艾司摩爾 (ASML) 2012 Q4 法說會逐字稿

Ladies and gentlemen, thank you for standing by.

Welcome to the ASML 2012 fourth-quarter and annual results conference call on January 17, 2013.

Throughout today's introduction, all participants will be in a 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, Peter, and good afternoon and good morning, ladies and gentlemen.

This is Craig DeYoung, Vice President of Investor Relations at ASML.

Joining me today from our headquarters here in Veldhoven, the Netherlands, is Mr. Eric Meurice, ASML's CEO; and Mr. Peter Wennink, ASML's CFO.

As the operator suggested, the subject of today's call is ASML's fourth-quarter and annual 2012 results.

This call is also being broadcast live over the Internet at www.ASML.com, and a replay of the call will be available on our website for approximately 90 days.

Before we begin, I would like to caution listeners that comments made by management during this conference call will include forward-looking statements within the meaning of the federal security laws.

These forward-looking statements involve material risks and uncertainties.

For a discussion of risk factors, I encourage you to review the Safe Harbor statement contained in today's press release and presentation found on our website at www.ASML.com, and in ASML's Annual Report on Form 20-F and other documents as filed with the Securities and Exchange Commission.

The length of the call will be 60 minutes.

And before we begin the Q&A, I'd like to turn it over for a brief introduction by Mr. Eric Meurice.

Thank you, Craig.

Before we begin the Q&A session, Peter and I would like to provide an overview and some commentary as usual.

Peter will start with a review of our Q4 financial performance with added comment on the short-term outlook.

I will complete the introduction with a brief status update on EUV.

So, Peter, please.

Thank you, Eric.

And welcome to everyone.

Our fourth-quarter results are very much in line with expectations that we set at the end of the third quarter.

Our fourth-quarter sales results came in just above our guidance.

This quarter's sales again remained largely skewed towards the foundry IDM sectors, which was 87% of total.

And that included non-critical KrF systems, which supported capacity additions.

The combined memory represented the balance of 13%.

By the way, we are now reporting memory sales, bookings, and backlog as one total value, due to decreasing transparency around system allocations between DRAM and NAND.

The ASP of all systems recognized in Q4 was EUR22.5 million, a decrease of about 10% from the previous quarter, affected by an increased percentage of used systems.

Service and field option sales hit a record level of EUR257 million this quarter, which I will come back to in a moment.

Fourth-quarter net bookings came in at EUR667 million for 32 systems, excluding EUV, with booked ASPs of EUR20.9 million versus EUR25 million in the third quarter.

The quarter's bookings ASP was impacted by a higher percentage of KrF AU systems -- clearly, not indicative for future bookings, which will reflect the secular trend of higher value technology buys.

Our order backlog at the end of Q4 was EUR1.2 billion, including 46 systems, with excluding EUV.

And the backlog profile at quarter's end to remain very similar to that at the end of the prior quarter, with a slight further movement in favor of foundry and IDM.

Regarding our fourth-quarter and total 2012 income tax charge, I'd like to point out that we released almost EUR120 million of our accrued tax liability during the fourth quarter, as a result of the favorable conclusion of tax audits in various tax jurisdictions outside the Netherlands.

And this resulted in a nearly complete offset of the total income tax due on ASML earnings for the year 2012.

Regarding share buybacks, we completed our (technical difficulty) [EUR1.

billion] program during Q4, with a repurchase of approximately 4 million shares, reducing our total share count to about 407 million as of the end of the year.

As a result, ASML has now returned more than EUR4 billion in cash to shareholders through share buybacks and dividends since 2006.

Furthermore, due to ASML's strong financial position and operating cash flow prospects, we intend to continue our policy to return excess cash to shareholders through dividends and share buybacks.

However, for regulatory reasons, ASML will not announce any new share buyback program before Cymer's extraordinary general meeting of shareholders, which will be held on February 5, 2013.

ASML intends to again increase the dividend by 15% compared to last year.

Therefore, we will submit a proposal to the 2013 annual general meeting of shareholders to declare a dividend in respect to 2012 of EUR0.53 per ordinary share compared to a dividend of EUR0.46 per ordinary share paid in respect of 2011.

As to the outlook, we currently anticipate the sales levels for 2013 to come in at an approximately similar level as in 2012.

However, second-half 2013 sales will be higher than first-half sales, largely due to the timing of next note investments of our foundry customers.

Therefore, we anticipate Q1 total sales coming in at about EUR850 million with a gross margin of about 38%.

The gross margin in the first quarter is temporarily affected by the underabsorption of our fixed production costs due to the lower sales level in the first quarter.

And on top of this, the underabsorption is fueled by the EUV production infrastructure, which will only start seeing cost coverage once we start shipping EUV, which will be in Q2.

The impact of underabsorption accounts for approximately 2.5 percentage points of gross margin.

