Good afternoon, ladies and gentlemen. Thank you for standing by, and welcome to the ASM International Q4 2018 earnings call. At this time, all participants are in a listen-only mode. There will be a presentation followed by a question and answer session, at which time, if you wish to ask a question, you will need to press star one on your telephone and wait for your name to be announced. I must advise you that this conference is being recorded today on Friday, the 22nd of February, 2019, and I would now like to hand the conference over to your host today, Victor Bareño. Please go ahead, sir.
Thank you, Valerie. ASMI issued its fourth quarter and full year 2018 results last evening at 6:00 Central European time. For those of you who have not yet seen the press release, it, along with our latest investor presentation, is accessible on our website, asm.com. As always, we remind you that this conference call may contain information relating to ASM's future business and results, in addition to historical information. For more information on the risk factors related to such forward-looking statements, please refer to our company's press releases, reports, and financial statements, which are available on our website. With that, I will turn the call over to Chuck del Prado, President and CEO of ASMI. Chuck?
Thank you, Victor, and thanks to everyone, for attending our fourth quarter and full year 2018 results conference call, and for your continued interest in the company. Let's start with a review of the highlights of last year. 2018 has been a successful year for our company, with sales growth of 17% on a U.S. dollar basis, we outperformed the wafer fab equipment market. We progressed well towards our strategic targets as we further expanded our addressable market in ALD, and as we strengthened our position in the other product lines. We stepped up our investments to prepare our company for the next stage of growth, and at the same time, we returned more than EUR 600 million to our shareholders in 2018. Supported by a record high backlog, we have entered 2019 on a solid footing.
We would like to thank our employees for their continued dedication and hard work during the year. Let's now review our fourth quarter and full year 2018 financial results, starting with the fourth quarter. Net sales in the fourth quarter increased strongly to EUR 255 million, up 30% from the third quarter, and up 40% from the fourth quarter of 2017. Sales in the quarter were slightly above the top end of our guidance, which was a range of between EUR 220 million and EUR 250 million. Quarterly sales in Q4 hit a new record high for ASM and were 19% above the previous high. Sales in Q4 continued to be led by our ALD business, although sales in the other product lines were also strong and up significantly from the quarter before.
By industry segment, the revenue stream in the fourth quarter was led by logic, followed by foundry, and then memory and analog. Logic sales increased strongly quarter-on-quarter. Foundry and DRAM sales were steady compared to Q3, while NAND sales was weak. Following the strong order intake in the third quarter, sales in the analog segment increased to a record level in the fourth quarter. The gross margin increased to 41.7% in Q4, up from 40.9% in the third quarter, mainly driven by a better mix and in line with our previous indication for a slight improvement. In terms of operating expenses, R&D increased 18% sequentially, which in part is explained by one-off expenses related to the new facility in Korea, as well as extra material costs related to development tools.
SG&A increased 13% sequentially, which is explained by higher legal costs and the higher overall sales level. Operating income of EUR 46 million in the quarter increased strongly by 64% compared to the third quarter. Let's now have a closer look at ASMPT, ASM Pacific Technology. Normalized results from investments, which reflect our share of the net earnings from ASMPT, amounted to EUR 6 million, down from EUR 17 million in the third quarter. In the fourth quarter, ASMPT reported sales of $609 million, down Q-on-Q was 8% and up 12% year-on-year, and in line with ASMPT's guidance. While SMT, so the surface mount technology Q4 sales were strong, up 13% year-on-year. Back-end equipment sales dropped 7%. Profitability decreased in the fourth quarter due to a 5 percentage points decline in the gross margin.
Factors causing this decrease in gross margin included impairment cost, PPA cost due to acquisitions, costs related to new market penetrations and under absorption as ASMPT works down inventories. For the full year 2018, ASMPT increased sales by 12% to $2.5 billion. Net profits decreased by 21%, mainly impacted by the lower profits in the fourth quarter. Important event was the acquisition of NEXX, which substantially strengthened ASMPT's position in the market for advanced packaging. Now turning back to ASMI's consolidated operations. In the fourth quarter, we booked EUR 302 million in new orders, up 17% from the third quarter and up 48% from the first quarter of 2017. This number clearly exceeded our guidance, which was a range between EUR 240 million and EUR 260 million. The upside is largely explained by pull-ins of orders that were originally expected in the first half of 2019.
