ASML Holding Management Present at Deutsche Bank dbAccess 2012 Technology Conference (Transcript)

Sep.13.12 | About: ASML Holding (ASML)


Deutsche Bank dbAccess 2012 Technology Conference

September 11, 2012 7:00 PM ET


Craig DeYoung – VP, IR


Kai Korschelt – Deutsche Bank

Kai Korschelt

Okay. Yeah, welcome everyone. My name is Kai Korschelt. I’m the European Semiconductor Analyst at Deutsche Bank. It’s my pleasure to have Craig DeYoung here, like last year, present to us, and I think Craig is going to give about 10 minutes – 5, 10 minutes presentation and then we’ll do...

Craig DeYoung

10 or 15.

Kai Korschelt

Yeah, okay, 10 or 15.

Craig DeYoung

No, I’ll try to go through it quickly. Part of this presentation is anticipation of your questions as well. So hopefully I’ll get some of those covered, and then Kai and I will talk a little bit afterwards. So appreciate your coming this afternoon. I know you would rather be enjoying the Nevada sunshine, but since there isn’t any.

So here is the obligatory safe harbor statement. Just everything that I speak about of course will be covered by this. So I want to do a brief business summary and I’ll talk to you – give you an update on our co-investment program, which I think all of you are probably familiar with, at least in part; talk about the business environment that we are in today, talk a little bit about our strategy on the product side and then just remind you of our outlook.

So in terms of the business summary, this is a repeat of what you’ve already seen, but I thought it worthwhile to remind you that we did about EUR1.2 billion in sales in the second quarter, kind of a typical quarter relative to the last six or seven or eight. So you’ll see that in a bit. 43% gross margin, operating margin of 27%. We booked just under EUR1 billion in systems and that left us with a backlog of EUR1.5 billion coming out of the quarter. And we generated in the first half about EUR470 million, of which a majority returned to you through a combination of buybacks and dividends in the first half.

So here is our sales over the course of the last few years. If we do as I guided, which I tried to indicate the range of guidance, because the second half we guided EUR2.2 billion to EUR2.4 billion in revenue. So you see that if we achieve that, which I believe we will, it will be our second best year on record. So it’s still quite exciting for us.

System sales broke down this way, just still a lot of systems to the foundry, about 61% of our shipments in the quarter supporting ramp of the 28-nanometer node, not so much to the NAND and the DRAM guys, nor the IDMs, but we’ll see that change.

If we look at our backlog, you see growth in the IDM space. Foundries reduced a bit and we expected that second half deliveries to the foundries would be less than the first. There were a confluence of events in the first quarter that drove foundry quite aggressively. The challenges of meeting demand in a relatively low-yield environment and in a state of heightened competition between all the – at least the four foundry players created kind of a fervor in ordering and deliveries in the quarter. So we see that moderate a bit in the second half. Again, that’s what we anticipated would happen.

We took a number of orders in the DRAM and the – sorry, in the NAND space. I wouldn’t be confused or misled by that. Just by way of example, we took 10 orders in the NAND space, half of which were emerging, the other half were KrF tools, which suggest an addition of capacity in an environment where you wouldn’t think capacity was being added. But in fact that was over three different customers.

So if you take, that’s one-and-a-half tools or so each – of each tied to each customer. So it doesn’t represent much of a capacity addition. So again, I wouldn’t want that to be misunderstood. It’s probably more just about balancing out fabs and adding some capacity in the leading-edge nodes, et cetera, to meet the demand. I’ll talk about what that will do for the different memory sectors in terms of bit supply opportunity in a moment.

Regards the co-investment program, everybody is aware that we’ve raised almost EUR1.4 billion in R&D from three of our largest customers, Intel, TSMC and Samsung. Those will be used for two things in the main – well, three: current EUV, next-generation EUV and the development of 450 millimeter. We’re obliged now to develop in the 2015 timeframe 450-millimeter capable tools on three different wavelengths of light, so that at least one customer can start to debug 450 processors out in a couple of years from now.

That will result in minority equity stakes by the three customers actually totaling 23% in aggregate. The current status is as of Friday, we got an authorization from our shareholders to issue additional shares. I won’t talk about the mechanics of this and waste our time, if you will, here. All of that’s available on our website, the timing, the mechanics and everything. So I’d refer you there. If there is any questions after that, then I’d be glad to answer them.

