two reasons.

And that brings us to the point where we will reach that EUR450m breakeven point by the end of the year, which is then your driver, what we call the direct material margin, which is currently just a touch over 55%, which should go to 60%.

And with our current cost structure of, let's say, EUR275m per quarter you can then calculate that will bring us to breakeven point by the end of the year.

Okay, perfect, thank you.

Next question comes from Mr.

Simon Schafer.

Please state your company name followed by your question.

It's Simon Schafer with Goldman Sachs.

I was just going back to this question about NAND flash, just wondering what the famous ASML model says about under or overcapacity in that segment, because clearly it seems as if the variability towards the end of the year could be coming from that customer base, and as to whether that supply and demand is more in balance.

Thank you.

Okay.

First of all, thank you very much for calling our model the famous model, which I take it as a compliment.

And the famous model, let's say, is basically saying that with the current reduction of capacity in the world because of the retirement of 200mm capacity, as well as the transition to 3x and 4x we estimate the total worldwide capacity between 2008 and currently 2009 to have shrunk by about 20% at this moment.

So again the numbers of wafers per month that -- the installed capacity that the world allows is 20% less wafer per month.

So that 20% down on capacity is now hitting the demand curve, which is -- has always been sustained.

We estimate, I mean, we think that the demand curve of DRAM has been -- sorry, of flash NAND flash has been about stable-ish to minus 3% to 5% 2009 to 2008.

So this demand has not been shrinking as much as the capacity has, which again leads us to believe that the -- more than that -- this is why we start seeing some activity.

But this, again, booking activity is clearly, as I said, not for immediate shipment.

So there is not a full urgency.

The customers are still willing to wait a bit until they see the famous Christmas season.

That will confirm that we are back into a growth situation, which again will lead us to resume shipments into the NAND sector.

Got it, thanks so much.

And my second question would be just on this assumption of the EUR450m I think you are saying is -- or I think you are still characterizing as an investment cycle that is really for technology purchases.

And so I was wondering whether that means that your mix is still very much skewed towards immersion deliveries, partly because the NXT is kicking off in Q3.

But does that mean that your average selling price can be sustainably as high now, just at least over the near term as it was in Q2.

I know that will change once capacity comes back into the picture.

But does that mean in the next couple of quarters it stays roughly around the same level, thanks?

Yes, that is correct.

Our famous EUR450m level is based on about 10 to 12 machines.

And as you can see you need to be around on average EUR25m to EUR30m to get to this level.

So we do expect this to be sustained.

We expect even the ASP to start growing up a bit more as we -- our mix increases in NXT's versus XT.

The XT ASP, as I said, is EUR25m to EUR30m the NXT will be in the EUR35m to EUR40m.

So you will start seeing this ASP going up towards a sort of mix between XTs and NXTs into, I would say, Q4 and Q1, Q2 to next year

Got it.

Thanks much.

Next question if from Mr.

Philip Scholte.

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

Yes, gentlemen, a short follow up on the NAND flash market.

Does your guidance assume actually that you will see some orders kicking in already in the second half of this year from the NAND flash market, or could that be on top of your current sales guidance?

And the second question would be maybe a short indication on longer term margins.

If I make the right calculation in terms of your cost savings kicking in I would say that your next cycle peak margins would be close to 30%.

Is that something you could confirm?

I will leave the margin question to Peter.

Regarding the bookings, yes, there will be bookings of NAND into the second half.

Those bookings will translate into sales in 2010.

And yes, there is potential for upside above EUR450m per quarter on the NAND, but then I would call it capacity drive.

So if I am right and the business recover and goes back to a curve of growth of units in flash then it will add up above EUR450m.

Peter, regarding margin.

Yes.

Clearly the -- what I would say the peak operating margin is, of course, very much dependant on what you believe the peak sales will be.

But to make things easier for you, you can do your own calculations.

We have a direct material margin target of 60% and we will get there.

And so if you take the 60% direct material margin on a sales level that you regard the next peak, then you need to subtract per quarter, and I talk about quarterly cost, between EUR290m to EUR295m of cost per quarter.

