Good day, and welcome to the ASM International Q2 2018 earnings call. Today's conference is being recorded. At this time, I would like to turn the conference over to Mr. Victor Bareño. Please go ahead, sir.
Thank you, operator. ASMI issued its 2018 second 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, asmi.com. As always, we remind you that this conference call may contain information relating to ASMI's future business or results, in addition to historical information. For more information on risk factors related to such forward-looking statements, please refer to the 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. Thank you all for attending today's ASMI Q2 2018 conference call. Of course, we'll start with some introductory remarks and subsequently, of course, together with Peter van Bommel, our CFO, we're able to answer any questions you may have. Let's now first review our second quarter 2018 financial results. Net sales in the second quarter increased to EUR 209 million, an increase of 31% compared to the first quarter. Sales in the quarter were in line with our guidance, which was a range of between EUR 200 million and EUR 230 million. On a constant currency basis, sales increased year-on-year by 3% in Q2, and by 6% in the first half of this year. Equipment sales increased 34% sequentially.
Service and spare parts accounted for 23% of total sales in the second quarter and were up 16% year-on-year in the first half at constant currencies. In terms of product lines, our ALD business was again the key driver. We also had a decent contribution from the other product lines. To highlight one example, in furnaces, we have seen solid sales levels in Q2 and also in Q1, driven by advanced logic as well as so-called More than Moore applications such as sensors and IoT products. Compared to Q1, sales in all industry segments were higher, led by strong increases in logic and foundry. Within memory, sales were at healthy levels in both DRAM and 3D NAND. The gross margin was 42% in the second quarter.
In line with our expectation, gross margin improved substantially compared to 38% in the first quarter and returned to the target range of low to mid-40s. The sequential improvement in gross margin was a result of a few factors. First of all, moderation of the impact from new products from 2% in Q1 to one percentage point in Q2. Secondly, the sales mix of existing products compared to Q1 returning to the normal levels as seen in the past years. Third factor, was the reduction of the cost associated with our preparations for expansion. We had a strong pipeline of new products and applications that we are going to introduce in the forthcoming periods, and that will add to ASMI's growth potential. It is possible that some of the next quarters will show the impact of higher associated installation expenses of these new introductions.
That can lead to a somewhat higher volatility in gross margin than what we have seen in previous years. Having said that, we expect the gross margin to remain in the low to mid-40s range. SG&A expenses increased by 9% compared to the first quarter. While R&D spending increased quarter on quarter with 5%, the reported R&D expenses decreased with 5%. This is due to the higher R&D capitalization as more of the new products and applications that we have been working on are getting closer to the point of commercialization. We generated operating income of €38 million in the quarter, with an operating margin of 18%, up from 7% in the first quarter. Financing results in the second quarter included a currency translation gain of €8 million compared to a loss of €8 million in the first quarter.
As a reminder, we hold the substantial part of our cash balances in US dollars. Net earnings, on a normalized basis, amounted to €62 million, up from €19 million in Q1. In Q2 of last year, our normalized net earnings amounted to €61 million, excluding a book profit of about €100 million, related to the sale of a 5% stake in ASMPT in that respective quarter. Let's take a look briefly at ASMPT. ASMPT delivered again, solid results in the second quarter. Normalized results from investments, which reflect our share of approximately 25% in the net earnings from ASMPT increased to €22 million for the quarter, up from €16 million in Q1. We're talking about euros here.
In the second quarter, ASMPT's sales rose to $672 million, up 21% from the first quarter and up 19% year-on-year, and in line with the company's earlier guidance of somewhere between $650 million and $710 million. Following the strong level in Q1, bookings in the second quarter increased 10% year-on-year, slightly above ASMPT's guidance of single-digit growth. Let's now turn back to ASMI's consolidated operations. Our order intake in the second quarter amounted to €176 million, down from the record high of €207 million in the first quarter. As such, well within our earlier guidance of a range between €160 million and €200 million. Looking at the breakdown in bookings by industry segment, foundry was the largest segment, followed by memory. Foundry orders increased versus Q1 and were driven by the continued ramp of seven nanometer, as well as some early demand related to the next five-nanometer node.