We will return all the proceeds from the issuance. We’ll have to issue shares, sell them to the equity investors, our customer investors, and then we’ll take that money and return it to you through what might seem a light bit complicated mechanism we call it a synthetic share buyback where we return capital to you and then do a reverse split. Those numbers, again, are all defined in the timing of that on our website, if you’d like to look at that.

So the current business environment, so we are going to ship based on our current backlog and our – and the forecasted – forecast from our customers enough equipment at the current nodes for the DRAM industry to create supply at about 40% bit growth. So just so, that’s kind of a frame of reference.

You might be aware that we have models that can, I think, pretty accurately, based on history, determine the amount of capacity that we’re adding by way of shipment of tools, given all the different wafers installed at the different nodes within our customer base. So we anticipate again about 40% bit growth capability or possibility in equipment we’re shipping. I don’t think the forecast of late are that – quite that high. So we’re – we’ll probably remain through the balance of the year at least in an oversupply situation.

In the NAND sector, this year we’re going to be able to – our customers are going to be able to grow vis-à-vis the tools we’re shipping bits to about 50% through shrink alone. By the way, the 40%, as mentioned here from DRAM, is just through shrink. So the NAND guys, about 50% through shrink and an additional 30% based again on the shipments that we’ve shipped – the tools we’ve shipped and the forecasted shipments, 30% in additional big growth through capacity additions.

Now bit demand, I think, if I’m not mistaken, is trending at about that 80% level. I do believe that the Q4 expectation is less than that. So the forecasts for the end of the year I think are about 65% growth. So there will be, again, apparently if these numbers are right, a bit of oversupply in the NAND space.

Just to speak to next year a bit, the current calculations for next year indicate that through shrink alone, DRAM will be able to grow bits by about 30%. So there’ll be a very nominal amount of investment required to grow bits to 30%. I think the current forecasts are for about 25%. So it appears today that there could be adequate capacity for next year with what I would consider a minimal amount of DRAM spend.

In the NAND sector, bits can be grown by about – through shrink alone by about 30% to 50%. That’s quite a range, I understand. But depending on how much 3-bit-per-cell production there is, that would make the difference between the 30% and 50%. And so that’s kind of uncertain to us and probably a bit even to our customers at this point in time.

Anything that’s above 30% then in DRAM and nominally 30% to 50% in NAND would require capacity additions or additional investments. And there is no clarity on our part or our customers’ part regards what the real demand is going to be for next year. It just gives you a couple or two or three frames of reference.

We do expect foundry spend to continue at least through the middle of the year, as they continue to ramp the 28-nanometer node to what would be considered a mature level of installed capacity based on our view of the 45-nanometer and 65-nanometer node at about 300,000 wafer starts. So again, based on our backlog today and forecasted orders for the next couple of quarters, we expect to ship kind of continuously in a pretty linear fashion between now and then to the 300,000 wafer starts.

And we then expect also public statements by TSMC, at the very least, to indicate that they’ll start to ramp – not ramp. They’ll start to bring in equipment to build capacity for – at the end of 2013 for – in preparation for 2014 real ramp of the 20-nanometer node. The question today is about how much capacity they’re going to add. I haven’t heard them publicly say. Certainly they haven’t even in private indicated to us. So there is again a level of uncertainty about the amount.

I would guess that the other foundry, at least one or more of the other foundry players would do the same. But again, the level of capacity additions in this node are unknown, but we do know it’s going to be quite expensive for these guys. So the amount of equipment, litho equipment required will be anywhere between 1.5 and 2 times, at least the critical layer tools to support the 20- nanometer node versus the previous – prior and current 28-nanometer node.

So people say where are we in this cycle? I don’t know where we’re on this cycle. This is a what cycle, which cycle. These are – show our shipments in time. You can see that today we’re shipping a lot of foundry and very little memory. And I would anticipate again, based on what I just said that this will continue for a while at exactly the same or exactly what levels, we’re uncertain, but certainly foundry will continue to spend pretty aggressively.