ALD was the main driver behind the order intake in the quarter, but we also recorded strong momentum in each of the other product lines, furnace, epitaxy, and PECVD. By industry segment, equipment orders in the quarter were led by logic, followed by foundry, and then memory. Logic orders remained relatively stable compared to the record high level in Q3. Foundry orders also remained relatively stable compared to Q3, with an increasing mix of 5 nm related demand. Orders in memory increased compared to the third quarter, mostly in DRAM, but were relatively modest compared to the combined orders in logic and foundry. Analog had again an important contribution to the order intake in the quarter. Now let's discuss the full-year results.
At EUR 818 million, our net sales in 2018 reached a new high that year, increasing by 12% on a reported basis and by 17% on a U.S. dollar basis. In terms of industry segments, we booked record high sales in logic, last year, driven by strong growth in demand related to the advanced nodes. Foundry decreased somewhat for the full year but remained at a high level and picked up strongly in the second half. In DRAM, we booked strong growth compared to a more modest level in 2017. In NAND, sales dropped in 2018 following a strong 2017. Analog, while smaller than the other segments, showed a strong increase driven by demand in areas such as power devices and IoT. Sales in the full year were led by our ALD product line, which continued to represent more than half of our equipment revenue.
2018 was, however, also a notably strong year for our other product lines, with vertical furnaces, epitaxy, and PECVD all showing solid sales increases. The key driver here was the increasing demand in logic and analog, where we traditionally are strongly positioned with a broader range of products. Last but not least, in 2018, we achieved our target to more than double our sales in the Chinese market compared to a more limited base level in 2017. Gross margin decreased slightly from 41.3% in 2017 to 40.9% in 2018, which is explained by the impact from investments in new products and growth initiatives. Last year, we talked about the investments that we have been making in the growth of our company and the new products we have been bringing to the market.
Looking at 2019, these investments will remain at a higher level, and our pipeline of new products and applications continues to be healthy. As a consequence, the gross margin in some of the quarters of 2019 may show some fluctuations. In addition, as always, gross margin can vary from quarter to quarter, depending on factors such as mix. Our gross margin target remains a range of low to mid-40s%. Operating expenses remained under control during the year. SG&A expenses increased by 22% due to the cost related to the new PLM system, product lifecycle management system, also targeted investments such as in the Chinese market, and higher legal expenses for the patent litigation cases that we initiated in 2017. Total R&D expenses increased by 10%, excluding the capitalization effect under IFRS. Operating profit for the year increased 14%.
Results from investments on a normalized basis decreased from EUR 112 million to EUR 61 million. This was the combined effect of an aforementioned decrease in ASMPT's net profits in 2018 and the full-year effect of the reduction in our stake in 2017 from 39% to 25%. Our normalized net earnings also excluding the book profit on the ASMPT stake sales in 2017, decreased by 9% to EUR 170 million in last year. Let's now look at our balance sheet and cash flow in a little bit more detail. At the end of December, cash amounted to EUR 286 million, up from EUR 266 million at the end of September. This increase was largely the balance of substantial positive free cash flow of approximately, EUR 60 million in the fourth quarter, and a cash outflow of EUR 46 million for share buybacks.
For the full year of 2018, free cash flow decreased from EUR 32 million to EUR 23 million. This is explained by an increase in capital expenditure from EUR 43 million to EUR 63 million, and additional investments in evaluation tools, particularly by the higher profits and improved working capital management. Okay. I trust that is clear, and in case there is any unclarity, you can, of course, follow up with a question after this introduction. Feel free to do so. Let's continue talking about our investments. In terms of investments, we completed last October our new facility in Dongtan, Korea, which includes an expanded state-of-the-art R&D lab.
