Good day. Welcome to the ASM International 2015 Q4 earnings call. Today's conference is being recorded. At this time, I would like to turn the conference over to Victor Bareño. Please go ahead.
Thank you, Cleona. ASM issued its 2015 fourth quarter results last evening. For those of you who have not seen the press release, it, along with our latest investor presentation, is accessible on our website, asm.com. We remind you that this conference call may contain information relating to ASM's future business or results, in addition to historical information. These forward-looking statements involve risks and uncertainties that could cause actual results to vary materially from those expressed or implied in such statements. These include, without limitation, statements relating to revenues, margins, cost reduction programs, liquidity, break-even levels, strategies, and economic conditions. For more information on the risk factors that could affect results, please refer to the company's press releases, reports, and financial statements, which are available on our website.
All forward-looking statements are based on information as of today, February 24th, 2016, and the company assumes no obligation to update these statements. With that, I will now turn the call over to Chuck del Prado, President and CEO.
Thank you, Victor. I would like to welcome you all. Thank you all for your presence in today's call. Looking back to the year as a whole, 2015 was another year of healthy growth for ASMI. For the third consecutive year, we grew our revenue by double digits and outperformed the broader equipment industry. We continued to benefit from strong momentum in ALD, which is more and more a key technology for our customers in the scaling of semiconductor devices. Profitability further improved, even with a significant increase in R&D investments. Cash generation was again solid and allowed us to start another share buyback program. On top of that, we announced our proposal to increase the dividend to EUR 0.70 per share.
The success in 2015 has been made possible by the continued dedication and hard work of our employees, for which I'd like to thank all of them. Let's now review our fourth quarter and full year 2015 financial results, starting with the fourth quarter. Net sales in the fourth quarter came in at EUR 145 million, down 11% from the third quarter, but up 16% compared to the fourth quarter of 2014. As such, net sales were at the top end of the range that we had guided for. In terms of product lines, the key driver was again our ALD business. By market segment, fourth quarter revenue was led by memory customers, followed by the foundry, and then logic segments. The gross margin increased to 44.8% compared to 43.7% in the third quarter. The lower activity level in the fourth quarter was more than offset by favorable mix effects.
SG&A expenses decreased by 4% compared to the previous quarter. As a reminder, in the third quarter, SG&A included close to EUR 1 million in one-off items. R&D expenses increased by 22% compared to the third quarter. Part of this increase is explained by a one-off EUR 3 million write-off of the remaining 450-millimeter assets. We generated operating income of EUR 12.5 million in the fourth quarter, down from EUR 23 million in the third quarter, and almost unchanged from EUR 13 million in the fourth quarter of 2014. Financing results in the quarter were EUR 6 million positive and mainly consisted of a translation gain of similar size. As we discussed in earlier quarters, a substantial part of ASMI's cash position is denominated in US dollar, and the related translation differences are included in the financing result. In the third quarter, the financing result included a EUR 3 million translation gain.
Tax in the fourth quarter has been positively impacted by EUR 5 million in one-off benefits, resulting from the recognition of a part of the tax losses that were incurred in the past in the Netherlands. That compares to a one-off tax benefit of EUR 9 million that we reported for the third quarter. Results from investments, which reflects our share of approximately 40% of the net earnings from ASMPT, amounted to EUR 2 million for the quarter, down from EUR 8 million in the third quarter and EUR 9 million in the fourth quarter of 2014. These figures exclude the ongoing amortization charge, which amounted to EUR 7 million in the fourth quarter. For 2016, this amortization charge is projected to be EUR 28 million. Note that ASMPT's net earnings in the fourth quarter were impacted by a few one-off items, including a restructuring charge of EUR 3 million.
In the fourth quarter, ASMPT's sales fell to HKD 2.9 billion. This was 9% lower than the HKD 3.2 billion reported in the third quarter, but better than what ASMPT had guided for. Compared to the fourth quarter of 2014, sales decreased by 15%. ASMPT reported bookings of HKD 381 million for the fourth quarter, up 11% year-on-year and flat compared to the third quarter, and as such, also better than earlier guidance. Turning back to ASMI consolidated operations. ASMI's net earnings on a normalized basis amounted to EUR 25 million in the fourth quarter, down from EUR 42 million in the third quarter and EUR 27 million in the last quarter of 2014. Our new orders in the fourth quarter amounted to EUR 135 million, which was at the higher end of our guidance.
Orders were 9% below the level of the third quarter and 12% lower than the fourth quarter of 2014. ALD was again the key driver behind the overall order intake. Looking at the customer segments, equipment orders in the quarter were led by the memory segment, both DRAM and NAND flash, followed by foundry and logic. Let's discuss the full-year results. Our net sales in 2015 increased by 23% on a reported basis to a new record high for our ASMI operations. On a constant currency basis, our sales increased by 14%. This compares with a relatively flat development for the overall wafer fab equipment market in 2015. For the third consecutive year, we have grown our sales by solid double digits, and for the fourth time in five years, we have outperformed the market. Gross margin in 2015 improved by 120 basis points to more than 44% year-over-year.