And I don’t know that the DRAM and the NAND guys can spend much less than they’re spending today, but certainly they could continue based on what the real demand is. And we just don’t know. We have very little visibility of next year outside the foundry space, which I just discussed in terms of the 28-nanometer capacity.

So – but it’s interesting each of these guys have their own, what we call, a heartbeat now, where five-plus years ago, the whole industry kind of moved in sync every couple of years and everybody would do kind of the same thing at technology node transition. So you can see here that that’s not happening anymore.

So in terms of our technology or product strategy, so we’re left to support two – and we’ve talked about this on the last two or three quarters or more. Both advanced emerging systems in support of this very high-cost multi-pass patterning techniques that will necessarily be employed in logic for sure at the 20-nanometer node.

So actually one of the investors – the co-investors has asked us to focus an amount of their R&D and their incremental R&D spend on just this, on improving the throughput and the overlay capability of our immersion tools to try to: one, would throughput mitigate the cost, because we’re going to have to pass the wafer multiple times through the system and the increase in throughput couldn’t help mitigate the cost.

And maybe as importantly, we need to improve the alignment capability, image placement capability, because as you’re bringing two images back together and the wafer is going out and getting etched and other masks are coming in, everything has to come in just exactly correct relative to the other. So we have to place that wafer right back in the exact same position. We have to bring the new reticle right back in the perfect place in order to allow these guys to yield in these multi-pass patterning techniques.

So there is at least one customer that’s encouraging us and – to again improve the performance of these tools, at the same time, working on the current EUV systems in order to – in the longer term address the ever increasing cost of imaging in the logic area.

So those are our two focuses. We’re shipping 230 wafer per hour immersion tools today, and we will be – did I go the wrong way, sorry – we will be shipping – oh, I got these mixed up, sorry. So we’ll be shipping 250 wafer per hour tools next year with, again, ever in the – ever improving overlay requirements.

By the way, all of these are upgradeable in the field. So as you see, our service, the current and past immersion tools of a certain vintage are upgradeable in the field. So as you see, our service revenue going up toward EUR225 million, EUR230 million level. This is in part responsible for that. So our service revenue should continue at least to EUR220 million level for the foreseeable future, maybe even a little bit higher, as again the installed base gets upgraded a bit, to EUR230 million and eventually to EUR250 million as we go forward in time.

Again, to go backwards here – sorry, the 3300s, just to remind you, we’re working vigorously and hard and long on trying to get these tools out into the field with acceptable throughput capability. We’ll ship 11 tools next year. These are essentially all process development tools. Six of them we already have in the field, as you’re aware, but these are higher numerical aperture and more capable imaging and overlay systems than the prior version. And in the future, they will be much more productive as well.

Also, based on the performance of the existing tools and the promise of the 3300s, we’ve received a commitment for four orders for delivery in 2014 for implementation at the 20-nanometer node at a major DRAM manufacturer. Our commitment to them is to deliver 70 wafers per hour in Q1 of 2014 to support that production implementation.

So last slide, just to remind you of our outlook, unchanged, EUR1.2 billion roughly in this quarter, gross margins 43%. We will pay – be paying – spending our kind of normal R&D at about EUR145 million.

Just to comment about what the incremental spend in R&D may be next year, I can’t tell you what exactly what it’s going to be, but what I will tell you is that now that the shareholders have authorized the issuance of the share in the ownership by our customer investors, we’ll be sitting down with them starting essentially today through the balance of the year, identifying the programs, identifying the targets of the programs that we’ll be working with on each of these guys, milestones, et cetera. Based on our resourcing capability, et cetera, we’ll come and identify a spend amount that will support the programs that we’re able to support throughout 2013.

And then that will become – there’ll be an associated incremental spend against which each of our partner customers will be contributing, and we’ll identify that as soon as it’s defined. There is no reason, of course, that we’ll keep that from you, but it’s just not known yet, because we have to, again, negotiate, discuss and decide along with our customer partner. So next year, of course, the R&D will go up.

I won’t talk to you about how the accounting will work on this, but I am available at any time to discuss it. It also is available on our website, the detailed accounting. It’s a bit complicated, especially in Intel’s case, because they are considered a significant related party with a share ownership above 10%, and we’re involved in a multi-element transaction, and you bring those two together and it makes for some kind of ugly accounting. It’s all U.S. GAAP-driven.