And that is true up to and including a sales level of around EUR800m.

So when we are at EUR800m, 60% direct material margin between EUR290m, EUR295m cost level and you can make your own calculation which is actually quite easy.

And how much above the EUR800m you want to go is up to you.

That is your own insight.

But then you can use those maths whereby cost levels will go up somewhat above that level, but that will be based on variable cost.

So we would need some more people but those would be flex.

Now, how do we come to EUR290, EUR295m?

We are currently at EUR275m which is just over EUR50m lower than where we were in the second half of 2008.

We are focusing -- actually the cost reduction program has now confirmed that we will sustained 75% of that EUR50m per quarter going forward, even at EUR800m sales level, which would then bring our cost level down to that EUR290m maximum, EUR295m level.

Now -- and if you run your numbers, and you would take, let's say, a peak sales level that's above EUR4b then you could get to the numbers that you just mentioned, that's pure mathematics.

Right.

Right, thank you very much.

The next question comes from Mr.

Sandeep Deshpande.

Please state your company name, followed by your question.

Go ahead, sir.

Yes, hi, this is Sandeep Deshpande at JP Morgan.

Hi, Eric, just a question again regarding your famous model, given that semiconductor -- there is some incredubility among investors whether this semiconductor equipment market itself will grow in the next five years versus the past five years.

What does your model says in terms of if the NAND flash market is going to penetrate the notebook as well as server markets, what would be the size of the opportunity for you?

Okay, hi Sandeep.

Now we can use the term very famous, because your colleague said famous, now we go up.

And the very famous model is clearly confirming our target sales of EUR5b to say EUR6b in the next three to four years.

And this is based on an average semiconductor unit growth of about 8% to 9% per year, so we will have to get back to that.

And in order to get back to about 8% to 9% if I remember by heart, we had a growth of NAND to be in the, I think it was, 25% to 30%.

And that would correspond to NAND reaching -- sorry to solid state drive to reaching a total market share of 10% of all laptops.

So if you do that and this happens in the next three to four years, then we get towards our EUR5b to EUR6b sales, which is based on a 70% market share, so you can recreate the total market from there.

So we are doing this simulation on a regular basis based on what we learn from the different technology mix, because as you say the simulation -- as you imagine the simulation tool is as good as the assumptions we've put in technology transition multiplied by the price of each of the tools.

And so if you -- what we are -- have today in the next three years or so is a significant growth of technology or double patterning of some sort, whether it is spacer double patterning or more sophisticated double patterning with a big usage of the NXT machine, which is this machine that offers double patterning to an overlay level of 3 nanometer, which is like half of what is happening today.

We've confirmed from the customer that double patterning will require such aggressive overlay.

We are the only one offering this type of overlay.

We therefore believe that we enable the market to do this technology -- aggressive technology, and it reinforces our market share.

So in conclusion the simulation tool is always telling us about the same results.

If sometimes we change the assumptions on technologies we still come up with the same numbers, because at the end of the day it -- the lithographic is blessed by the fact that this technology transition requires an ASP increase.

So we are supported all the time by either the units or the ASP.

But to multiply -- if you multiply both you get to about the same level.

Thanks, Eric.

Just one follow up, Peter, actually following up from the previous question if you look back at previous cycles for ASML, the cycle -- the two cycles in the 90s saw a peak EBIT margin for the Company at only about 20%, whereas in 90 -- 2007 you saw over 20%, 24% in 2006 EBIT margin.

Clearly a market share gain related thing associated with your revenues.

So when you talk about this 30% EBIT margin you didn't give this figure yourself, but is this because of new cost cuts which you are implementing now or are you talking about that your market share -- it is a function of your market share which will be higher in this -- in the upcoming cycle?

Yes, I think that is -- the latter is certainly true.

When you talk about 2006/07 we had between 55% and 60%, I think we are now closer to 70%.

So that is sheer volume.

And as you know there is always a high leverage in the income statement.

Like I said if we are over our breakeven point then we would have 60% direct material margin.