Apart from positive mix effects during the year, we also benefited from ongoing efficiency improvements. As discussed at previous occasions, we started a number of efficiency programs in 2013, including new outsourcing initiatives and the migration of a large supply base to Asia. Successful execution of these programs has contributed to an improved and competitive cost structure, as well as structurally increased and more robust growth margins. Our target remains to run the operations at a gross margin percentage level of low to mid 40s, barring a market downturn, for which the target is high 30 levels of gross margin. On a quarterly basis, gross margin will continue to be impacted by factors such as sales mix and utilization. SG&A expenses as a percentage of sales dropped from 15% in 2014 to 14% in 2015. R&D spending, however, increased significantly from 12% to 14% of sales.
Apart from currency effects and the one-off charges in the fourth quarter, we stepped up investments in R&D during the year in response to customer requests for new applications and engagements. Operating profit improved from EUR 91 million to EUR 106 million. The operating margin remained relatively steady at 16%. Results from investments, on a normalized basis, decreased EUR 61 million to EUR 44 million. Sales of ASMPT decreased by 9% to HKD 13 billion, reflecting the slowdown of the back-end equipment market in particularly the second half of the year. Looking at our consolidated numbers again, normalized net earnings increased by 13% to EUR 181 million for 2015. On a per share basis, normalized net earnings increased by 15% to EUR 2.87. Looking at our balance sheet and cash flow.
At the end of December, cash amounted to EUR 447 million, up from EUR 428 million at the end of September. This increase is mainly explained by solid cash generation during the quarter and positive currency effects, in part offset by the share repurchases during the quarter. At the end of December, net working capital stood at EUR 111 million, slightly down from EUR 113 million at the end of the third quarter. The number of outstanding days of working capital, measured against quarterly sales, increased from 63 days at the end of September to 69 days at the end of the fourth quarter, but decreased compared to 78 days at the end of December one year earlier, December 2014. Cash flow from operations amounted to EUR 34 million during the fourth quarter.
For the year as a whole, we generated EUR 145 million in cash flow from operations, up from EUR 110 million in 2014, driven by continued strong profitability while working capital remained well under control. Free cash flow increased by more than 30% to EUR 103 million in 2015. During the year, we spent a total of EUR 116 million on dividends and share buybacks together, up from approximately EUR 60 million in 2014. After the completion of the previous share buyback program earlier in 2015, we announced the authorization of a new buyback program last October for another EUR 100 million, as part of our continued commitment to use excess cash for the benefit of our shareholders. We started that program at the end of November, and during the fourth quarter, we spent approximately EUR 9 million to repurchase some 250,000 of our own shares.
Including the share repurchases so far in the first quarter of this year, we have completed approximately 24% of the program as per the end of last week. As announced earlier, the buyback will end on completion of the program, but ultimately in November of this year. As a reminder, a weekly update of our share repurchases can be found on the ASMI website. During 2015, we paid a dividend of, as you know, EUR 0.60 per share. This year, in 2016, as a reflection of our continued confidence in the sustainability of our profits, we will propose a dividend of EUR 0.70 per share, an increase of 17%, for approval at the AGM in May. Let's now briefly address the change to IFRS reporting. Up until 2015, ASMI has reported financial results according to both US GAAP and IFRS, with US GAAP as the primary standard.
Starting in 2016, ASMI will switch to IFRS as its only reporting standard. The appendix of the Q4 2015 press release shows a condensed P&L according to IFRS for the last eight quarters. If any of you have questions about this change in reporting, Peter van Bommel, who of course, has joined me on this call, can further discuss this during the upcoming Q&A. Let's now briefly talk about ALD. 2015 was another year of success for our ALD business. Despite moderating conditions in the broader market during the year, momentum in the ALD market was again solid. On the back of a continued strong and leading position, we were able to take advantage of the growth in ALD demand, as evidenced by the double-digit growth in our net sales in 2015. ALD was again the key driver behind the growth in revenue last year.
ALD accounted for clearly more than half of our total equipment revenue in 2015. ALD is now firmly established as a key enabling technology. In logic, foundry, and memory, the leading customers in the world have already ramped several technology generations based on our ALD equipment. In the more recent years, we have also broadened our customer base beyond the traditional top three. In 2015, looking at our total revenue base, we again had a growing contribution from the top four to top 10 customers as some of these newer customers invested in our ALD systems in substantial volumes. In combination with the expanded client base, we have also achieved a more robust and balanced customer mix over the years. Apart from expanding our presence in the logic foundry segment, we have built a much stronger exposure and customer relationships in the memory sector.
Our ALD equipment is an enabling technology for spacer-defined multiple patterning and used by virtually all of the memory customers. Equipment bookings in 2015, for our company as a whole, were led by the memory segment, followed by foundry and logic. Building out our leadership in ALD remains a key element in our strategy to grow our company. As reflected by the increase in R&D spending in 2015, we continue to invest significant resources in further enhancing our leading platforms and to grow the pipeline of new ALD applications that will support our customers in transitioning to the future technology nodes. The structural growth prospects for the ALD market are strong.