So – and again, just to remind you, we are on track for our H2 sales between EUR2.2 billion and EUR2.4 billion. I’m sure we’ll take a question about Intel’s comments and pushing and pulls of foundry, et cetera, et cetera. I’ll be glad to answer them. But in short, it does not affect to date what we envisioned when we started this quarter. So EUR2.2 billion, EUR2.4 billion we’re still on track for.

So with that, I’ll take a question or two.

Kai Korschelt

Okay, great. Thanks, Craig. Maybe just on the first question on that guidance, particularly for the fourth quarter. I think at the – on the Q2 call, Eric was indicating that there may be potential upside to the fourth quarter, I think, the implied guidance of EUR1 billion to EUR1.2 billion in revenues based on the foundry – now maybe a pickup in foundry activity. Just wondering what the likelihood of that scenario is now. And then maybe another one on Intel, obviously the inestimable one, if you’ve seen any change in their behavior.

Craig DeYoung

I actually think – well, there is two things that could drive foundry spend higher in whatever timeframe. And that is there are some people that suggest that the 300 million wafer starts could be higher, so in time, that the demand, based on real demand, die size implications, et cetera, at least in the short term, could drive installed capacity above the estimated today 300,000 wafer starts per month level.

I think actually Eric was a little bit more bullish – or set an expectation – not more bullish, but he set an expectation that we could see some NAND orders in the latter part of the year. That’s what I recall. And I haven’t talked to him recently about his thoughts on that, but I’d think they’re probably tempered a bit.

I would – I still – I don’t want to speak for him, but I still think he would probably believe that in time that will happen, but maybe not at the end of this year. He did leave room open I think in the call that that could happen, and if not late this year, early next year. But I think we’ll have to wait and see on that front.

The question about Intel, they never like to us talk specifically about them, but somehow when they talk about reuse of tools and stuff, it kind of puts me in a position where I have to talk a little bit about it, so – at some risk. No, I mean, I don’t, I shouldn’t probably respond to, but I would just tell you that – Intel, if you would have looked at their website about five years ago, you would have seen a whole presentation on cascading of equipment. So that’s been their history to reuse equipment, and again their term is cascading and so – and everybody does this, by the way. I mean they’re – nobody throws this equipment away. It stays in its initial fab for at least 10 years in some applications.

So generally customers do that, specifically, yeah, I don’t know. We – most of what I better stop. But I – so let me just tell you one thing that I think is fair and you can see this is most of what we’re going to ship to them through the balance of the year is in preparation for the next node, so whether a 14-nanometer node. And if you look at the history of spend on equipment in general, but certainly on lithography, that’s what – that’s really the only thing I am aware of. Intel usually spends more in the first year of the ramp than their second year of the ramp. So if you consider that – and again, this isn’t speaking to their comments about 2012 CapEx, because I don’t know what that means. I can’t talk to that.

Intel usually spends more in the first year of the ramp than their second year of the ramp. So if you consider that, and again, this isn’t speaking to their comments about 2012 CapEx, because I don’t know what that means, I can’t talk to that. But in general, I would expect that next year, because it’s the first year of the new node, that it would be higher somehow, I won’t say significantly or I won’t put an adjective to it, but it will be higher than this year if history repeats itself.

Because again it’s the first year of the node, they want to get capacity installed. This is – this year is the second year of their 22-nanometer node. So again, if you look at their spend history, you’ll be able to see that it suggests that the first year of the node (inaudible) – next year is going to be a higher spend year for them than the previous node.

Kai Korschelt

Next time I ask – I’ll make sure talk about large MPU customers...

Craig DeYoung

Yeah, that’d be helpful.

Kai Korschelt

Sorry. Okay. Then maybe just on foundry I think as you said one of the positive long term drivers for ASML clearly is growing lithography content particularly in foundries, which I think is the only customer segment which is still doing single patterning at 28 nanometers and then moved to double patterning and 20 nanometers will materially increase, the lithography content.

But it also seems to be increasing the cost per wafer so maybe there’s still some pricing discussions there the foundries have to figure out. I’m just wondering in terms of the linearity of the 20-nanometer ramp, as you mentioned, there will be some activity next year, but will it be more sort of pilot type capacity or could we potentially think about a volume ramp already maybe in the second half next year?