We expect our gross margin to recuperate in the coming quarters when sales volumes increase and EUV shipments (technical difficulty) starts ramping.

R&D for the quarter will be about EUR185 million with other income, which consists of contributions from participants of the customer co-investment program, of EUR16 million -- sixteen.

We intend to spend a total of between EUR750 million and EUR800 million on R&D (technical difficulty) ['13], with our co-investment partners contributing EUR200 million in cash, thereby fully matching our incremental R&D (technical difficulty) spend on a cash basis.

However, as explained in previous quarters, as a result of the required accounting treatment for the co-investment contribution, the effect on our statement of operations will vary, based upon the business volume of one of the key co-investment partners.

The contribution of this partner will be included in the gross margin, whereas the contribution of the other two participants will be accounted for as Other Income under the Operating Income line.

Actuals of the co-investment contribution to R&D will be reported at the end of each quarter.

And SG&A will continue at the slightly elevated level of EUR63 million, as we anticipate about EUR6 million in additional expenses related to the pending Cymer acquisition.

Now going into 2013, we see sustained demand for the logic sector for continued 28 nanometer node RAMs.

We also recognize the strategic importance of the transition Bio-logic makers to the litho-intensive 20 nanometer and below nodes, and the associated requirements for a minimum credible capacity which would be installed by year-end 2013.

This, along with our EUV system delivery, plans up to 11 NXE3300 systems allow us to confidently plan our 2013 business as being very similar to 2012.

This includes a very low level of demand from both memory sectors, with -- however, with a possibility of upside in NAND systems.

That demand could (technical difficulty) later this year.

And that is not right yet reflected in our current assessment of the 2013 potential sales levels.

Lastly, I would also like to (td) high level of service and field options sales as we saw in the fourth quarter.

Enabled in large part by the continued adoption across the installed base of our performance and housing field options, including throughput and overlay upgrades, this level of server sales will continue through 2013, supported in large part by our holistic lithography products, with some customers beginning to implement our integrated metrology and feedback loop solutions, providing a much-needed (technical difficulty) level of process control at the 20 nanometer node and below.

And with that, I would like to turn it back over to you, Eric.

Thank you, Peter.

So let me talk a bit about EUV but first about Cymer.

In Q4, we announced our intent to acquire the light source supplier Cymer in a cash and stock deal.

We currently are awaiting several international regulatory authority approvals while the clearance has been granted by CFIUS, the Committee on Foreign Investment in the United States, and the German antitrust authorities.

In addition, a vote on the (technical difficulty) of the deal by the Cymer shareholders will take place on February 5. We expect that this acquisition will therefore be completed in the first half of this year, after having obtained the necessary approval in the US and the antitrust authorities, and in the different remaining foreign jurisdictions.

In the meantime, we are continuing our long-term close cooperation with Cymer and, of course, are working on EUV.

Both on the source and on the scanner, we moved ahead significantly in Q4 towards our 2014 production ramp goal.

Firstly, progress has been made towards confirming what we call the source proof of performance, as we were able to double the source power in tests and simulation up to 60 watt with no indication of performance degradation.

We, of course, need to reach our long-set target of 105 watt, which is equivalent to about 69 wafer per hour, but we are very encouraged, of course, by the data produced this quarter on 60 watts.

Secondly, we have operated the first integrated UV source with a stable exposure power up to 40 watts with good dose control, enabling full field exposure over an extended period of time.

Again, a fairly good piece of information here.

Thirdly, we have completed the integration of our first scanner machine, the NXE3300, and have now met a number of significant overlay and imaging performance milestones.

We have achieved imaging down to 14 nanometers, and mix and match immersion to EUV overlay results of below 5 nanometers.

These three areas of progress underpin our confidence to reach the 105 watt or the 69 wafer per hour target for 2014 production ramp.

And these areas of progress support the preparation for shipment of our 11 production, NXE3300's.

These 11 production units, I remind you, are not specified to reach immediately the 69 wafer per hour target specification at shipment, but will ship when assembly can be completed.

Regarding the production planning phase, our goal is to ship about 8 to 12 systems in total in 2014.

We have already booked 4 of these NXE3300 production systems, and intend to book several more in the next few months, as the requirements for our customers for EUV are becoming (technical difficulty) with each performance progress step being made.

DRAM manufacturers, for instance, currently suggest a potential need for imaging (technical difficulty) to four layers, which is a bit better than we believed before, to be dedicated to EUV, at high 1X nodes.

And the logic devicemakers may require between five, up to 17 EUV exposures at the nodes sub-14 nanometers.

So parallel to EUV -- which remains the only cost-effective emerging solution for customers to do the next generation IC scaling any way -- we are developing 450 millimeter wafer form factor exposure tools.

As Peter mentioned, our customer co-investment program is indeed enabling this, enabling this huge investment in normal business, investment in EUV, investment in 450 development.