Within memory, DRAM accounted for the largest part of the order intake, and NAND bookings, after having been at solid levels for multiple quarters, slowed compared to Q1. Taking a look at our balance sheet and cash flow. At the end of June, the cash position stood at EUR 651 million, down from EUR 741 million at the end of March. This decrease was largely the result of a total of nearly EUR 80 million spent on dividends and share buybacks, as well as a negative free cash flow during the quarter, and partly offset by EUR 15 million in dividends received from ASMPT. The free cash flow turned to EUR 37 million negative in the second quarter, compared to EUR 15 million positive in the first quarter. This is primarily the result of working capital, which increased from EUR 156 million at the end of the first quarter to EUR 213 million at the end of June.
The largest part of this increase was in turn caused by accounts receivables, which rose by EUR 42 million due to the strong sequential increase in sales and by the fact that Q2 sales were heavily back-end loaded. In the second quarter, we paid EUR 37 million in dividends or EUR 0.80 per share. We also spent EUR 39 million during the quarter to buy back more than 700,000 of our shares at an average price of approximately EUR 52. These share repurchases were primarily related to the EUR 250 million share buyback program that we announced on June 5th of this year. As of last week, we have completed 31% of this program. As communicated earlier, this share buyback program is funded with part of the proceeds of the 9% stake sale in ASMPT that we executed in November last year.
We will use the other part of the proceeds for a tax-efficient capital repayment of EUR 40 per share, which was approved by the AGM in May. The indicative ex-date for the capital repayment is August 7. The AGM also approved the cancellation of 6 million treasury shares, which is expected to become effective as of August 1st. Let's now more closely look at the dynamics in our key markets. Despite the number of macro-related risks, such as potential impact from trade tariffs, overall conditions in the semiconductor end market still look healthy, supported by broad-based demand drivers such as cloud computing, data centers, automotive, and industrial applications. The total IC market is expected to grow low double digits this year.
Looking at the WFE market, the wafer fab equipment market, while 2018 is still expected to be a growth year, the increase this year is likely to be somewhat lower than expected earlier in the year. That's due to the lower than previously expected spending in the second half of the year, in primarily the memory sector. General expectations are now for the WFE market to increase by a mid- to high single-digit percentage in 2018 in U.S. dollar terms. Drivers behind the growth this year are still expected to be the logic and DRAM segments. For ASM, we continue to expect sales in the second half to be higher than in the first half of this year. As we indicated in April, expected higher sales in the second half will be driven by logic foundry and DRAM, while NAND sales are still perceived to be lower sequentially.
For the full year, we still aim to outperform the WFE market. On balance, WFE expectations for the logic foundry markets have remained largely unchanged. While the mix of spending on the different nodes this year now looks somewhat different than earlier assumed, overall spending within advanced logic is still expected to be up this year. In foundry, the spending in 2018 is related to the ramp of seven nanometer node as well as the first investments in the next five nanometer node. Investments for the five nanometer node are likely to further increase in the course of 2019. As we highlighted earlier, we believe the five nanometer foundry node is going to be an important transition for our company.
Looking at the intensity of R&D engagement, we see very good opportunities to expand our position to multiple new applications across our ALD portfolio, and to increase our overall share of wallet in the foundry sector with the transition to this more advanced node. Turning to 3D NAND. Growth in end market demand continues to be healthy, but following strong levels of spending in the 3D NAND space in the last couple of years, WFE spending in this segment is expected to decrease in 2018, mostly impacting the second half of the year. Customers are preparing for the ramp of the next generation devices, which are likely to see volume buildouts in the course of next year. Looking at these next generation devices, we feel we are well-positioned.