Craig DeYoung

For the 20-nanometer. Yeah, yeah.

Kai Korschelt

For 20-nanometer just like we saw I think in Q4 of last year already pick up in foundry activity maybe in preparation for the 28-nanometer. So there seems to be quite a long lead time for lithography.

Craig DeYoung

Yeah. So TSMC has again said publicly that actually at the end of this year they’ll be doing what they call risk launch and I think that roughly translates into product sampling to various customers.

So they already have the equipment set, which is similar to the – it’s the same tool set as the 28-nanometer node, but they just need more immersion tools to be doing support to double patterning. But they can do the – they’ve done the development activity apparently and they can do low level of sampling in what they have today.

So next years would be about real capacity ramps. And they’ve again said publicly that they intend to bring in volumes of tools to support the real ramp in – or let’s say revenue recognized shipments in 2014. So – but again, I don’t know what – to what level they’re going to do it, whether it’s 10,000 or 20,000 or 40,000 wafer starts and that can make a big difference.

Because the difference between – well, not them, but in foundry generally, a 20-nanometer node, we estimate the litho purchases for all the equipment in a 45,000 wafer start fab is about just a tad over a EUR1 billion, just for the litho equipment. So depending on how much they bring in, you obviously can see that it could be quite a significant spend.

And I don’t know what to – to what level they’re going to bring it in, but I’m fairly sure based again on history that the ramp will happen. The initiation of the ramp will happen. They won’t delay a ramp of a node of hardly if ever, ever seen that done. So they want to get a critical amount of capacity in there.

And if you think about – if we think together actually, about what they – of their experience of the 28-nanometer node collectively and the yield issues et cetera, I would – and the fact that they got caught a little bit short or a lot short on supply, I would anticipate they might be even a little bit more aggressive in the beginning in terms of the tool intake, especially if they have the process defined, because the last thing they want to, the foundries collectively want to have happen again, is to get short on capacity.

So they have to anticipate yield issues. They have to try to predict real demand. And in all of that – again, I don’t know exactly what die sizes are going to look like, but they have to also consider what their die size is.

But I think they’re going to err on the safe side as they start to ramp this next node and make sure they have ample capacity to cover ever, real demand, loss of yield or whatever it is. So I would – I don’t know what I’m trying to suggest in terms of a level of capacity installation, but I think they’re going to err on the safe side and put more as opposed to less in. It’s my guess.

Kai Korschelt

Any question from the audience?

Craig DeYoung

There’s one here.

Question-and-Answer Session

Unidentified Analyst


Craig DeYoung

To current shareholders, to you guys?

Unidentified Analyst


Craig DeYoung

So what will happen is basically we’re going to buying a share, our new investors, our co-investors with all of you guys, are going to be buying the share from you. We’re going to be doing it actually, we’re going to be executing the buy and it’s going to be actually a capital repayment.

So we’ve agreed with all three customers, customer investors, to a share price of EUR39.91 per share, price per share. So the way we’ll – so we’ll issue shares, they will pay us EUR39.91 per share. And by the way, this was this 20-day average at the time that Intel acted. We felt it responsible to offer that same pricing for 45 days to the other partners. We started to talk to them about this proposal at the same time. The fact that they acted in a different timeframe than the others, we didn’t think it was wise to penalize them now for that. You might think that’s not wise, but it we thought it was. So we did.

And so they’ve all executed their – the contract with us and so they will all buy. Up to 23%, by the way. Samsung decided instead of taking 5%, which they had available, they only took 3%. So 23% shares will be issued, EUR39.91 per share will be paid to us, and then we’ll take that EUR39.91 times X number of shares and we will pay that back to you in a capital repayment.

So, a repayment of capital per share. And then at the end after we pay you, then we’ll do a reverse split on the tail end. So it’ll reduce the share count then back down to the original share count of today. So it will be non-dilutive in the total overall share count.

Unidentified Analyst


Craig DeYoung

No, no. Sorry, sorry. That’s the equity. The real deal presented to them was, we decided we’d make 25% of the company available to customer partners. The program said, if you buy EUR1 of equity from us on top of that you give us EUR0.30 of R&D. So that’s the formula.