We also made the first investments for a new expanded manufacturing facility in Singapore. As we further invest in the construction of this facility, we project in 2019 a continued higher level of CapEx compared to 2017, in line with our earlier indications. In the fourth quarter, we managed to reduce working capital, including higher customer prepayments and the effect of initiatives that we started earlier in 2018 to structurally bring down inventory. Measured in quarterly sales, days of working capital was reduced to 72 at the end of the year, down from 103 days at the third quarter and down from 89 days at the end of 2017. In the fourth quarter, last October, we completed the second EUR 250 million share buyback program that we started last June. In 2018, in total, we spent EUR 355 million to repurchase 7.2 million of our own shares.
In addition, we used more than EUR 200 million to distribute EUR 4 per share to our shareholders through a tax-efficient capital repayment. As most listeners on the call are aware, we reduced our ASMPT stake during 2017 in two steps from 39% to 25%, which generated total proceeds of approximately EUR 700 million. In line with the commitment we made at that time, we have returned all the proceeds to our shareholders. During 2018, we spent in total EUR 607 million on share buybacks, capital repayments, and dividends, more than double the amount we spent the year before in 2017. Looking at 2019, we plan to increase our regular dividend by 25% to EUR 1 per share. This underpins our confidence in the prospect for ASM. In addition, we will also propose the AGM this year, the cancellation of 5 million shares.
This will reduce the number of issued shares by 9%. Let's look at our strategic achievements in last year. 2018 was also a year of important progress in a number of strategic areas. An important pillar of our growth strategy is to drive up our addressable market in the single wafer ALD space. An important step in this context, as we discussed earlier, was the launch of our new Synergis thermal ALD tool last July at SEMICON West. Customer demand has been strong, and our backlog at the end of the year already included multiple tools. Furthermore, last month, we announced the formal launch of the XP8 QCM tool for high productivity plasma-enhanced atomic layer deposition applications. This newest addition to our portfolio of leading ALD platforms allows for the integration of up to four modules, each containing four process reactors on a single platform.
The XP8 QCM is already in high volume manufacturing at multiple customers. In terms of applications, during 2018, we were selected for multiple new applications for the 5 nm foundry node. In DRAM, we continued our focus on developing new ALD applications beyond traditional patterning. In NAND, we made further progress in a number of applications in which we are engaged, which we expect to contribute step by step once the 3D NAND industry moves to next generation higher stack devices. Next to ALD, as part of our growth strategy, we're also focused on driving structurally higher sales in our other product lines. Epitaxy is a key growth area for ASM and delivered again, a solid performance in 2018. After we started shipping our Intrepid ES system in 2017, we focused last year on supporting the high volume manufacturing ramp of this leading foundry customer.
In addition, we improved our position going from 7 nm to 5 nm. In vertical furnaces and PECVD, we are making targeted investments to enhance also our niche positions in these markets. Let's now have a more detailed look at the trends in our markets. WFE spending, wafer fab equipment spending, further increased in 2018 with an estimate growth of around 10%. While the first half of the year was still strong, overall WFE spending in the second half started to slow down due to a substantial weakening of the memory segment. Investment in the NAND flash segment weakened in 2018 and were particularly lower in the second half of the year.
As we discussed in the last few calls, this was not unexpected as customers made strong investments in both new capacity and 3D NAND conversions over the last few years, for which they now need some time to digest. In the DRAM segment, WFE spending was up for the full year, but also started to slow down in the second half. Advanced logic WFE was up for the full year and even accelerated in the second half, driven by investments in the advanced nodes. Foundry spending decreased, but accelerated in the second half as the mix of spending started to shift toward early investments in the 5 nm technology generation. While WFE spending started to slow down in the second half, ASM clearly ended the year on a strong note.
Our second half year sales increased 22% compared to the first half, in line with the guidance we issued earlier in 2018. While we incurred a decrease in our memory sales in the second half, mainly in 3D NAND, this was more than offset by robust traction in the logic and foundry segments. This is explained by our strong exposure to the logic foundry space. As we indicated earlier, our total non-memory business accounted for clearly more than half of our equipment sales in 2018. We therefore strongly benefited as logic and foundry customers stepped up investments in the second half. Apart from an overall increase in spending, the mix in logic and foundry shifted in the second half towards the most advanced nodes in 10 nm and 5 nm respectively.