The 450 millimeter system's timing is targeted for 2016 pre-production, with production systems targeted for 2018, if, of course, customers confirm the ramp interest in an agreed-upon timeframe.

So, with this update on the technology and the product, with Peter, we are -- we would be pleased to take your questions.

Thanks, Eric.

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

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

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

Now, operator, could we have your instructions and then first question, please?

(Operator Instructions) Jerome Ramel.

Please state your company name followed by your question.

Jerome Ramel from Exane BNP Paribas.

One question, Eric.

How do you see the installed capacity for the 28 nanometer node?

And going forward, how do you see the size potentially in terms of wafer start demand for the 20 nanometer node?

Okay, this is a very good question.

So we estimate that we'll see a shipment that we are supposed to do up to Q3 of 2013.

The total customers will reach about 320-ish-thousand wafer installed base of the 28 to 32 nanometer node, which 320,000, 340,000 wafer per month type capacity is a good, good node -- on the high side, but not exceptional.

What surprised us, and the reason why we've been bullish this year -- and this is the first year in eight years or so, nine years -- we've guided the full-year, is indeed to your question.

We are now seeing the buildup of capacity for 20 nanometer -- 20 nanometer, 22 nanometer -- in the same environment, in the foundry logic environment.

And this is starting earlier than we would have expected in the first place.

We can discuss later about the drivers of that.

We expect the numbers of wafers to be installed with the current forecast that we have to be in the area of about only [60K] wafers -- well, I would say it's [40K to 60K] wafers -- will be installed by the first quarter of 2014.

This is very important.

First, [40K, 60K] is not a very large number, but it requires a lot of lithography machines, because each wafer of 20 nanometer, 22 nanometer, requires [double] (technical difficulty) machine, and a significant amount more of holistic lithography products, which is good margin products, to make the same numbers of wafers.

So, you have a factor two here, which is, again, why we've said with some confidence that although we do not see good business in the PC arena, we do not see a pickup of the memory business at this time -- that doesn't mean it won't happen, but today, we don't see it.

There's no signal.

But the only thing we see is 20 nanometer, 22 nanometer, and this is good.

And the multiplier intensity force us to guide that, in fact, the second half of the year will be so good that it compensates for the beginning half, which is slow.

And just maybe a follow-up on EUV.

What's going to be the maximum capacity of shipments you will have by 2015?

Are we talking about 24 units?

So, for 2015, our simulation says that the market will probably need about 30 to 36, so we will have the capacity for 30 to 36.

We could always do a bit more, but we don't think it would be necessary to overheat on that one.

And then we will go towards a capacity about 60, 66 in the 2016, 2017 range.

Thank you very much.

Stephane Houri.

Please state your company name followed by your question.

Stephane Houri, Natixis.

The first question I would have is on the visibility that you have on the orders.

Because you have a rebound in sales starting in Q2, if I understand well, and an acceleration in [H2] because of the 2014 nanometer transition.

Given your lead-time, is it fair to assume that the orders should rebound as early as in Q1?

Thank you.

Oh, absolutely.

The orders will pick up in Q1.

In our opinion, with Peter, we don't think the orders will pick up as much as they should.

In another terms, the customers who we know, of course, will be somehow putting the orders within lead-time.

This project that we're talking about is such huge strategic project that we are building on commitment and on requested PO, buildup is more the administrative process that will go with it.

So indeed, there is going to be a pickup in Q1 and we expect this to continue.

But again, as you will not be able to easily use a standard lead-time of 6 to 9 months to guide to know what the revenue is going to be.

The PO's will be coming, in fact, late to the ramp, because the ramp is so obviously discussed, committed, planned, agreed, that the PO has become a bit irrelevant on these one or two fundamental projects.

Okay.

And you've been talking about 8 to 12 EUV tools in 2014.

Can you help us with the price they will be sold or they might be sold?

A bit early, because these tools will have a huge amount of options, because you're talking about first production tools.

And in this situation, usually, the customers will be very customizing the content of them.

It -- I would say EUR80 million to EUR100 million could probably be an acceptable range.

But unfortunately I won't be more precise at this moment.

And it could vary from customer to customer.

Okay.

Okay, thank you very much.

Gareth Jenkins.

Please state your company name followed by your question.

Yes, it's Gareth Jenkins, UBS.

So just a quick one on EUV again, coming back to the 8 to 12 tools for next year.

I just wondered if you've seen any changes in terms of logic adoption of the EUV?

Or whether we should still expect the vast majority to be DRAM?

And if so, what's changed for logic?

Thank you.

No, in fact, we may have started to hint to last quarter, if I remember that DRAM was also keen with EUV because there is no choice.

They absolutely have to have at least one layer on EUV just as a feasibility statement, so DRAM is always out.

However, logic has, over the past, I would say, DRAM in a sense that, first, they also have now a need which they have realized, which is a minimum of, as we said, five layers to seven.