Looking at the technology nodes that will move to HVM in 2016 and 2017, we are confident that the number of ALD process steps will show a healthy increase relative to the previous generations, and that consequently, the share of ALD of the total deposition equipment market will further increase. In logic and foundry, the transition to 10 nanometer node will be an important step in the expansion of the ALD market. The combination of smaller geometries and increased complexity means that in these next generation FinFET devices, more deposition steps and more layers require single wafer ALD. At the 10 nanometer node in logic and foundry, the number of ALD applications for which we have been selected is significantly higher as compared to the previous 16, 14 nanometer generations.
As we highlighted in the previous quarters, ALD-based multiple patterning is part of these new and significant opportunities at the 10 nanometer node. More of the critical patterning steps at 10 nanometer will require spacer-defined double patterning based on single wafer/mini-batch ALD. With the leverage of our strong track record in the memory market, where our equipment has already supported the ramp of several technology generations, we expect to benefit as our logic and foundry customers also expand their use of ALD-based multiple patterning. Multiple patterning also continues to be a key enabler for customers in the DRAM sector to transition to smaller geometries. At the next technology node, the number of multiple patterning layers is expected to further increase. We are well-placed to support our DRAM customers in their ever-increasing technology and productivity requirements, as soon as equipment demand for the 1X node starts to pick up.
In NAND flash, we have taken important steps in the past few years to strengthen our participation in next generations of 3D NAND device manufacturing. In the course of 2016, we expect to gradually increase the contribution from new 3D NAND applications. When taking a longer-term view, we remain confident about the growth potential of the ALD market. As our customers move along Moore's Law, further miniaturization, introduction of new device architectures and new materials will only further increase the need for superior precision and conformality offered by ALD. We still expect the ALD market to double in size in a three to four-year period compared to the baseline in 2014. As a market leader, we believe our company has strong opportunities to benefit from continued solid growth in the ALD market. Now let's briefly talk about the market environment.
Looking back at 2015, conditions in the semiconductor end markets slowed down in the course of the year. The market for wafer fab equipment ended the year relatively flat as compared to 2014. Following an uptick related to 14/16 nanometer in the first half of the year, wafer fab equipment spending in the logic foundry segment was sequentially lower in the second half. DRAM in the second half was lower, following a very strong level of 20 nanometer spending in the first part of the year. For the year 2015 as a whole, the memory segment was the key driver for the WFE market. Looking at 2016, at the current year, against the backdrop of an uncertain macroeconomic outlook, market research firms such as VLSI and Gartner are projecting the WFE market to be down by a low to mid-single digit percentage compared to 2015.
Spending in the logic foundry segment has the potential to increase in 2016. Key driver in this segment will be the transition to the 10 nanometer node. While the visibility for the size of this ramp in logic foundry going into the latter part of 2016 and into 2017 is still limited, investments in the near term are picking up. With the current limited visibility, DRAM industry spending is likely to be down a double-digit percentage compared to a strong level in 2015, according to the market research firms. NAND flash spending for the WFE market as a whole is likely to be up in 2016, with the mix further shifting to 3D NAND. Let's look at the guidance ASMI gave as part of our press release that we announced overnight. For Q1, we expect sales between EUR 135 million and EUR 145 million.
For Q2, we expect sales between EUR 140 million and EUR 150 million. In line with our earlier views, based on our current visibility, we expect 2016 to be more revenue backloaded. The Q1 order intake is expected to be in the range of EUR 150 million to EUR 165 million. All these outlook related figures are on a currency comparable level. At this point, Victor, I propose we take any questions the audience may have.
Yeah. Chuck. Okay. All right. Clearing up. We are ready for the first question.
Thank you. If you would like to ask a question at this time, please press star, followed by the one on your telephone keypad. Please ensure the mute function on your telephone is switched off to allow your signal to reach our equipment. If you find that your question has already been answered, you may remove yourself from the queue by pressing star two. Again, please press star one to ask a question. We will now take our first question from Pieter de Pous from Kepler Cheuvreux. Please go ahead. Your line is open.
Yes. Good afternoon, gentlemen. Maybe first on ALD. Could you maybe share your thoughts on the development of your market share in ALD, both in 2015 and in 2016, how you see that developing? Somewhat related to that, a question on the outlook for the full year. Your U.S. competitors in deposition, Lam and Applied, have said that considering a kind of flattish wafer fab equipment market this year, they should be able to show some sales growth and be able to outperform the market. You think you will be able to outperform the industry as well, like you did in recent years? My final question relates to China. It seems there's quite some investment activity in China. How does your positioning in China compare to other parts of the world?
Do you have the infrastructure in place there to support international customers that are looking to build facilities in China? Or may we see some increase in OpEx as you expand your service and support staff in China? Thank you.
Yeah. Pieter, okay, thanks for your question. On market share ALD, we are very confident we maintained a strong position in the ALD market. As we shared with you before, we are really ready for the shift to 10 nanometer, from 16, 14 nanometer in logic to 10 nanometer, we expect to strongly benefit from that. Also, we are ready for several transitions in the memory market. We feel very good about our competitive position in the different markets. At the same time, as we shared with you before, as the market grows from $600 million in 2014 as a baseline to north of $1.2 billion three, four years later, likely our percentage market share in ALD may go down somewhat. That is not something we are focused on.