Compared to the previous logic foundry nodes, our share of wallet in these newest nodes has increased in a meaningful way as we are engaged in a substantially higher number of layers. This also contributed to our strong performance in the second half of last year. In addition, our analog business also did well. Looking at this year, looking at 2019, average expectations are for the WFE spending to show a year-over-year decrease of a mid to high teens percentage. This is expected to be the balance of a substantial decline in memory spending and a more resilient outlook for logic foundry. With this 3D NAND market still in oversupply, spending in this segment is likely to further drop in 2019. DRAM investments are also likely to drop in 2019 compared to the strong level in 2018.
WFE spending in the combined logic foundry segment is expected to be healthy in 2019, driven by spending on the most advanced nodes. We expect our strong position in logic foundry to support our sales performance during this year. Let's now look at our guidance as included in our press release. For Q1, on a currency comparable level, we expect sales of between EUR 225 million and EUR 245 million, while we expect sales in Q2 to be at EUR 200 million - EUR 230 million level. For Q1, on a currency comparable level, bookings are expected in the range between EUR 200 million and EUR 220 million. For 2019, general expectations are that the WFE market, the wafer fab equipment market, will decline with a mid to high teens percentage, as we just alluded to also. Based upon this current market view, the company expects to outperform the WFE market in 2019.
That's what we wanted to share with you as part of the introduction. At this point in time, like we always do, of course, we are happy to answer any questions you may have, and that applies to Peter, our CFO, and myself.
We'd like to ask you to please limit your questions to not more than two at a time so that everybody has a chance to ask a question. All right, Valerie, we are ready for the first question.
Thank you. Your first question comes from the line of Sandeep Deshpande of JPMorgan. Please go ahead.
Thank you for letting me on. I have two questions, if I may. In your presentation, you've talked about your customer concentration in 2018. Was your customer concentration in 2018, Q3 and Q4, similar to what it was for the full year? The reason I'm asking this question is that based on your guidance at this point in the first half of 2019, you're at the midpoint of the range, guiding to about EUR 450 million of revenue. If that is the case, then you are very substantially outperforming what is happening in the industry. I'm trying to understand how that can happen, in an environment where memory shipments are down so significantly based on what your peers are saying.
Clearly, is it that your customer concentration is very different in the first half of this year compared to what it was in the full year of 2018? Sorry.
Yeah. I think, in general, apart from Q1, I think if you look at the remainder of the year as of Q2, I think it was clear that foundry, logic and analog really stepped up to the plate. I trust that answers your question. That's what I recall. I was not completely prepared for this question, but I think apart from Q1, the picture was pretty steady throughout the rest of the year.
How does this play in the first half of this year?
It was less memory-oriented as of Q2 already. We were very strong in logic foundry from Q2 onwards. The majority of the equipment sales from that moment on was logic foundry-centric.
Thank you.
Thank you. Your next question comes from the line of Tammy Qiu of Berenberg. Please ask your question.
Sandeep, does that answer your question? I just wanted to make sure that He's cut. His line is cut off now?
Yeah.
Okay. Please get back on the line, Sandeep, later if it's not fully clear. Okay?
Thank you. Your next question comes from the line of Tammy Qiu of Berenberg. Please go ahead.
Hi. Thank you for taking my question. The first one is, I have seen that your number has been strong recently. Can you actually comment about your market share? Over the past two years, there has been some discussion about your market share has been going down, in front of the U.S. competitors try to persuade their way into this market. Also secondly, about your gross margin. We have been seeing this low 40% gross margin for a while now. With your top-line growth coming back, are we likely to see more gross margin growth in the future? I.e., to what extent can we actually expect the expansion of gross margin, together with the top-line leverage? Thank you.
Okay. Tammy, thank you for your question. First on the market share. Well, Tammy, it is too early to conclude full calculations on market share for 2018. We can provide you a little bit more color looking at the dynamics in the different industry segments, because the most important element in driving our market share in the short term is mix-related. In that respect, if you look at the full year of As you know, our market share in memory is healthy, but not as strong as in logic and foundry. That is historically the case. If we then look at the full year of 2018, the mix for the full year continued to be a little bit of a headwind.