So in the end actually, based on the R&D programs that each of them wanted to run, had a specific cost defined and therefore they had to make a contribution and then that determined actually their ownership position based on that formula. So that’s the formula that everybody worked to. So we actually get EUR0.33 for every EUR1 that’s invested. So the EUR1 investment goes back, the EUR0.30 not of that but in addition to that, is the R&D investment, which equals EUR1.38 billion.

And that will be put to use as they prefer. So publicly they’ve each talked about in press releases what their interests are; I alluded to them in part of my presentation. Intel’s primary interest in the beginning is in 450 millimeter, development of 450 millimeter; the other guy’s primary interests are a shorter near term EUV.

So again, those are the programs, how we get – how we satisfy them. And by the way, the EUR1.38 billion will be spent over a five-year period; the investment’s over a five-year period. So how we satisfy them each year will be determined at the end of a given year and the incremental spend above our normal, let’s call it, nominal 150 per quarter will be determined.

The account – the part I didn’t want to get into was the accounting part, how it’s accounted for. I’d be glad to if there’s time, but I wasn’t trying to avoid it. It’s just kind of tedious.

Yeah, Ian?

Unidentified Analyst

Could you just speculate, it isn’t so much an ASML question, but I asked you this question before. If they are making – clearly the – you came to this decision because you looked at the last few nodes and said, this is too expensive for us to carry on our own and we want to share it with our customers. But if they are making this investment in this R&D grant to you, what about the other players that you need for the whole chain to work? I mean, how are they going to make this terrific investment to get to 450 millimeter or whatever without a similar type of R&D contribution from their customers?

Craig DeYoung

And by the way, we could have afforded this. So it wasn’t about affordability. We could have afforded it. We wouldn’t have done it according to the timelines that they wanted and so that’s how this helps us. But I think one of the other primary interests in the co-investment on some of these programs is there’s risk involved in these. So we sat and we talked about a five-year activity to develop lithography for at least five-plus years out.

And so there is risk, there is risk in 450 millimeter. There’s no consensus today on when and if and who is going to use 450 millimeter. If there would have been consensus on 450 millimeter, we probably would have acted somehow on our own. But we begged the industry to find a consensus about what was required, when and how what was going to be done, and even today there’s no consensus amongst the three guys that we’re talking about, let alone the whole industry.

So there is some – there is risk. There’s real risk. So it’s going to cost us about EUR1 billion to develop 450 millimeter systems. So without some commitment of the industry, it would be – we didn’t think very wise for us to proceed with development to that level.

So the fact that one of our partners now, financial as well as technical partners, wants to go ahead with it and to burden – carry some of that risk burden, that’s what allowed us to go forward with it.

The other part – the other question?

Unidentified Analyst

I wasn’t being critical (inaudible). I’m just wondering...

Craig DeYoung

Did I sound defensive? Okay.

Unidentified Analyst

(inaudible) how will the other players that you...

Craig DeYoung


Unidentified Analyst

Need to make the whole thing work...?

Craig DeYoung

Yeah. You know, oddly enough – yeah, oddly enough a couple of those that I won’t name are already in – started 450 millimeter development activity, committed the number of dollars strangely enough without a commitment from the industry in terms of timing. So I don’t know exactly. You’ll have to ask them. I don’t know what the costs are, but I think they are prepared, at least at a certain level, to start work on their own for whatever reason.

I suppose also if you look at it, there’s – this provides opportunity. There’s certain times in – there are certain places in time that allows the equipment suppliers opportunity. Usually a node transition is one of those, certainly in lithography, but probably similarly in the other equipment suppliers’ areas, whether it be etch or implant or whatever else, I assume, and wafer size changes, it provides opportunity for somebody.

So I think probably people will identify what their real opportunity is and be willing to invest against that opportunity, but exactly what they have in mind and how they are going to do it, I’m not sure. But at least they know now when the litho tools are going to be there, and without the litho tools there’s no real sense in having it because you can’t debug a whole fab without all the litho tools, and Intel realized that and that’s why they decided to invest the way they did.