In the transition from 16, 14 to 10, we see that the number of applications that we are engaged in with our customers is going to increase in a significant way. Also, our engagement in VNAND is really starting to take off. Coming from a very low level, but it's starting to happen now. In DRAM, as soon as the industry picks up again, going to do 1X nodes and starts to use even more multiple patterning, we are ready. We are ready to engage. From that point of view, we feel good about the competitive position. At the same time, of course, we are dependent on, let's say, the volume at which our customers are going to kick in these new nodes. That will ultimately determine our top line and bottom line results for the company.
Looking at the WFE market as a whole, as we shared in our introduction, we do expect that the market as a whole, looking at, let's say, the industry analysts, they foresee that the market will go down somewhere between, let's say, well, basically with a low to mid-single digit percentage compared to 2015. That is the guideline we assume at this moment in time. Our ambition in that respect has not changed, that our ambition is for this year, based on all the assumptions we have made for 2016, our ambition is to outgrow the market. On the China infrastructure. On China in general, you asked if we have the infrastructure. Well, we've been asked before. I think there's one fab that has created a lot of attention on its refurbished fab for VNAND purposes and so on.
We have basic activities there, but we will not necessarily benefit in the first phase in that fab of CapEx expansion. That just has to do with the fact that four or five years ago, the company was less focused on VNAND. The last two, three years, that has changed significantly. As a result of that, our VNAND engagement on, let's say, transitions that are now going to higher stacks and with new players, we are ready to engage. On these older nodes, our benefit will be low. As such, we don't foresee any challenges in terms of our infrastructure in China on the short term. There are other memory fabs that have been there on DRAM and flash for a longer period of time, that we have been engaging on for multiple years. There we have a very healthy existing infrastructure.
Also there, we don't foresee any issues. I trust that answers your question. Yes, it does. Okay. Thank you. You're welcome.
Our next question comes from Mathias Santos Silva from Morgan Stanley. Please go ahead. Your line is open.
Hello. Thank you for taking my question. Strong gross margin performance surprised us to some extent as you weren't helped by the high level of top-line growth as we saw earlier during the year. You say mix was beneficial. Could you shed some light on what type of mix benefit you experienced in the quarter? That also takes me to my second question. How do you view the competitive environment in each of the different application areas, such as high-k metal gate, spacer-defined multiple patterning and ion doping? Are there material difference to the competitive intensity? Thank you.
Yeah, let me answer the question first of the gross margin. As what we said in earlier calls, the gross margin is not dependent on the sort of business that we have. It's not product line driven. It's more driven by the fact that certain applications for certain customers might have a higher margin or a lower margin than other things. In this quarter, we see two big benefits. The one is the fact that our constant cost reduction efforts that we are taking now already for quite some time, which [Chip] was also referring to in his remarks, is improving. That's an important thing, of course. On the other hand, we have seen that the mix in the products, in combination with certain customers, was a little bit better than what we have seen in previous quarters. It's fluctuating.
I think it's relatively stable or has been relatively stable during the year. When you look to our gross margin development in the course of 2015, the fluctuation has been around that 44% level, with an up of 1% and a down of 1% for each individual quarter. I think that's the guidance in a good year that we have given earlier to the market, where we say, okay, in a good year, we expect low to mid-40s as a gross margin. I think that's also reflected in the Q4 numbers. On the competitive landscape. We, of course, face, as we shared in earlier calls, we face competition in each of the markets that we are engaging in. You mentioned high-k metal gate, you mentioned, I think, the patterning market, and we have engagement of competitors in all those markets.
We try to know, understand very well what our competitors are doing and anticipate and react on that in the best possible way. I don't think our competitive position has materially changed over, let's say, the last six months or so. We are confident that, on one hand, we know that competitors will take a share of our markets, of the markets that we engage in. Our view is that we are confident that we have the right products and roadmaps in place to really grow based on the huge potential that the ALD market represents in the coming years. That's what I would like to say. In general, we don't comment on specific competitors and specific products, but I trust that this call provides you enough insight.
Thank you very much.
Our next question comes from Nigel van Putten from ING. Please go ahead. Your line is open.
Hi. Thank you. Good afternoon. I have a question or more of a follow-up on your competitive positioning going towards the 10 nanometer. Where do you see most growth in the number of applications? Is that more towards the front end of the line or more towards the back end of the line? If we segment the market that way, how do you see your relative positioning?
Could you repeat one more time the specifics of your question? We could not hear you completely.
Oh, sorry. Yes, it's from the 14 nanometer towards 10. You talk about growth in applications. Do you see a difference between growth of application in the front end of the line, so more towards the gate or the back end of the line? As a follow-up, do you see any difference in your relative positioning competitively if we segment the market that way?
Well, to start with the latter, I think, in general, our view is that our position is strengthening going from 16, 14, to 10 in logic foundry. That answers the latter part of your question.
Okay.