Until the -- we hit the EUR800m per quarter ceiling there is an incremental addition to the operating income line of 60% of every EUR100m, which is EUR60m for every EUR100m that you will add.

So that goes quick.

So I think pure sales volume will indeed help, and it will be driven by what Eric said, the growth of the market and our 70% market share as compared to the previous cycle where it was indeed lower.

That is a big driver.

The other driver is the cost structure.

I think we are focusing on the 60% direct material margin which we were not at in previous cycles.

Also the addition of more software related products driven by solutions we have in the holistic lithography space will also help.

And those three things, so basically the new suite of products with more software-like margins, cost reductions that will come through plus the market share are drivers for the higher EBIT and EBIT margin as compared to the previous cycle.

Thanks a lot, Peter.

Next question comes from Mrs.

Mehdi Hosseini.

Please state your company name followed by your question.

It's actually Mr.

I have a question going back to your model, and I apologize if I am repeating the question, I joined in late.

With migration to 2x node for flash what kind of dollars of CapEx are we looking at?

And I would also like to ask the same question about DRAM.

I do understand the $5b to $6b revenue opportunity several years from now.

But what happens with the flash and DRAM spending over the next 12 to 18 months?

Any color would be appreciated.

You take me a bit aback, because I will have to answer you a bit scientifically, understanding or reminding myself of all those numbers.

I guess that in our model we probably have, I think, total conversion to another technology is probably in the order -- in flash is in the order of EUR1b to EUR1.5b.

But to be honest I would have to confirm that to you, I don't want to give you wrong numbers.

We are very disciplined as to what we call technology transitions which we define a certain way.

And then on top of that you have capacity adders.

So I have to be cautious that the technology transition is like recession independent, which is why we call with very strong statement that it is for EUR50m.

We have seen this 10 times.

But the additional capacity built up on the nodes depend on the numbers of units that this node is going to go for, and that depends on the recession or the getting out of recession.

So I don't want to talk about that at this moment without coming back to my lawyer the (multiple speakers).

And also I think you need to remember that those transitions don't take place at -- all at the same time.

Customers have different investment patterns so that is also spread over a certain period, like it was very clear in the DRAM space where you have leaders and you have the followers, that could be easily 18 -- 12 to 18 months in between.

Sure.

Let me rephrase the question.

Back in 2006/2007 DRAM spending was between $25b to $30b.

Given this technology migration do you think that over the next couple of years we are going to meet or exceed the 2006/2007 DRAM CapEx spending?

I think we will probably meet it but not for the same reason.

In 2006/07 there was a clear overspend, every one of us was surprised if you remember, even our model didn't show that it would go as fast as this.

2007 at the end was a -- probably 50% of the spend when compared to the, I would say, natural growth.

But, we are pretty optimistic as to DRAM re-build up of CapEx mainly because they are ramping now major technology difficulties.

Their transition to 50 nanometer is not easy.

This is being done by a current leader, two leaders.

Those guys are now running towards 40 nanometer.

And they are discovering how complicated it is to do these new nodes.

So they are ready to go for 40 and 30 type nanometer with double patterning until they transition to EUV, which again is an expensive transition.

So through the ASP, the node through the units we are going to reach the same type level of sales that we had in 2007.

Okay, thank you.

Your next question comes from Mr.

Tim Arcuri.

Please state your company name followed by your question.

Timothy Arcuri, Citi, thanks.

How confident are you that you can sustain order growth through the end of the year if in fact utilization at foundries begins to roll over in Q4 as there are some signs that that will happen?

Do you think that the upgrade cycle going on right now in DRAM and later on in NAND is that actually strong enough to sustain bookings growth in the fact of rolling over utilization at the foundries?

Definitely.

The current slight pickup that you saw in bookings have been done based with halfway through one foundry, and halfway through the second tier DRAMs.

So the next batch of orders in the next six months will be the foundry business competitors catching up a bit, as well as the first tier DRAM and, as we have just discussed, NAND.

There's now -- there is therefore much more opportunities, more segments, which will pick up in the second half than the current ones that we have seen in May that were current bookings, a bit better.