If you look at the growth of around 10% in the WFE market, for the year as a whole, it was again largely driven by memory. If you look at a little bit more detail, if you look at it on a first-and-second-half basis, then there were, of course, clear differences. In the first half, it was still strongly memory-driven, which was in a way unfavorable for maybe our overall ALD market share. In the second half, memory clearly has slowed down, as we all know. Logic foundry stepped up in a significant way. Not only in general spending, but also in terms of transitioning to more advanced node. That was very favorable for our share.
If you extrapolate that trend into 2019, we all know that 2019, at least that's the way we see it will be more logic foundry-centric than memory-centric, because we think memory is still likely to be down significantly. That will play, in our opinion, in our favor for market share developments in 2019. Okay. Then apart from that, of course, beyond 2019, as you know, we are seeing evidence of that now as the market transitions in logic to 10 nm and in foundry to 5 nm, but also as we go to 7 nm and to 3 nm. We are aggressively working on new projects to further expand our served available market in single-wafer ALD, that will ensure that we can protect our market share in the right way going forward, also beyond this year.
When you look to the gross margin discussion, Tammy, basically having a look at it, two major reasons for that. The first and utmost important one is the introduction of a lot of new products. That has also driven our growth and driven our order book. When you have all those new products coming in, then there is a combination of a few extra costs that you have. You have initially the evaluation costs, so the tools which are at your customer, where you have to pay the depreciation on. Especially when you have new products, you have two initially, in a lot of cases, especially on complex products, two additional extra costs. That's the first install and qualification costs. They are mostly higher for completely new products than for existing products.
The other one is when you do all those new product introductions, initially, that might lead to changes, last-minute changes, which putting a little bit more pressure on the first tools that you're going to deliver. That, I think, is the most important element. Due to what I already mentioned earlier, the very high volume of new product introductions that we have seen, that has impacted our gross margin in the course of 2018, and that was also the main reason for all the fluctuations. Beside that, we have a second element. That was especially visible in the first half of the year. Our customers came very often back last-minute with changes, that led to pull-ins or push-outs for certain products. As a consequence of that, we had quite some inefficiencies in our operating organization.
You can imagine that you have, on that moment, the people available, but last-minute, the materials were not available, so the people have to wait for that. That had also a certain impact on the gross margin. I think the two elements that we have seen in 2018, last-minute changes, a lot of new product introduction, were unique for 2018, but especially the new product introductions will continue also in 2019. Hence the early remark that Chuck has made, that we also expect in 2019, that we will have somewhat more fluctuations.
Okay. Thank you. One follow-up, if I may. Regarding market share, I noticed that your Japanese competitor has been talking about they are now the biggest market shareholder in the ALD market, and I know they only do the mini batch tool. However, they said their mini batch tool fits into a lot of the chip makers' criteria because it's actually more efficient compared to the single-wafer tool. What is your view from a market demand from end market perspective for single-wafer and multi-mini batch, multi-wafer tool in general?
Yeah. I don't know what specific tools you are referring to, but our comments are always related to the single-wafer ALD market. Of course, next to that, there is a batch market. A batch market, if you look at the last few years when NAND was really growing, a lot of spending was done in NAND. Quite some investments were done in batch tools. We have seen in those years that the batch market next to single wafer has been benefiting from that. Yeah, our comment that I just made on the single-wafer ALD market are not any different, given your comment on a competitor with a mini batch tool. Our comments are not different in any way.
Okay. Thank you.
I'm happy, Tammy, on maybe offline, through Victor, to discuss this a little bit more. Again, the general comment for the whole audience is not going to change based on your comment of this competitor.
Okay, cool. Thank you so much.
Yeah. You're welcome.
Thank you. Your next question comes from the line of Peter Olofsen of Kepler Cheuvreux. Please go ahead.
Good afternoon. My question is on the outlook for logic foundry demand. Obviously a very strong Q4. Looking at the Q1 and Q2 sales guidance, we are seeing sales are trending slightly lower sequentially. Does this mean you see logic foundry demand gradually leveling off as the year progresses? As such, is there a risk that H2 could be materially weaker than the first half? Do you see a fairly steady logic foundry investment throughout the year?