The first part of your question, I think what you can say is that, in the FinFET device itself, the need for ALD is increasing. In the device itself, going from 16, 14 to 10, and at the same time, you see increased, as we shared, I think before, also the patterning needs in single wafer ALD is really there. Is increasing going from 16, 14, to 10. Those are both important elements, that contribute to the growth that we foresee in the level of penetration. Yeah?
Yes. Thank you. Yes. Thanks. Maybe a quick follow-up. I know you don't comment on the competitors. I think one has said that they now have the lowest cost of ownership, and the other said it's 2 to 4 times faster. I'm not going to ask you to evaluate those claims, but do you have also a unique selling point, along those lines?
I think, whether we have a unique selling point, I think, ultimately customers decide and only the numbers speak. Yeah. As long as we are looking at the number of applications that we are engaged in, is growing from 16, 14 to 10, tells us that we are doing some things right, in terms of providing unique technological solutions to the customers. At the same time, from a cost point of view, also serve their continuous drive for lower costs, because as Moore's Law continues, cost becomes increasingly important. It's not one or the other. Both need to be addressed continuously by suppliers in maintaining a position at our customers, and that's what we are working on day and night. I think that's the best answer we can give.
That's very clear. Thank you.
Okay. You're welcome.
Our next question comes from [Philip Scholz] from Kempen. Please go ahead. Your line is open.
Yes. Good afternoon, everybody. Two questions. First, on ALD. Can you maybe give your best guess maybe about the size of the ALD market in 2015 versus the baseline, 2014? Can you maybe talk a little bit about your view, you continue to say that you expect that ALD market to double in the next three or four years. What could be a next trigger, or what do you really need in order for that market to double in three years instead of four years? My second question is on the transition to 3D NAND. You said that currently your involvement is actually quite limited. What is actually the real transition going on in that 3D NAND market that means that your involvement will be bigger? Is there a specific technology node, or how you say that, where you can mention on that?
Okay. Thank you, Philip. On the size, we're looking at those, where actually since the year has just ended, we are, of course, very well aware of our own numbers. We are still waiting for more detailed numbers on the market as a whole, which we always compare with, let's say, more independent agencies. We're not ready to answer your question on the 2015 market as a whole from an absolute point of view. What we can say, it's really a solid double-digit % with which the market has grown year-on-year. As soon as we can provide an update on, let's say, the 2014 baseline and the EUR 600-EUR 1.2 billion guidance that we gave to the market, we will do so. We will try, this year, whether we can provide an update. Unfortunately not at this moment in time.
Clearly, a solid double-digit % growth has taken place in 2015. On 3D NAND, you're asking what's really going on there and what determines our growth. That's how I perceive the question. I just want to be very transparent on that. Again, until three, four years ago, we were not that focused on that market. Let's say two, three years ago, we really started to focus on that part of the memory market. As you know, seeding efforts always take two, three years, at least, in whatever market you are engaged. As a result of that, you see now that as of late 2015, revenue starts to kick in V-NAND. It was in a very modest way towards the end of last year, and more meaningful contribution will start to develop in the course of 2016.
It has to do with our focus on that market. It has to do with that more than one player are now, let's say, starting to get up to speed in this market and also with, let's say, higher stack solutions where the technologies become more and more complex and where, let's say, single wafer ALD might be a more suitable solution than the technologies they have been using so far. That's how you should see it, and that's why the contribution to our P&L will gradually grow as our engagements strengthen.
They strengthen also based on the fact that we build relationships with these memory customers over time through our double patterning programs that we have had in place for multiple years, through which these customers have got to know the company and have got to know the tools and the capabilities of this company in general. That's how you should see this. As a result of that, the earlier refurbished fab that we talked about, we are not immediately engaged there, because we didn't have the historic relationships. That's just a fact, whether we like it or not, and I don't want to be too secretive about that. Yeah. That's what you have to take into account in your models.
All right.
Does that answer your question?
Absolutely. Thank you very much.
Okay. You're welcome.
Our next question comes from Marc Hesselink, from ABN AMRO. Please go ahead. Your line is open.
Yeah. Thank you. My first question is on the technology ramp of 10 nanometer in the second half of the year. You're guiding for quite a strong order intake in the first quarter. Does it mean that the ramp will be relatively aggressive, or do you expect it to be more gradual and then building up into 2017? The second question is your demand for ALD in the medium term. Now it seems that for seven nanometer, at least partly there will be multiple patterning. How does that relate to your guidance of the doubling of the market in a three to four-year period? How is EUV related into that guidance? Did you include any multiple patterning on seven nanometer? Then finally on R&D investments, growing quite rapidly in the last year, what is your expectation going forward if you're building out the applications?
You continue to have to invest extra in R&D there?
Okay. Mark, yeah. Thanks for your questions. Three questions. I'll let the R&D be answered by Peter. First on the other two questions. 10 nanometer, we think that our view now is that from a revenue point of view, likely there will be a healthy contribution throughout the year, in each quarter. Again, that's the visibility we now have. There will be a healthy contribution throughout the year, each quarter, and not a hockey stick effect or so. Secondly, the initial capacity that will be built in 2016 is just, let's say, the initial phase of wafer starts. As far as we can see, again, based on the visibility we have from customers, more significant wafer starts capacity will be built in 2017. For our customers, depending on their end user demand.