As customers move to higher stack devices of 90 layers and above, the increasing complexity and aspect ratios will gradually fuel the need for a growing number of advanced single-wafer ALD solutions. Over the last few years, we have invested in new solutions and hardware improvements that will increase the part of the 3D NAND ALD market that we can address, both in terms of existing and new applications. In the DRAM segment, overall growth in WFE this year is still expected to increase even though some investments have been pushed out to 2019. Technology transitions continue, customers are making relatively higher investments in new capacity this year, which is driving for us a healthy increase in new ALD tool demand for double and quadruple patterning. In R&D, we remain focused on broadening our portfolio of ALD in this DRAM segment.
We continue to expect the first contribution of new non-patterning solutions in next year. Looking at the longer term, we remain confident about the double-digit growth potential of the ALD market. As the industry continues to transition to smaller geometries, new materials, and more complex device architectures, more and more deposition steps are expected to require the superior film precision and control offered by advanced single-wafer ALD with lower temperature processing and high levels of productivity. An important element of our growth strategy is to expand our SAM, our served available market, within the total ALD market, as we shared with you before, including parts of the markets that we have so far not yet addressed. An important step in this context is the introduction of our new Synergis tool that we recently announced at SEMICON West in San Francisco earlier this month.
The Synergis combines the core ALD capabilities of the Pulsar and EmerALD reactors, with the proven high productivity of our proven XP8 platform and includes innovations such as optimal thermal control in a low volume reactor with improved purge efficiency, contributing to excellent film uniformity and chamber repeatability in a lower cost of ownership configuration. Furthermore, Synergis supports a wide range of films, film properties, and material compositions, and as such, will enable a meaningful expansion of our served available market. We have R&D engagement for our Synergis tools ongoing with several customers. Looking at the longer term, next to ALD, we also expect a structural higher contribution from our combined other product lines. In epitaxy, our Intrepid systems are currently supporting the HVM ramp, as you know, of the leading foundry customer in the world.
For the next device generations, we continue to be engaged in R&D with multiple customers. Let's now close this introduction with looking at our Q3 guidance as we included in our press release. For Q3 on a currency comparable level, we expect sales of between EUR 180 million and EUR 200 million, and an order intake between EUR 200 million and EUR 230 million. Q3 reflects still some uncertainty around the exact timing of individual tools. At this time, we are more than happy to meet together with Peter van Bommel to answer any questions you may have.
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. Okay, operator, we are ready for the first question.
Thank you. If you would like to ask a question, please signal by pressing star one on your telephone keypad. If you are using a speakerphone, please make sure your mute function is turned off to allow your signal to reach our equipment. Again, press star one to ask a question. We'll pause for just a moment to allow everyone an opportunity to signal for questions. We will now take our first question from Kwan Lee from Credit Suisse. Your line is open.
Hi.
Please go ahead.
Hi. Thank you for taking my question. Chuck, if I could move to your comment. You said that you remain confident about the double-digit growth in single wafer ALD market this year. Am I correct? Could you clarify that?
Yes. Thank you for asking this clarifying question. It's more the, let's say, the double-digit calculated CAGR over, let's say, a four or five-year period of time. Because it's reconfirming our statement that with 2017 as a baseline, we are of the view that the single wafer ALD market has the clear potential to grow to around EUR 1.5 billion by 2021. That's what we meant with that statement. Thank you for asking this clarifying question.
Thank you. In that case, what is your projection for this year and next year growth, then? Do you have anything in mind?
Yeah. If you look at the ALD market, it's of course a little bit too early to do a complete forecast for the year to estimate, forecast the complete outcome of the year. If you look at it by industry segment, then we expect that the logic contribution, the advanced logic contribution to go up this year. We expect the DRAM contribution to the single wafer ALD market to go up. Our opinion now is that we expect the NAND contribution to go down, and likely the foundry contribution also temporarily to go down also. That's I think from an ALD point of view, I think our current view.
I see. If I could clarify, you also said that-
At the same time, if you then look at next year. Next year, if advanced logic is able to re-accelerate the ramp of 10 nanometer, and foundry really steps up again on investments N7 plus, but especially N5, there is a big opportunity for those segments to develop very well going into 2019. That's also a very interesting area for us to be focused on and to work on increasing our share of wallet.