So I would say the current bookings at EUR390m, EUR400m or something, EUR400m of system is in the low side of the equation as I could project in the second half.

And on top of that, Eric, if I may add, when you talk about foundry utilization, you have to realize that the leading edge utilization on 45 nanometers has been very, very high.

And then the introduction of the 40 nanometer to 45 nanometer products is now ramping, which the installed capacity at 45 nanometers is still extremely low as compared to total installed capacity.

So you could indeed see that if total installed capacity utilization would go down, doesn't mean at the leading edge, which is basically only, like Eric said, only one foundry, that hat would go down also.

We'd be ramping introduction of new 40 nanometer to 45 nanometer products.

Agreed.

I just asked because we've never before seen a cycle where utilization has kind of rolled back over and bookings have not rolled back over.

So I'm just kind of wondering why that would be different this time (multiple speakers).

But only if you're shipping, Tim, the only thing we are shipping is 45 nanometer technology.

We were not shipping any additional capacity.

Right.

Just as a quick follow up, do you think that you're going to grow immersion backlog in both September and December quarters?

Yes.

Okay.

Thanks a lot.

Next question comes from Kai Korschelt.

Please state your company name, followed by your question.

Yes, hi.

It's Deutsche Bank.

My first question is, when you talk about the timing and magnitude, potentially of the 4X migration DRAM with the Korean vendors, what are they telling you, roughly, about those two aspects of it?

And then I have a follow up, please.

What I think we are getting is a lot of [float].

I'd say -- realize that it is strategically important for them to run this new node a bit earlier than planned for this obvious strategic reason at this moment, while the DRAM business is somehow consolidating in other terms, where the leaders have to show their colors.

So you are getting much more pressure than those guys saying, "Well, let's accelerate those nodes."

But as I said, the technology is now extremely complicated and therefore they need the latest technology.

So you get much more interest in our machinery, in our high-end machine and with the -- also holistic lithography which enables them to do those designs.

So you see that happening.

Clearly, there is going to be a second wave of activities with -- by them when they are finishing up the recipes.

And depending on the success of the yield, we will have a different rate.

So in other terms, I expect a first wave of orders and then followed by another one.

And again, the other one, we don't know yet how successful it will be.

That's great, thank you.

Then my second question would be when do you expect government bailout-funded DRAM vendors, so to speak, in Japan and Taiwan to come back with orders?

That is also a very good question.

So I would try to avoid any specifics.

Of course this is now customer data, but you have heard from the Japanese side that there is interest to invest the money they receive by the government into also 40 nanometer type technology.

And that will be very much done very near in parallel with the first tier, so in other words, ramping into Q1 and Q3 of 2010.

Regarding the Taiwanese part, there are, of course, as you know, two Taiwanese parts.

One is one aligned with the American vendor and the other one aligned with the Japanese vendor.

The one aligned with the American vendor is, at this moment, planning a ramp of 50 nanometer, which is, I would say, more of a volume question rather than a technology question.

The 4X at this moment is not yet in discussion.

Okay.

That's great, thank you.

Your next question comes from Mr.

Francois Meunier.

Please state your company name, followed by your question.

Yes.

It's actually Francois from Cazenove.

A quick question on the margins.

I'm looking at your accounts in IFRS.

And I was wondering when you'd officially move to probably only IFRS reporting, maybe next year.

And do we have to model something like EUR25m or EUR30m of incremental EBIT per quarter, which it's probably linked, well a bit of options accounting, but also mostly to amortization of R&D.

I know it's not cash, but probably on the headline numbers -- your headline numbers in the coming years will be around EUR100m to EUR120m higher at the EBIT line.

Yes.

I think clearly we don't run the company on the accounting rules in a sense that we would just pick the most favorable one.

I think we are on US GAAP for the simple reason that all our peers are on US GAAP.

And as you correctly pointed out, if we would go to IFRS, we would have to capitalize our perhaps a big part of our R&D spend.

And then amortize that over the expected lifetime, which would mean that our R&D costs in the IFRS income statement would be significantly lower than we currently show in the US GAAP.