They all know -- so see that they have an economic issue, so there is now another incentive to go from [57 to 17].

And they have also a huge business attraction, meaning logic, is in fact, doing good at this moment.

And they see themselves going in another node after the famous, what we call, 20 nanometer, but which will be called 1420.

That node will then be followed quickly by a node which will be called something like 1011 or something.

And that is the UV node.

And there is going to be, or at this moment, the customers believe that there is an economic wave that will drive this.

So, logic becoming hugely interested, putting most of our attention now on logic.

But DRAM will get in.

And if DRAM's business becomes more attractive, the PC business basically gets back and puts some attraction there, we could see DRAM trying to still be three, six months ahead of logic because they do have a need in it.

So, Eric, can I just follow up on that then for production on the 10 nanometer, 11 nanometer node for EUV, is that a 2015 production?

Well, so you will see the chips, the dye, the wafers in 2016.

The buildup of the capacity of the factory will happen a year before.

Great, thank you.

Sandeep Deshpande.

Please state your company name followed by your question.

JPMorgan.

Quick question for you, Eric.

Would you say at this point that you are now more confident about doing the 69 wafers per hour production tools by the middle of next year compared to, say, six months ago?

(technical difficulty)

(technical difficulty) dangerously because you -- we learn something and what you learn may take time to repair.

And the way that we've moved to 60 watts, we've been able to justify in the convert our simulation systems.

And our simulation systems and models now are converted up to 60 watts.

And in fact, it can be easily now projected towards the 105 watt; even if we cannot prove it yet, the models are proven checked at 60 watts.

So first 60 watts is nearer to 105 watts, and secondly, it is very good theoretically.

It has proven a model.

So therefore, we absolutely strengthen our point that 2014 production should now be much higher probability than ever, indeed enough that the customers are, in fact, continuing to discuss with us this order placing for 2014, which are meant to be between eight and 12 in total, which by the way, is a good number.

We don't need more.

(multiple speakers) Customers don't need more any way; that would be the normal ramp of that technology.

Thanks, Eric.

I mean, just one follow-on to that.

I mean, you're talking about the 60 watt having been shown in your labs.

Is there a substantial technical difference between a 60 watt laser and 105 watt laser, I mean, based on gas or what power has to be -- I mean power clearly is different, but from any other elements are just substantially different, which could cause a risk from moving from 60 watt to 105 watts?

Absolutely not.

The concept is now going to remain the same up to 210 watts, which is that architecture that we just talked about.

But from 60 to 210, we should not change the concept.

Thank you very much.

Mahesh Sanganeria.

Please state your company name followed by your question.

RBC Capital Markets.

Eric, question on your 20 nanometer.

You said there will be [40K to 60K] by Q1.

Can you tell us how many customers will that be who will be installing 20 nanometers?

Oh, these are complicated questions.

When now you are in the world of two customers, it's always difficult to tell you how many, but (laughter) -- how can I answer that question?

I -- the days clearly (multiple speakers) --

(multiple speakers) It's multiple customers, I would say.

(multiple speakers) So let me say I think the question is that every one of the fundamental logic players is going to buy a [number submission], okay?

But there is one or two who is going to buy more.

Because at this moment, they have more confidence that they have reached numbers of design wins in their own environment.

Okay.

And then one question for Peter.

In terms of R&D expenses and modeling for Q1, you said you're getting EUR16 million recovery in Other Income.

Can you give us some sense of what we will see that in the rest of the quarters for 2013?

About similar.

It's about the same number.

Similar?

Yes, yes.

So that -- like I also said in the introduction, the other part of the contribution, which I say one of the key contributors to the program will be in gross margin.

And that will vary upon the business volume that we have with them, because that's for the particular accounting treatment that we (technical difficulty).

So, this other income is only for two of the participants.

The other contribution will be in gross margin, which will be -- we'll clarify at the end of the quarter based on the actuals.

And would you say that that -- the gross margin contribution comes a little later in the second -- more in the second half in next year?

No.

It fully depends on the timing of when the shipments happen.

So basically, it's because that particular customer is also very large is a larger shareholder than the others.

And that means that the accounting requirements state that you have to allocate that contribution to the business that we have with them.

So if we start shipping tools, that means that those tools will get a, let's say, lift-up of the contribution that was intended to be the coverage for the R&D cost.

(technical difficulty) That's per system, you will have an extra margin.

You will -- and that is the way it is going to be accounted for.

So it means (technical difficulty) depending on the business volume that we have with that customer, we are going to, you know, recognize the accounting-wise; we're going to recognize the contribution of, you know, the customer.

Cash-wise is not going to be an issue.

Cash-wise, it's going to be matched.

So you will have, you could say, for some time, you could have a deferred accrual account, a deferred revenue account on the balance sheet, which will then be released to income in the next quarter when the shipments happen.