Like we tried to say before, if you look at the outlook for the second half, of course, we are not able to give sales guidance for the company specific, but if you look at the climate for the industry segment in general, Peter, we can say, first of all, let's first start for the company specifically. If you look at, we started, of course, the year 2019 with a record high order book. If you look at our guidance for Q1 and Q2, at the midpoint, that's EUR 450 million. If at the midpoint, sales in the first half of 2019 are expected to be only marginally below the strong level in the second half of 2018. Up by about 20% compared to the first half of last year.
If you look at the trends in the different segments, we expect WFE in logic and foundry, it's expected to be solid in the first half. It is possible that spending in the second half in logic foundry will be somewhat lower. With the visibility we have now, still we foresee it to be at a very healthy level in the second half. Next to that, if you look at analog. Analog, of course, growth in the latter part of 2018 has been very strong, the outlook for the first half of this year is also healthy. It's likely that demands for analog in the second half will be lower, at least based on the visibility we have today. Next to that, because it's probably good to give you a complete view, if you look at memory.
Memory is weak in the first half, investment levels are especially weak in 3D NAND, this moment in time. Also tracking meaningfully lower in DRAM in the industry is our perception. Based on current visibility, it's possible that DRAM demand will be slightly higher in the second half. That's very well possible. Again, that's just an estimate based on our current visibility. For NAND, visibility is lower than for DRAM. Based on that, it's not impossible that there will be some recovery in spending in 3D NAND, coming from a very low base, by the way, today, towards the end of the year. That's a little bit too early to tell at this moment in time. I trust that provides some good sense.
DRAM, possibly to show some improvement, 3D NAND, maybe towards the end of the year, but too early to tell. Logic and foundry, staying healthy throughout the year, but maybe at a somewhat lower level, in the second half compared to the first half. That's the way we see it today, Peter.
Okay. Yeah, thanks for that very extensive color. Maybe a follow-up on logic foundry. You talked about share of wallet gains with the move to the most advanced nodes of 10 nm and 5 nm. Were you specifically referring to ALD? Or is it also referring to Epi, for instance?
Well, we are growing in foundry. We are growing our participation in Epi, going from 7 nm to 5 nm. It's in a modest way, but the level of engagement is growing in terms of applications. That's a great step forward. Our aim is that it doesn't stop there. We're engaged heavily to further increase that participation towards future nodes. The participation has increased, going from 7 nm to 5 nm. In ALD, as we said in the introduction, in both logic and foundry, the number of layers that we have, are engaged in, has gone up in a significant way. That's what we have always been telling, that we were focused on that, to make that happen. We finally see the fruits of all that hard work of our employees at those customers.
Okay. Thank you.
Yeah. You're welcome.
Thank you. Your next question comes from the line of Stéphane Houri of ODDO. Please go ahead.
Yes. Hello, good afternoon. My questions are about the level of OpEx and notably R&D, which has increased in the last quarter. You have alluded to a one-off. The question is, what is the right level for R&D on a quarterly basis going forward? Here I have the same question for G&A, so it's more to understand where to put the OpEx level in the next quarter. Thank you.
Thanks for that question. First of all, let's start with the R&D. As we always have said, our R&D is expected to be in the low to mid teens. When you look now to what happened between the third quarter and the fourth quarter, when you look to the net increase that we have seen there are basically two major reasons for that. The one, and that's visible as well in R&D as in SG&A, we moved operations from within Korea to a new office and a new lab environment. As a consequence of that, we had in the existing operations that we had, we had to write off certain CapEx. That was a few millions. Distributed over R&D and SG&A, that's really a one-off. In R&D, we also have seen this quarter that we had a lot of bare wafers coming in.
In our development area, we need wafers so that we can do test runs, et cetera. They are not evenly distributed, coming in in our fabs or in our lab environment quarter by quarter. We had rather high invoices for those bare wafers in the fourth quarter. It's a normal cost for the year, but not normal cost for the quarter. That is as far as R&D is concerned. When you look to SG&A, SG&A was increasing quarter-on-quarter with EUR 4 million. When you would take there the incidents and the one-offs off, then the increase would have only be EUR 1 million. The other EUR 3 million is related to that depreciation that I just mentioned.