That of course, depends on macroeconomic demand for smartphones, et cetera. That's how we view it. Initial capacity being built in 2016. For sure, assuming normal economic conditions, that will not be the end of building 10-nanometer capacity this year, on the contrary. Okay? That's on 10 nanometer. Again, that's based on the visibility we have now. It could be that the customers start to pull in demand or push out demand based on macroeconomic developments in the course of this year. Then we would update you of that in the course of this year, of course. On EUV patterning, whether that has an impact on the doubling of the market, that guidance that we provided, compared to 2014? The short answer is no. We don't foresee that that has any impact on that prediction.
That's again, based on our assessment today. R&D?
Yeah. When you look to R&D, I think a few things that I would like to give a little bit of color about. First of all, when you look in absolute terms, you see that our R&D was increasing from EUR 64 million to EUR 95 million. That's an absolute amount and a substantial amount of nearly 50%. There are two important differences in that. On the one hand, you have an impact of nearly EUR 10 million simply due to currencies. We are basically a dollar company, which is reporting in euros. While our R&D is distributed, our resources are distributed well over the world. The impact of a weakening of the euro plays an important role. Second thing is that we had, of course, in this quarter, a one-off.
We decided, based upon the uncertainty, around 450-millimeter, to write off the remaining part of the assets that we had in the 450-millimeters. When you exclude that, you still have a substantial increase. That increase is driven by what we mentioned in earlier calls, for a big part, by the requirements of our customers. We see that more and more customers are attracted by our solutions. We are serving the whole top 10. That requires, because you have to align their process developments with our products, that require more resources. Moreover, we have seen during the year that they came back towards us and asked us to help them to explore more and more alternatives and more opportunities that they saw of using our equipment.
That's a long answer. That drives that development, and we expect that those requirements will increase in the forthcoming period. That is important, I think. On the same moment, we have indicated towards the market that we expect R&D to be in the low to mid-teens as a percentage of sales. That's still the goal that we are striving for. Again, that doesn't mean that every quarter that that will be exactly at that spot, because you build up R&D not very quickly. You reduce it also not that very quickly. When you look at this over an annual level or a longer period of time, that is the goal where we are striving for and what we are managing our company on. I trust that that gives you a little bit of color on how R&D is developing.
Yeah, it was very clear. Thanks.
Our next question will come from Tammy Qiu from Berenberg. Please go ahead. Your line is open.
Hi, guys. Thank you for taking my question. Firstly, we'd like to get more understanding on the 10 nanometer ramp-up pattern. Basically, you have been saying that 2016 is only the initial phase for what if we start building. Does that mean if the trend continues to continue in 2017 from orders perspective, will we see continuous order strength driven by 10 nanometer throughout 2016? Another question is for Q2 revenue, the growth wasn't significant if you compare to the level of order expecting for Q1. Does that mean you are actually running at a lead time of more than three months? Hopefully we can see impressive revenue growth in second half from Q3 and throughout 2017, hopefully. Is that correct understanding?
Well, thank you, Tammy, for your question. There are multiple questions in your comments. The best way to answer it is that you would like to get a little bit more color on 10 nanometer. It is the case that we like to reiterate that we anticipate, we expect that indeed orders will continue to happen in 2017 related to the 10 nanometer ramp, assuming that the modest capacity that There's a lot of noise in the background.
Yeah, I can mute it. Sorry.
Please. Please do so. Thank you. All right, that's more pleasant also for the rest of the audience. 10 nanometer, again, we foresee a steady contribution in revenue throughout the year based on the visibility we have from our customers now, and we do expect that to continue into the next year. Indeed, some orders may be placed ahead of time, but maybe you are referring also to the high level of orders in Q1 that we gave as a guidance compared to the guidance we gave on revenue for Q1.
Yes.
To explain that, part of the bookings in Q1 are likely indeed for shipments in the second half. Another part of the upside are bookings that are expected to immediately turn into revenue in Q1. That's a little bit of a mixed bag that we, at least based on current visibility, for the remainder of the quarter, foresee. I trust that that answers your question.
Is it now still correct to understand you are running basically a three-month lead time or slightly shorter?
It depends on the product and on the customer, because some customers only provide us guidance and place the order, and then immediately want the shipment, and others place way ahead of time. That really differs a lot by customer. That's why sometimes we also have to carefully choose our wording to give the right guidance to you as an audience.
Okay, thank you, guys.
In general, it's three to four months max. In general, that is the case, Tammy.
Okay, thank you.
You're welcome.
Our next question comes from Jim Fontanelli from Arete Research. Please go ahead. Your line is open.