Okay.

Thank you very much.

Thank you.

Jagadish Iyer.

Please state your company name followed by your question.

Your line is open, sir.

Go ahead, please.

(multiple speakers) Peter, you want to go to the next one?

Next question.

The next question will then come from Mr. Simon Schafer.

Please state your company name followed by your question.

One more try, Simon?

Mr. Schafer, maybe you have muted your own line?

Let's go to the next one.

Okay.

Mr. Francois Meunier.

Please state your company name followed by your question.

Thank you for taking my question.

The first one will be regarding EUV shipments for this year.

I think the plan since last year was to ship 11 machines this year.

If I read the statement this morning, it sounds a bit shaky like it's going to be 11 at best.

Does it mean that there is some risk of delay into 2014?

And if so, if you could confirm that some other shipments -- for some of those shipments, all the revenues will not be recognized this year.

And that's my first question.

I've got a follow-up.

So, Peter will address the revenue recognition because I don't understand anything on this.

So I will address the shipment part.

(laughter) So the machine itself, as I said, has been purchase ordered on an as-is basis.

So, we will ship when they are ready -- finished, assembled, as I said in the speech.

The machines themselves, the scanners or the certainty of being available for next year is very high.

I mean, we don't see a problem achieving that.

The matching scanner to the proper source could be still a risk of execution.

Whether we ship all of them or we miss two or three or four depends in fact what we're going to choose to do with the sources we can ship them here, and test them and qualify them, but then that takes more time to ship later to the customer.

We can decide to, what we said, dropship the machine to the customer -- meaning you ship the machine, the scanner, on one hand and the laser on the other.

But then you have to recognize -- well, in fact, I'm starting to answer your question (laughter) -- the machine when they are tested at the customer.

So that may take more time.

So we are a bit less in control of what we wanted to be in the first place, as to the logistic of this.

And then Peter will have to put the economic view to this, which is the recognition of revenue, to the logistic nightmare that I just discussed.

Yes.

And it's what you know, Eric said -- a shipment for us is defined as the shipment of, you could say, the metal.

That could be in two ways.

That could be like Eric said, being integrated metal from Veldhoven from the Netherlands, or be (technical difficulty) parlayed, customers could say please save time.

Don't ship everything to Veldhoven, integrate there, test it and then take it apart and ship it to us.

So, please ship the scanners separate from the EUV source, which we will call a dropshipment of the source -- which then means that the integration and showing that it works can only happen at the customer side, which also means that the revenue recognition takes place at that moment in time.

And since we aren't completely sure as to what customers want, and what particular configuration of the source that they want, there is a risk that, from a revenue recognition point of view, some of those shipments may end up in the first quarter of 2014.

And that might happen.

So there is a bit of caution there.

So you notice that.

And as a matter of fact, in our guidance, we took account of that, so we took a conservative number for the EUV shipments to be recognized in 2013, because it will always be a possibility that while we have shipped it, that we're still in a process of putting it all together and qualifying at the customer side, which might then happen in Q1 instead of Q4.

Okay.

That's very clear.

Now I've got a question for Eric if I may, about EUV and 14 nanometer, especially because, Eric, you say that 40 nanometer and below they could be like 5 to 17 EUV layers per chip.

So basically, how much will be left for immersion?

And what would be the share in your view of EUV versus immersion and all the other 14 nanometer when it ramps?

Let me answer the real question you have.

Is it going to be growth for us when we ship EUV because immersion will still remain important?

The answer is yes.

I think we said it two or three quarters ago, that when EUV is introduced within its normal activities, there will be EUV, there will be immersion, there will be drive.

And in this environment, we think we're going to have to go to another plateau of sales.

And I said, by the way, I would give you the new plateau solid on EUV we know for sure what we ship.

But yes, it will be bigger than the plateau that we have already touched, which is the 5.5-ish.

We think we'll get to another level of sales because, yes, immersion will remain.

And when we ramp EUV, I think we also said it before in 2015, 2016 -- that EUV is definitely in 2015, 2016 is going to be significantly over the half of our sales in terms of the value.

And in 2015, it could easily be up to that half level.

So, yes, EUV shipments when we are ramping are going to be the most significant part of our sales numbers.

Okay, so immersion goes down basically, compared to --?

Well, what Eric said is that the total sales will go up.

So yes, we will cannibalize some of the leading-edge layers clearly.

But don't forget we also see that if complexity increases, which it does, the number of layers also goes up.

The total will increase.

(multiple speakers) Yes, okay, sure.

So, it's -- yes, it could be lower but not -- but it would be very strange if it were more down half as compared to where we are today.

So, it will be lower but it's going to be manageable.

And on top of that, we see the EUV sales.

Okay.

That's very clear.

Thank you very much.

Andrew Gardiner.

Please state your company name followed by your question.