The write-off in the Korea area, or to the additional legal costs that we have for the cases that we are running or some extra amortization costs that we took on IT systems. I think that gives you the color where you are looking for.
Okay, thank you. I have a follow-up, if I may, about the contribution of ASMPT. You have said that basically there were some one-off costs also in the profits of ASMPT. I'd like to understand what we can foresee for the near future for this contribution. Thank you.
Yeah. The details of that, of course, should come from PT in itself. What we have learned from the discussions that we had with them is that they have written off certain activities. They bought, a few years ago, a new activity, and that included some solar activities, and they have looked at that, if that could be of interest or not. They concluded in the fourth quarter that it was not worthwhile to do that, and the impact of that was around, I believe, the EUR 10 million. That's an important impact. They acquired NEXX last quarter, that had also some impact on their margins. I think that's the color that they have provided also towards the outside world.
Okay. Thank you very much.
Thank you. Your next question comes from the line of Quang Le of Credit Suisse. Please go ahead.
Hi. Thank you for taking my question. Most of them were already answered. Coming back to your question regarding the contribution from ASMPT. We see it was significantly lower than what consensus expected and also versus previous quarters. My question is, what's your plan from here? Are you still going to hold it as I said, it was in your strategic plan, or are you maybe willing to divest it more?
Yeah. As we go to the 25% stake, our position has not changed compared to earlier statements that we made. I trust that you're familiar with the earlier statements that we made, and that it's just a strategic stake that we still have in that company and that on a regular basis as part of our strategic planning, we review the stake and the importance of the stake, and at the same time, the development of the valuation of the company.
Got it. Thank you.
Yeah. You're welcome.
Thank you. Your next question comes from the line of Marc Hesselink of ING. Please go ahead.
Yes. Thank you. Marc, first question is on the comment that you made that you see the logic foundry market for ALD doubling
Between the 14 nm, 16 nm and the 7 nm. I want to make sure that I completely understand that comment. Is that the number of layers doubling in that transition? Because 14 nm, 16 nm was obviously a very large node. Does it mean that if 7 nm is similar size node, then it's like the total market doubled, or is it just talking about those layers? The second question is on your spares and service business, which also grows decently last year. What are your expectations there going forward? Will it continue to grow faster than the rest of the business in the medium term? A short question on the CapEx. You already explained that the trends moving in the OpEx are in the SG&A, also same question for the CapEx. You already said that 2019 will still be a relatively high one, what will happen afterwards?
Okay. Marc, on the logic foundry going from 14 nm, 16 nm to 7nm. The comment just meant that how significant the contribution from logic foundry is to the single wafer ALD market. That going in, like we always said, that we expect as we continue Moore's Law and we go to smaller geometries, that then more and more conventional deposition technologies will run out of steam, just from a physics point of view, and that the customers are forced for more and more applications, and especially layers, to go to ALD technology. The contribution from logic foundry to the single wafer ALD market more than doubled in that time frame. As a result, also, of course, the contribution to ASM has increased also significantly.
A big proof of that is shown now as logic transitions to 10 nm in high volume, foundry transitions to 5 nm in high volume, or at least in pilot production as we speak.
The question about spares and services, Marc. First of all, we think that indeed the spares and service business will grow in the years to come, albeit a little bit delayed, because of course, when we are installing a new capacity with our customers, then there is a warranty period. After that warranty period, normally spares are popping in. We think that will remain a healthy growth activity going forward. When you look to CapEx, we have seen in 2018, what we already mentioned in the prepared notes, also in 2019, we will see some additional investments.