Now my personal view is that I don't think that is the right thing to do.

I think that given the uncertainty in this industry and IFRS profile it is good to expense the R&D.

So for this particular reason, but also for the reason of comparability with our US peers, we stay on US GAAP.

We have also a very large contingent of US shareholders.

So you are right that the EBIT is indeed higher because of the R&D capitalization and the resulting amortization that hits our R&D is -- will go up going forward to a certain level where the capitalization will be higher than the amortization.

Therefore, a certain period of time, we will then have a positive impact on the IFRS results.

But that's not what we're focusing on.

Like I said, we're focusing on [book variability].

And I think also from a Company point of view, it's better to get rid of the R&D spend and not take this in as an asset.

Okay, good.

So when you're talking about margins, peak margins, you're talking about US GAAP.

US GAAP, yes.

Okay.

Now another question about -- you're talking about debooking in the presentation.

I guess that's order cancellations.

Could you give us a bit more flavor about those EUR71m of orders which have been debooked?

Where are they coming from?

This is something which is not an official debooking.

It's basically a scrubbing of the backlog that we're doing.

We're looking at certain orders where we think will the probability of those orders shipping fall outside the 12 month window.

If it falls out of the 12 month window, we take them out.

So it's our own choice.

And that has to do in fact with some orders that we received out of Taiwan where we said and Eric alluded to that the financing situation has not improved the government funding.

We aren't seeing yet the results of that, so that's why we've taken a more conservative approach on backlog.

And that should be cleaned up probably by the end of this quarter.

Yes, or by the end of the year.

To be clear, these orders still exist and the customers are still there.

It's just -- it's a delayed shipment and we want to be conservative as to measure this backlog.

Thank you.

Next question comes from Mr.

Andrew Gardiner.

Please state your company name, followed by your question.

Good afternoon.

Barclays Capital.

I'd just like a little more detail on the NAND flash business, if you could.

You've spoken quite confidently about DRAM NXT tools and the interest from the NAND side of the space there and in particular regarding the move to solid-state drives and that $1 per gig target.

I was just wondering if the lead customers or the lead vendors on that side of things are giving you any indication of when they might like to start shipping at those kinds of levels and therefore what kind of confidence you have in terms of the potential of the ramp.

No, unfortunately I will be disappointing you.

We're not a specialist of that.

So I gave you those pieces of data mainly based on editorial comments from our customers.

But they do not share with us a product strategy.

That is they just kick us because we're not shipping technology fast enough.

And so therefore we know they want the machine and we try to make logic as why they want this and why they drive technology so hard..

But I can't really tell you much more than that.

The 10% market share target is a well-known target by most of those NAND players, as you have heard.

And I now understand also that they think they will reach that type of 10% range with a target price of X, I don't remember for a big D drive of 60 gig and 120 gig.

And all this will be based on 20 nanometer type technology.

But again, don't look at that especially to the application.

I can't read that.

Okay.

Perhaps on a slightly different angle as relates to the customers then.

You highlighted in the presentation keeping capacity in your system or in manufacturing side at around the EUR800m revenue level on a quarterly basis.

Clearly we're quite a ways below that.

But I presume you wouldn't be keeping it at that kind of level if, say, over the next three, four, five quarters you didn't think you could get somewhere close to that.

So what is your thinking around that sort of thing at the moment?

Okay.

Well, it is a very important question because we are asking ourselves that question to manage the company, how fast a move back to EUR800m or more can happen.

And the message that I gave in the speech at the beginning was as follows.

It says, well, we have a base of EUR450m.

We think that's the base.

It carried us for a long time.

And on top of it will be a capacity buy.

And those capacity buys could happen within a quarter.

We can get a surprise set of orders.

That is definitely -- we could definitely will be towards EUR800m, so in a flash.

So we absolutely need to have this capacity of EUR800m and above in view of that.

And the reason we think it can be very quick is that first of all, we just had one or two flashes of that already where we have had questions about shipping within a lead time or in fact within six months specifically certain numbers are tools for capacity reasons.