It's Barclays.

I was interested if you could provide a bit more detail around the visibility being provided by some of the foundry customers at the moment.

You've clearly indicated that your expectation for revenue there to ramp in the back half of the year, given the 20 nanometer buildout.

I think looking at the way the foundries are communicating at the moment, one is being particularly vocal.

The other perhaps slightly less so, given a customer shift.

But how much of your expectations for that second half ramp are being driven by that one who is being more vocal relative to the other one?

Or are you assuming within your guidance that there is a strong ramp across those 2 leaders?

We are in our current guidance where we said at this moment, identifying about the same level of sales in the year, we are only inserting one large 20 nanometer buildup and the rest being very, very, very small.

So in other terms, our point has always been to be conservative.

So if the other players were to try to move to their natural market share versus the current one who is driving, we will sell much more units.

Understood.

Well, just a quick follow-up, Peter, regarding tax.

We've seen in the past, when your sales dropped to these kind of levels, that the tax rate increases.

I was just wondering if you can give us any sort of steer as to how we should be modeling that for 2013, in particular, the early quarters when revenue is lower?

Well, as a matter of fact, at these sales levels, you are referring to times when sales were really dire.

You talk about the first six months of 2009.

And then you could argue that the effective tax rate had no relevance at all -- which, at the levels that we're talking about in Q1, is not the case.

I think the fact that we have currently a resolution on the tax issues that were outstanding, in particularly Asia, means that we can now confidently recognize the taxes in that area on a lower level, which means that going forward for 2013, we guided you at around 12% tax.

You could take 2 percentage points also around 10% -- 9% to 10% would be a good number for the total year.

For the year as a whole?

Okay, thank you very much.

Satya Kumar.

Please state your company name followed by your question.

From Credit Suisse.

Thanks for taking my question -- a couple of them.

I guess one of the customers that have indicated this very strong ramp of 20 nanometer has also indicated that the capital spending will be heavily first-half weighted, but two-thirds of their spending in the first half.

From your commentary, it's clear that your second half that is stronger than the first half.

Could you perhaps explain why there might be a bit of a difference in terms of what happens to you, versus what they're saying for their CapEx?

We are a conservative bunch.

Maybe that's what I can tell you.

We are -- I know it's difficult to -- it may be also -- remember, we do not represent 100% of the CapEx.

And if I'm not mistaken, there is some brick and mortar into the pie, which means there is potentially upfront thing that is done before we start shipping machines.

But no, I would not say -- I would also be realistic -- we just finished, the world finished to -- not even finished -- the world is in process of finishing 28 nanometer.

It would be a bit aggressive for me to say that we now ship hundreds of millions of 20 nanometer today, that that would not be natural.

But for the customer in question, they may have to have significant CapEx to prepare the situation before the machine comes in.

Okay.

I was intrigued by one of the slides you showed on your presentation, which talked about the integrated metrology that they're used on.

Could you talk a little bit about what they're doing to -- if you could quantify, that would be great -- to the ASPs of the emerging systems as we go from 28 to 20 nanometers?

And from a process flow standpoint is -- what exactly is that metrology doing?

Is it replacing an optical 3D metrology that your company would have bought from another equipment supplier?

Or is it in addition to other equipment that the chip companies are going to be buying for metrology?

Okay, so let me answer the first question first.

And Peter will discuss the margin impact of this type of product.

No, we would not say that it directly replaces metrology.

Metrology -- the segment metrology, as it is known in the industry, is diagnostic metrology segment.

It's every day or so, you take a wafer, you make an extremely deep analysis.

You understand what the recipes impact are.

You change the recipe.

So it's an involved diagnostic, long lead-time, deep, involved process capability.

And we don't have that.

What we insert are machines which are able on-the-fly to correct a process failure by resetting the numbers of knobs into a scanner.

We call it metrology because indeed you have to measure and correct, but the real term should have been a closed loop system.

Before, they didn't need to use closed loop, because the machines did not need to be corrected.

But now the processes are so tight, the window of processes are tight, that they need to be expanded and they need to be controlled.

And therefore, there is a new activity that is called closed loop.

And we are happy to be a big player into this.

And it is true that we've done, in the past year or so, a very small amount of business in that segment.

In total, I would say, with up to 200 million or so in total business of this type of activity -- 200 million, 220 million.

Now, it's real.

So we know the terms are -- there is going to be significant impact as we go.

And when the business goes from 20 nanometers, it's already big.

And at 14, it's going to be even worse or better for us.

So we will go north of 400 million in due time.

So this is real additional business, which indeed have an impact in margins.

So, Peter?

Yes.

In -- when you talk about the ASPs and not so much margin, because there are higher margin products.

When you look at the particular slide that you are referring to, where you see these two machines, the scanner ran alongside and what we call the YieldStar metrology tool on the other side, you have to remember that is not one single tool that we are selling.