2018, that was driven by the Korea facility, what I mentioned earlier, so that the investments took place during 2018. We got that new fab of that new lab environment in October 2018. We moved in. On this moment, those costs are gone, but were impacting the 2018 CapEx. In 2018, we also bought access to land in Singapore, and the plan now for 2019 is to build a complete new fab, preparing ourselves for the next growth, as what we have mentioned earlier. That highly likely will be delivered in early 2020. With that, we have done, I think, the big investments, what we have indicated a few years ago. I would expect that after 2020, or in 2020 already, the CapEx will normalize more towards the level that you have seen in the years before we started those big CapEx programs.
Okay. Thank you.
Thank you. Your next question comes from the line of Robert Sanders of Deutsche Bank . Please go ahead.
Yeah. Good afternoon. My first question was just on your new products in ALD that you've launched recently. You had a follow-up to the Eagle XP8, and some other products. It'd be great just to hear what the reception has been and what that does to your ASP and your gross margin. Then I have a follow-up. Thanks.
Yeah. Okay. Robert, thank you for your question. You're referring to the XP8 QCM. Well, reception has been very good. As we shared in the introduction, we are engaged with multiple customers in high volume. That's in basically all segments. We're engaged in DRAM, 3D NAND, and foundry with this product. That's an indication that customers are really willing to adopt this tool. Of course, we are seeding with more applications, more processes, with this tool towards future nodes in the different industry segments. That's from a top line, from a market development point of view.
Would you like to say something, Peter, about the margin?
Margin development, yeah. I think ASPs are healthy, Rob, that's not the issue. When you look to the margins, I can only repeat the remark that I made earlier. In those initial phases, you have some extra costs, we expect that those extra costs will disappear on the moment that we are installing more equipment with existing customers. Moreover, as I said earlier, due to a lot of last-minute changes, our procurement organization can step up on the plate on this moment when the volumes are further increasing, also to put somewhat more pressure on the total supply chain. That's for sure going to happen.
Yeah. Okay. That's great. Go ahead.
This tool, Rob, it's a great product, again, as part of our whole program to expand our served available market in ALD. We do that on the process side, and we do it on the equipment side, because everything counts. It's functionality from a process point of view, but it's also productivity, cost of ownership, and let's say, hardware capabilities from a platform point of view.
My follow-up would just be, on the early pilot production you're doing on N5 with foundries, how are you thinking about your ALD intensity and patterning versus non-patterning? There had been a worry out there that EUV adoption, the early starting of EUV adoption would lead to more reuse on that side for self-aligned patterning. Are you seeing that impact, but you're just seeing a whole lot more gate level usage of ALD, or is it not so simple?
Yeah. Your question mainly, because there are a few elements part of your question, but the main part is what's the impact of EUV?
No. I'm just thinking, when you look at N5 and you're saying you're seeing a lot of traction with ALD, single wafer ALD at N5, is that mainly driven at the gate level, where you're seeing either replacement of CVD for ALD or new applications? Is that really where you're seeing the big jump intensity?
Yeah, it's a mix.
Okay.
It's a mix of increased patterning demand and also new applications. Really, it's on both ends that we see developments.
Okay.
Yeah. If you look at the logic foundry space combined.
EUV hasn't really come through as a sort of negative impact on your patterning yet because of all the immersion tools that are out there that need to do self-aligned patterning. That's basically it.
If you look at the logic foundry combined, without being specific by segment, you see an increased demand in both patterning and non-patterning.
Okay. Thank you.
Okay, you're welcome.
Thank you. Your last question comes from the line of Nigel van Putten of Kempen. Please go ahead.
Hello?
Hello?
Oh, hi. Sorry about that. My questions have been answered. Thank you.
Okay. Well, thank you for attending.
Thank you. That does conclude the question and answer session. Mr. del Prado, the floor is yours for closing remarks.
Well, I would like to thank you all, as we always do, for attending today's call, especially on this late in the week, on a late Friday afternoon, early in the day for some of you people maybe that listen in from the U.S. Thank you again. Looking forward to meet a lot of you face-to-face in one of our roadshows in the coming weeks. Feel free to follow up also with us through Victor on any questions that have not been answered today. Thank you very much again, also on behalf of Peter and Victor. Have a good day. Good evening. Thank you. Bye-bye.
Thank you, sir. Ladies and gentlemen, that does conclude the earnings call. Thank you for participating. You may now disconnect.