Thank you. Afternoon. I just wanted to follow up. Obviously, you've covered up some detail on the opportunities at 10 nanometer and how you see that ramp unfolding over the course of this year. I just wanted to dig into that a little bit. It looks at the moment, certainly, that 10 nanometer, certainly from a foundry perspective, is going to be a short node, akin to what we saw with 20 nanometers going to 14, 16. I guess my question is, do you think you'll be able to carry your market share at 10 through to seven nanometers? Or are you having to re-qualify, in a kind of renew the process of record at seven versus 10? Or is it a carry through of the process at 10 to seven, and you'll effectively keep your share at 10 through to seven?
Secondly, at seven, you mentioned you didn't think there was much change in your potential opportunity, whether EUV was incorporated into seven or not. I just wanted to clarify why you thought that was the case. Again, I guess it looks increasingly likely that EUV won't be used for seven nanometers, and will be introduced to five, and that would significantly increase multiple patterning requirements at seven. I'm interested to see why you don't think, the presence of EUV or not at seven makes a difference to ALD penetration.
Okay. Jim, okay. First, on a short note, it's clear that every customer decides at every node what the best set of applications and supplier platforms they need. From that point of view, every node is a decision on its own. At the same time, it's not very attractive for a customer to change at every node, every application, from supplier A to supplier B. That would also be very expensive, especially when the node is that short. They have to write off the equipment in a very short period of time. Those things we all take into account in assessing our own chances. Our belief now is that we have a very good starting position, going into 10, and as a result, also to build on that going to seven, regardless of how long that 10 nanometer node will be.
We learn, of course, as we go into high volume, and customers will learn also on our performance of our tools, as they go into HVM with our tools. Our current assessment is that we know what we are doing and that we can build on the strengths that we have developed going from 16, 14 to 10. On EUV, at least your perception was that we tried to say that there was no impact of EUV on ALD, and I apologize if that has come across, because that's not what I meant to say. The question earlier was whether EUV would have any impact on the projections of the market going from EUR 600 to EUR 1.2 billion in the next three to four years, with 2014 as a baseline. Based on that, we said clearly, no.
Of course, when EUV comes, it may have an impact on the market. Our view is that it will take a couple of years. It will take still quite a number of years before EUV will be out there in a real HVM mode with a significant amount of layers. In the meantime, the patterning market will still grow in a significant way. That's what you and we are taking into account into modeling our own opportunities in the coming years. Secondly, that is the impact of EUV in the patterning market. We also have a lot of applications in ALD not related to patterning at all.
Those markets will only get fueled in a positive way to use ALD, because in those markets, EUV will accommodate Moore's Law to continue even longer in going to smaller geometries, which requires more ALD. In those markets, it serves the interest of growing our market. I trust that answers your question.
Yes. Thanks. Could I just follow up just very quickly on that 10 nanometer point? You mentioned there was a healthy increase in ALD use from 16, 14 down to 10. Do you think you could give us a slightly more specific guide on what healthy means in terms of process step increase from 16, 14 down to 10 in logic?
Well, it's way more than a handful.
That's way more than a handful incrementally?
Yeah, it differs a little bit by customer. Not applications, but I'm talking about layers. Yeah?
Yeah.
I should be clear. I talk about layers, not applications. In terms of layers, it's more than a handful of layers, depending on what customer you talk about.
Great. That's useful. If I may, just one follow-up question on vNAND. You mentioned a kind of slow ramp of engagement into vNAND over the course of this year. Obviously, if we look at the number of wafers being added in vNAND this year, it's significantly more than a slow creep. I guess we're going from 100,000 wafers to maybe 400,000 wafers over the course of 12 months. Is ASMI's engagement specifically around one or two customers ramping within that overall wafer growth profile? Or is it about where the overall customer base is in terms of ramping through 32 to 48 layers and then maybe into 64 layers? Is it more of a node-driven engagement? Or is it because, as you mentioned, specifically some customers are coming in late to vNAND and are only really starting to ramp in earnest this year?
It's a combination of both.
Okay. Do you see any opportunity of increasing penetration into the Korean marketplace for vNAND over the course of this year? Or is it going to be led outside of, is your specific growth going to be led outside of Korea by wafer additions from Toshiba and Micron?
Let's say, the only thing I would like to say is that we are focused on every part of the world.
Okay. Thank you.
All right. You're welcome.
Our next question comes from Edwin de Jong from SNS. Please go ahead. Your line is open.
Good afternoon, gentlemen. A few questions left. On ALD, the Epitaxy and the furnaces market, could you give some color on that and maybe if there are new applications there or some additional color, please? On the ASMPT, you're now the major shareholder, of course. The contribution of ASMPT has decreased significantly. What does that mean for the talks that you are having with management of ASMPT? Are you going to push for more cost savings, for instance? Or are you looking for new technologies that they have to develop? What's your position there? Finally, on vNAND. You're coming up with new applications in the course of the year. How many kinds of machines should we think of? Finally, if you go from 32 layers to 64 layers in vNAND, does that mean that you, for instance, need twice as many machines?
how should we look at that? Those were the questions.
Let me see. On V-NAND, you had two questions. I'm just writing down. Three questions. Why don't you start with ASMPT? Otherwise, I talk all the time.