It is hardware across a lot of software.

So it's a combination of products that we are selling.

But if you would have a fully loaded 28 nanometer machine compared to a fully loaded 20 nanometer machine, which would include this YieldStar and the feedback loops that Eric gave, you would look at, depending on the customer and the application, a difference in list price of between EUR7 million and EUR10 million between the two.

Now, of course, then it depends on the customer contact that you have, the volumes that they are taking, what discounts that they're getting, but that is the kind of the difference that you need to think of.

Okay, thank you very much.

Weston Twigg.

Please state your company name followed by your question.

Just a quick follow-on to the last line of questioning.

On the holistic lithography and the upgrades, and the revenue contribution, I think you said that you expect to stay at this EUR257 million level in terms of service and field options.

But wouldn't it make sense that that would increase as you get this 20 nanometer ramp coming into the back half of the year?

In other words, why would you expect that line of revenue to be flat?

Or is it -- does it just come into the (technical difficulty) sales revenue?

Well, you have to look at it in the context of that the fourth quarter was the highest ever.

Yes?

So I mean, this did indeed in the fourth quarter show some incidental extra revenue.

But where it would be when you take the average from 2011 and 2012, we are significantly lower.

Now I think that is -- that average will go up.

So that is basically what we are saying.

And yes, we are conservative in a sense, because like I said earlier, EUR7 million to EUR10 million addition to the sales price from a list price point of view, is also something that customers need to get their hands around, and basically need to realize that they need it.

We think that they need it, but we feel we can, on average, keep the [EUR250 million] level per quarter, which is a record.

Okay.

And then just staying on these features, is there something that your competitor, your main competitor in this space, has available?

Or is this entirely different than what they can offer?

Well, as far as we know, that's not available.

Okay.

And then just finally, when you talked about the 11 EUV tools that are expected to ship this year, and the contribution of EUR700 million, you also mentioned that perhaps you've slipped some of that EUR700 million into 2014.

Can you tell us how much you're expecting for 2013 and how much for 2014?

Good question.

We have and what we told you as sales levels for 2013, we took a conservative amount, about [EUR215 million] off the EUR700 million to come to the sales levels 2013 similar to 2012.

Very helpful.

Thank you so much.

Jagadish Iyer.

Please state your company name followed by your question.

Yes, Piper Jaffray.

Thanks for taking my question.

Sorry about the previous one.

Two questions, Eric.

So first, if you look at your sales guidance for 2013, and given your foundry war, which is ongoing and the capital intensity that is happening, given that there's going to be at least 1.7 times more intensive to the prior notes, your guidance seems to be a little bit kind of soft.

Is that a fair assessment that you're being very conservative here?

Can you talk about the puts and takes?

And then I have a follow-up.

Yes.

Well, our philosophy and style is not to bet and have an opinion, and then say it's a gutsy opinion or it's not.

We're trying to guide you only on factual data that we have.

And the factual data is, we have one segment and a half, which one segment is, you've all guessed it, [in the] 20 nanometer leading company.

And then you have a percentage of our 11 machines.

And Peter was nice enough to tell you exactly which percentage he talked about.

So you take this big account at 20 nanometer with our own understanding of how many units you really need to do to do these [40K] wafers.

You add this X percent of the EUV machine, you add the standard business that we have, and you come up with about the same sales level.

So that's what we guided.

When we saw that number, we said we have to guide you because it's a true statement.

It's a true number.

So you can translate my point by saying there is indeed upside opportunities.

The first upside opportunity is that the other logic people will consider this acceptable to leave only one and one of them take over the majority share of 20 nanometer so early in the game.

So, that's an upside.

The second upside, which we mentioned in the press release, is that if the PC business does pick up a bit, due to the new form factors, you will see immediately a reaction in every people who do microprocessors, you know, who will benefit from that, and memory.

DRAM will start benefiting.

But then you have -- means you have a multiplier in the NAND business for the people doing and successful at solid-state drive.

Because solid-state drive will be multiplied by the PC new form factor performance.

So again, that is not in the numbers and this is an opportunity.

Okay.

So that will be if you are -- want to hear from me the upside.

The downside is that we are in a business now which is extremely granular.

That is if we make a mistake in judgment, it's a big one.

And it could be a big one because like a fab is late by 3 to 6 months just because the concrete did not pour, or the contractor to the foundry business with whom we are working is having nightmares and et cetera, something bad has happened, and in which case, we have a negative to the guidance.

So again, we don't want to put any judgment to the upside and the downside.

But I hope by doing this very candid description of what is in our numbers and how we built it up, you, yourself, in fact, should make the number for the year.

That's very helpful, Eric.

So, as a follow-up, I just want to get an understanding.

In terms of the DRAM customers, what is the sense of urgency are you seeing for the DRAM guys to be looking at EUV, given that the industry is somewhat consolidated and it's in their best interest to see pricing continue to get better for their own profitability?