And we know why these things happen with a short lead time is there is no excess capacity anymore.

It seems that most of our customers have reached a level of utilization, in particular in the areas where it's hot, which is in new technology, which is too high for taking upside.

And therefore, if there is anything that happens in the industry in the world where you go back to level of sales of 2008, which would be 150b of units, if you go back to that level, we're at 130b units, then you will go to EUR800m very quick.

Okay.

Thank you very much.

Next question comes from Mr.

Didier Scemama.

Please state your company name, followed by your question.

Good afternoon.

It's RBS.

Thanks for taking my question.

I just have a couple of question, starting with the provisions that you took again on obsolete inventories in the course.

If you could just indicate whether you've booked any of that in the cost of sales in the quarter because if I would add back this EUR43m, your gross margins would be closer to 28% rather than 12.5%.

So I was just wondering where you booked that in the P&L and whether that has come in already.

After my first question, I've got a quick follow-up.

Yes, as you have seen, not only looking at this quarter, but at the previous quarters, we always book a certain level of provisioning.

So taking it completely out would be great because it means we would have absolutely zero obsolescence waste in this Company, which is not realistic.

But you are right.

We have a conservative policy.

We are taking provisioning on some of the R&D that we are doing.

We are taking provisioning also on some of the inventories that we are having against current low level of sales that we are seeing.

So we -- yes, we have a conservative policy there.

You could say, "Well if you would have been more aggressive, would the gross margin have been higher?" Yes, it would have been higher.

How much is difficult to say because we are consistent in the application of this policy.

That's quite interesting, thanks.

And if I basically summarize what we've heard today, are we talking about both halves back on the envelope peak EPS of EUR2 on US GAAP.

If I add that EUR120m that Francois was talking about on the IFRS, I get another EUR0.20.

And if take into account the provisioning that you have on probably capacity related tools that will come in in the course of 2010, you could easily see another 5% at least or 5 points at least on gross margins once you reverse those provisions.

Your EBIT margins could see quite spectacularly above 30%, in fact, in the next peak.

I think you just wrote the base for your own report.

It's an interesting following line of conclusions, but we cannot comment of course on that.

I think you can probably write something quite interesting on that piece.

Many thanks.

Next question comes from Mr.

Jerome Ramel.

Please state your company name, followed by your question.

Yes, good afternoon.

Jerome Ramel from BNP Paribas.

First question, do you expect more bankruptcy in the DRAM, among DRAM players?

There is a possibility that in Taiwan that one or two players would go.

And the understanding that we have is that if that were the case, the assets may be transferred to one of the four key players that remain at this moment in this industry.

And as you know, the four key players are at this moment considered the survivors of the current consolidation.

And again, the others are going to be aligned to one of the four players.

They can be aligned with a contract of some sort or they can be aligned just because their assets have been transferred after bankruptcy.

Either/or doesn't change much in what exactly is happening, which I repeat is four key players, four large companies are now going to go for control of this business.

Okay.

And just one follow-up concerning the NAND.

How do you see the cost curve of your NAND customers going forward, because in the past, the NAND guys were able to cut the costs by roughly 50% per year?

And it seems that it is slowing down to roughly 30% per year with [dual] exposure too, so requiring more (inaudible) steps for the 32 nanometer node.

It seems actually that the cost curve decline for the NAND players will reduce significantly compared to what happened over the last five years.

So how do you see that going forward?

This is a very complicated question because (inaudible) message.

Cost is a factor of volume and excess capacity.

So obviously, the NAND guys have been burned significantly by an excess capacity due to the fact that there were too many players and too much expectation.

So now it seems that the players have agreed to be less aggressive than DRAM guys.

So they have been more selective as to their buildup of capacity.

And I explained that in fact we see a shrinking of capacity more than anything else.

And we do not see crazy investment, we only see technology investment.

So we therefore expect the installed base to correspond to the demand and therefore, by conclusion, the cost per chip is going to become -- the full cost per chip is going to be under control.

Further to that, the design of the chips allows a shrink level which on paper has justified the price points of -- that would be allowing new applications like the solid state drive.