Unidentified Analyst

I can’t understand the – accelerating 450 millimeter development, I’m not sure how you can accelerate EUV, because you’ve been working on it for quite a while. It doesn’t seem like throwing more man hours is going to do the job, so I’m just wondering what exactly will you be doing in EUV to accelerate that.

Craig DeYoung

I can’t answer that question right now, because I don’t know. I really honestly don’t know, but I guarantee you that the R&D guys have plenty of ways to effect some positive change in that area. And we will be – when we have an agreement and some milestones for next year with each of the co-investors, I’d be happy to – it’s not something I want to keep from you for sure, it’s just at this point in time I don’t know what the plans are.

Kai Korschelt

Maybe one other question on memory. I heard you mention the sort of nasty O words a couple of times, oversupply, I think when you talked about DRAM and NAND, maybe later this year. So I’m just wondering, is that sort of your base case assumption that we’ll continue to see relatively weak sort of pricing on DRAM and NAND or do you think things stabilizing now and going into next year we should see some pickup in (inaudible)?

Craig DeYoung

To be honest with you, we really don’t know. We’re not that smart about the end demand drivers et cetera. We just are a simple equipment supplier. But the – what we’re able to do and the reason I share all these bit growth numbers at least to give us all some talking points and to give you guys a kind of a frame of reference on what could trigger spend beyond where we are today or the nominal spend to support shrink only next year, so that’s the reason we discuss it the way we do and that’s the way we view it internally.

Again, we try to be smart about what the demand drivers are and when it could change, but in the end we’re left with our customers’ indications to us and then we also kind of layer on top of that industry analysts’ forecast to try to get a range of opportunity, but that’s the best we can do. So I don’t think we – that the company overall has a view on when things could turnaround. Sorry about that.

Kai Korschelt

Yes, no, sure. Maybe one final question. I think for regulatory reasons, you’ve had to suspend – or temporarily suspend the buyback. Do you anticipate to bring that back on stream?

Craig DeYoung

Yeah, we should be able to bring that back on. It won’t be until I think late November, early December according to the timeline. We had to suspend any kind of buybacks, because you can’t, apparently, issue shares and buy back shares at the same time the way we’re doing for some strange reason. But no, so we we’re restricted there. So we suspended that.

Yeah, there’s no reason we shouldn’t pick that back up. I don’t recall off the top of my head. We still have some available tax free share buyback room. So just for those that aren’t familiar, there is kind of a crazy Dutch law, based on your dividend paying history, there’s a calculation. It’s really a funny and kind of contorted one that determines an amount of available tax free share buybacks.

The reason we have do these kind of synthetic buybacks is just to be efficient with the buybacks under the Dutch tax rules. But they do allow an amount of tax free buybacks based on your dividend paying history. So I think we maybe have – I don’t want to misspeak, but I think we might have at least about EUR300 million or something like that left.

So hopefully we’ll go ahead and start to execute that, if not late this year, when it makes sense in time. And we’ve been very disciplined and just done kind of – on every quarterly basis we tell the banks that we want to buyback X amount and just do the best you can and they’re incentivized to buy below the average et cetera. So yeah, we should reinitiate that.

And then we always have, by the way we just want to let you know we’re going to generate cash beyond our tax free buyback amount. There should be no reason we won’t do that. So our only vehicle then to buy back is to do this synthetic buyback again, which we did in 2007 and for about EUR1.2 billion.

It’s not something you want to do because it’s a little bit kind of again quirky logistically. You don’t want to do it for EUR200 million or EUR300 million you’d rather accumulate cash just to do it and make it practical to do. So if we start to accumulate, it’s just for that reason that we would anticipate doing a larger buyback using this synthetic approach. So thanks, everybody.

Kai Korschelt

Perfect. Thank you very much.

Copyright policy: All transcripts on this site are the copyright of Seeking Alpha. However, we view them as an important resource for bloggers and journalists, and are excited to contribute to the democratization of financial information on the Internet. (Until now investors have had to pay thousands of dollars in subscription fees for transcripts.) So our reproduction policy is as follows: You may quote up to 400 words of any transcript on the condition that you attribute the transcript to Seeking Alpha and either link to the original transcript or to All other use is prohibited.


If you have any additional questions about our online transcripts, please contact us at: Thank you!