Let me give that information. I think, first of all, what happened with the ASMPT numbers? I think to give you our perspective as a shareholder with the announcement that you saw earlier today from their side. First of all, I think the view that they have provided is a little bit more optimistic than what we have seen in the past period. The sales decrease in the fourth quarter, which is by definition their lowest quarter, was slightly better than what they expected originally and also what the market was expecting. That is, I think, a very important thing, their order intake was the highest ever that they have reached in a fourth quarter, and they expect now that when you have normally an uptick in the first quarter, that that indeed continues. I think important to mention there is that the gross margin improved substantially.
They had a very weak third quarter. When you look now to the fourth quarter, the uptick that they had, despite the lower activity level, that was nearly the highest that they had in 2015. They announced some restructuring costs. While the results in itself were not that good, the bottom line result, they were basically related to a few things. First of all, restructuring cost. Secondly, due to the distribution of the sales, they had in a net income, a relatively high tax rate this quarter. Beside that, I think that the underlying quality of the results is improving, and I think that as a shareholder, as we look at it, that's how
What are our most important tools in this respect, I think, is that things are improving or at least are becoming better than they were in the past quarters. I trust that gives you a little bit of color, Edwin, how that we look at it.
Maybe a little additional one. How many times do you speak with management? Do you speak to them once a month, or how does that work in your communication to them?
We speak on a regular basis with them. We have, of course, the official board meetings with them, but we speak on a very regular basis with them. Yeah.
Very regular. Okay.
Okay. On Epitaxy and vertical furnace, as you asked. Well, yeah, those product lines have really, in a very steady and healthy way, contributed to the 2015 growth of the company. Really literally in a very steady way. Also from a profitability and a cash flow point of view, those two product lines have been contributing. If you look, we are clearly investing in specific areas of both product lines. Yeah, we do expect that to make progress, to really grow our engagement in both product lines, in specific areas of the market. It's too early and part of the R&D investments that we have been engaged on in 2015, have been related also outside ALD to these product lines.
As soon as our view is that we are close to then those new initiatives contributing to growth in the top line and the growth of our addressable market, we will share it with you, but it's just too early to do so. It should be clear, healthy contribution from a profitability and cash flow point of view now, and we are investing in specific areas in both of those product lines. It's just not a stable business where the product lines are treated as pure cash cows. That's absolutely not the case. Okay?
That's very clear. Yeah.
On V-NAND. You had a number of tools in 2016. The only thing I can say is that the revenue will be significantly higher, again, with the visibility we have now, than in 2015 for vertical NAND. We do expect that the contribution in the second half will be bigger than in the first half for vertical NAND. That's what I can say. Your question was also on higher stacks. Yeah, that's interesting. I think that's too early to say. We are extensively engaged with multiple customers in qualifying for, let's say, the stacks they are working on. That is really by customer for a number of applications. Not one, but a number of applications by customer.
I think it's fair to say that we and the customers are learning how we can proliferate this technology further as we go to next technology nodes. I think it's too early to say, because, again, we're now in a phase of preparing for HVM of the current stack. That's too early to tell, but I can guarantee you that we are doing all we can to, of course, further penetrate into these markets, as we really have learned a lot in the last two, three years, starting from scratch in this part of the market.
All right. Thank you.
Our final question comes from Pieter [Oliphasen] from Kepler Cheuvreux. Please go ahead. Your line is open.
Yes, thank you. I wanted to clarify the outlook statement for the full year, where you said that sales will be back-end loaded. That means for 3D NAND, H2 will be better than H1. For 10 nanometer logic foundry, it will be kind of steady through the year. For DRAM, at this moment, you don't foresee a pickup in the second half. Is that the way to look at it?
That's a good question. I think, the industry watchers think that year on year, there is a meaningful double-digit decline of the DRAM market. For us, it's too early to say whether the second half will be as weak as the first half. In DRAM. That is, for us, that's still to be seen. Year on year, that's what industry watchers look at. How the second half of the year develops compared to the first half, for DRAM, specifically for DRAM, as you asked, that's too early to tell. The only thing we can say is that we are ready with the right solutions to engage there as soon as the customers are ready to invest. There is some contribution in the first half, I should say that, but not at the level, let's say, what the industry is looking for. I trust that answers your question.
Also the second half, again, also you said that you made a certain assumption on Logic Foundry, you can make that assumption maybe on, let's say, modest volume developments in the 10-nanometer node. Modest capacity development in the 10-nanometer node, as the industry sees it now. That can change. There, visibility is limited. If volume there would be accelerated, then it could impact also the industry in the second half. That's really too early to tell. As you know, in this industry, things can change quickly in a three-month period of time.
Okay. Thank you for the clarification.
Okay. You're welcome, Peter.
There are no further questions at this time.
Okay, we had an extensive call, one hour and 15 minutes. It's clear that there were a lot of questions on your mind. Thank you for sharing those. I trust that both of us, Peter and I, were able to address as much as we can your questions. Please feel free to follow up with us through Victor or through the roadshows we have, on any follow-up questions you may have. Final comment, thank you again for joining today and let's stay in touch and have a very good day or evening. Thank you.
That will conclude today's conference call. Thank you for your participation, ladies and gentlemen. You may now disconnect.