What is the sense that you see that in terms of them trying to accelerate EUV at this point of time?

Thanks.

I guess the DRAM business, or the memory business because it's more and more merging somehow in there, is looking for a differentiated activity.

So they have results probably -- as you call this, the consolidation.

I guess that they have numbers of players which have reached some kind of an equilibrium.

So none of them would want to continue losing money.

Why they continue losing money and the prices may take time to get up, because, by the way, the DRAM prices are getting up in the last three or four weeks.

So there is an upside coming on.

Something is happening at this moment.

The reason why they, however, had another bad six months or so is because the PC business, in fact, was surprisingly weak.

If you remember, we started the year last year where everybody was thinking that the PC business would grow by 5% or 10%, and then it happened to be flat or shrinking.

So that created a problem that they are trying to resolve.

So, what is happening now is a -- one will not continue driving the business with the same method which is overinvesting, so that they try to take the other guy out of business.

I don't think they will do this any longer.

But I start seeing more interest to differentiate themselves with the way their chips, which also include a bit of process.

This is very visible in the NAND business.

You have noticed that in the NAND business they have gone now to trying to differentiate on the design of their chips, on the numbers of bits per sale, on the software drivers, et cetera.

And you can start seeing this in the DRAM.

So if you have a DRAM idea on the EUV shrink capability is not only cost but it's also an enabler of something, like power consumption, speed of access, et cetera, then you could show some huge interest.

And we started to see some of that, but I must say I don't have specific to tell you that you should consider.

I mean I have -- there is something in there, I don't know.

I see questions but which leads to this idea.

Thank you so much for the comment.

Really helpful.

I think we, ladies and gentlemen, have time for one more quick question.

We'll try to squeeze another one in.

If you didn't get through and you have remaining questions, of course, feel free to call the Investor Relations Department, and we'd be glad to get back to you as soon as we possibly can.

So, operator, can we have the last call, please, or last question?

Of course, Mr. DeYoung.

The last question comes from Mr. Lee Simpson.

Please state your company name followed by your question.

It's Lee Simpson at Jefferies.

I really just wanted to try and ask a question around DFM risks.

I mean if you look to a lot of the EDA suppliers, they're talking about increasing DFM risks around advanced newer development, such as at 20 nanometers, 22 nanometers and more so at 14 nanometers.

Things increasing like DRC rules and the complication with fill flow for new features like finfeds.

I suppose history tells you there's always threats to the speed of design progression for new notes, but I wonder if you could shed some light on what might be different this time, and how this, if at all, really transfers to slow down in development of chips at each node, and whether or not that has an onward impact to the ramp of the 20 nanometer node up to, let's say, 320,000 wafer starts per month similar to 28 nanometer node?

Well, you put a tough responsibility on me.

Because usually when you close a call, you want to close on good news.

And you come up with a possibility for me to say something negative.

(laughter) No, I try my best.

The negative part of your question is really at the end of the day, it could create an inflation of cost.

You mentioned issues, design manufacturing issues, which require, in fact, investment.

So there is indeed a pressure on cost on all of us, our customers, but all of us are trying to solve that problem by bringing solutions to the market, which do not add an inflation of cost.

So that clearly is a pressure.

But let's conclude with the positive news.

We see this as a huge business opportunity.

The discussion we just had on this metrology closed loop is exactly your point.

It has triggered the possibility for us to go from EUR100 million to EUR200 million to EUR400 million potential sales of these high-valued product.

Secondly, at the moment when the wall is too big, EUV comes up and EUV does reduce the need for these competitive situations.

So, ASML is usually well-positioned to resolve the complexity that you just mentioned.

So I mean, this goes back to an earlier question I think you got that holistic lithography should probably increase, bearing aside that perhaps Q4 was the high tide so far in that sales level.

Absolutely.

Yes, okay.

And Eric had talked about the DFM solution, the holistic litho products have two elements if you look at our presentation.

One is to widen the process control window in your preparation for manufacturing, which really addresses your design for manufacturing issues.

But in rolling this out, we also discovered a second element, which is basically the control of your process parameters, while you are in manufacturing.

So, it has two elements to it.

And I think that is -- the latter is more a, let's say, a finding of the last few years, while the first part, you know, needing to find a solution for the design for manufacturing issues, that was really why we did the Brion acquisition.

But it evolved into the latter with the YieldStar product and the closed loops.

Great.

Gentlemen, thanks so much.

Good.

You bet you.

Thank you.

On behalf of the Board of management here at ASML, I'd like to thank you for joining us today on the call.

And operator, if you'd formally conclude the call for me, we would appreciate it, thanks.

Ladies and gentlemen, this concludes the ASML 2012 fourth-quarter and annual results conference call.

Thank you for participating.

You may now disconnect.