So those guys seems at this moment to be under control of their cost curve.

As you said yourself, technology goes down, help them to go down.

But mainly, mainly, mainly, they seem to be investing to demand.

They don't invest to market share, if you know what I mean.

Okay.

Thank you very much.

Next question comes from Mr.

Jonathan Crossfield.

Please state your company name, followed by your question.

Hi.

Yes, it's Bank of America, Merrill Lynch.

I guess my first question would just be on the longer term strategy, which is that as you see some margin expansion going forward and we're hearing increasingly that your competitors are struggling or falling further behind.

What do you see as the key challenges for ASML on the five-year view to maintain customer relationships as the competitive environment becomes potentially more beneficial to you?

Well, it is a very insightful question.

If you haven't guessed it, in fact, the last six months for us has been a reinforcement of what you said, which is that our technology investment or over-investment, I would say, is starting to pay.

The technologies requirements are very, very strong, concerning to the customers.

And now we're coming up with products which are absolutely ahead of our competition, whether it is in double patterning or it is in EUV.

So this creates an aspiration by every customer to have a significant strategic relationship with us.

And a strategic relationship is usually a better thing than a commodity-type relationship.

So we are thrilled by this way.

But it's a beginning and now we have to execute on it.

And execute means making sure the machine works and making sure we do not bring any fear to our customers that we will take advantage when the situation is better.

So we have to be -- wait on our tools to offer them the technology but also the best cost.

And you've heard Peter spending time explaining that we want our cost structure to be extremely under control.

And you've heard him say we don't want to be going for any inflation above the current recession level cost more than say 10%.

We can see that we will offer to our customers a cost and which also allows us to improve a little bit our margin.

The big negative, the concern I have as usual in (inaudible), we are a euro-based company.

And therefore, if the macroeconomic trends make the euro "overvalued" due to financial imbalances in the world, we will be having not such an easy time to enrich our margin.

That is our concern obviously.

We've been able to live through this in the past five years.

I think we've -- as I think you remember, I mentioned the fact that all competitors have been privileged by a depreciation of the yen by about 60% in four years or something.

We've been able to manage that by still improving our margin and market share.

But we would appreciate that it doesn't continue in that sense and that the euro stabilizes or weakens.

It would be good if the euro didn't get too strong.

Then we'll come back a bit this quarter and consider this a bit of a problem.

Thanks very much, Eric.

And just as a follow-up perhaps, could you talk a little bit about the impact of FlexRay and the baseline or announcements that came out yesterday on the financials, on a one or two-year view or how should we think about that as to margin effect and pricing.

I will voluntarily remain conceptual because the whole holistic lithography, which comes from our acquisition of Brion, is mixed, meaning two, three things.

It's mixed into the price of our machines, which do not -- that we do not discount as we used to.

So we strengthen our margins through a stronger pricing where we say, "No, if there is -- if that machine is compatible or ports to the software of Brion, then this customer you cannot get the same discount."

The second point is that we are selling, of course, Brion's stuff and our majority stuff, etc., so there's a bit of sales which at this moment is still under EUR100m.

But it is -- starts to be visible.

And the third aspect is it makes our machines more competitive.

These options, like as you said, the new illuminator, which is a completely pixelized illuminator, makes this machine another level of our competition.

So these three factors will -- are currently reinforcing the margin.

And this is why Peter, was starting to say, "Well, we are really going to try to improve our material margin by 2 to 3 or 4 points." And part of it is included into this value generation thing.

But again, it's hidden in those three aspects.

Great.

Thank you very much.

Ladies and gentlemen, I think our time is expired.

So if you were unable to get through on this call, we apologize for that.

And if you still have questions, feel free to contact ASML's Investor Relations Department with those questions.

And now, Operator, I'll ask that you close out the call.

And I'd like to thank everybody for joining us today.

We look forward to talking to you through the quarter, and look forward to talking to you again at the end of next quarter.

Ladies and gentlemen, this concludes the ASML 2009 second quarter results conference call.

Thank you for participating.

You may disconnect now.