I see. If I could also move back to your comment that all your segments have been increasing year to year in this quarter. Do you see any market share gains, especially in single wafer ALD, EPI, or PECVD, those segments where you are focused on?
You mean market share?
Yes
in this year?
Yes, this year. I mean year to date, that is.
Yeah. I think that's way too early to, because we cannot compare it with publicly available data because we always build also our intelligence on combining our own assessment with publicly available data. You can also do an assessment yourself, because our market share is mostly a function of overall mix of spending. You know that, for example, in 2017, our market share decreased somewhat because memory was the main driver of the industry. You understand that we are better able to, let's say, maintain and develop a solid market share when logic foundry, compared to memory, is doing well. Our relative market share in those segments is the strongest.
That's the color we can provide at this moment in time, for the rest, for now, we invite you to make your own assessment, as soon as we think we can provide more color on that, we will.
Thank you.
Okay. You're welcome.
As a reminder, if you would like to ask a question, please press star one. We will take our next question from Tammy Qiu from Berenberg. Your line is open. Please go ahead.
Thank you for taking my question. Firstly, can you talk about your comment regarding memory spending is lower than expected in H2 this year? Is that more of a push-out or a cancellation of spending? If it's actually a push-out, are we going to see 2019 higher than expected? Also, my second question is: Can you talk about your positioning in the new 96-layer generation 3D NAND, which was released literally, a week or two ago? Thank you.
Okay. First of all, memory in general, I think we have been pretty consistent on the memory market. Since the beginning of the year, our visibility was that NAND market was clearly weakening in the course of the year. Compared to April, it maybe has weakened a little bit more for the remainder of the year. On DRAM, that we foresaw clearly, and we still see that year-on-year, DRAM is clearly stronger. Also there in the second half, certain players have slowed down their spending. Since most of those players only place their orders relatively late, for us, those are push-outs and no cancellations. In general, we are of the opinion that, looking at 2019, because that's maybe what you're also looking for.
It's of course very early, the common opinion is that DRAM after two strong years may slow down year-on-year in 2019. NAND, initially, a lot of people were pretty negative on 2019, very aggressively negative on 2019. We said earlier that in talking to our customers, we don't see that same pattern coming back from our customers, especially since some of the slowdown in NAND already kicked in basically in this year. It could result in a soft landing early next year and then maybe even a pickup sometime next year. Again, we're not predicting that, but we don't exclude that from happening. Then for the rest, as I alluded to in the earlier question, for 2019, we see a big opportunity for Foundry to pick up driven by N7+ and N5, N3 spending.
Advanced logic could also be up if 10 nanometer really would re-accelerate. Coming back to your second question on 96 stack, I don't know exactly what you mean, we have seen the industry preparing for 96 stack already since early in the year. We have not seen a strong ramp there yet, and we don't foresee a strong ramp of volume there until early next year based on, again, the visibility we have today. Yeah, we have been positioning ourselves with certain applications in 96 stacks at multiple customers. Let's say the programs that we have been working on with customers, those programs are going very well. As a result, as we said before, we aim to gradually increase our presence in 3D NAND as we go from stack to stack.
We are also aiming, as we said, to expand our presence in, basically, build a presence in the leader in this industry, gradually as of 96 stack.
Okay. I see. What I meant by 96-layer was that previously you were involved in some of design wins with 64-layer 3D NAND. One of the biggest memory maker wasn't actually using single-wafer ALD at 64-layer. You were commenting that you were working with them on the next-generation, a 96-layer. I'm just wondering, have they started adopting single-wafer ALD at 96-layer design?
We are, let's say, on that specific application, again, they have not ramped volume significantly. What we can say is that that engagement is very much on track. For us, that is a very strategic path to get in, for the customer to familiarize with our tool, with our process capabilities, and then gradually further build our presence within that customer. Okay?
Okay. Thank you.
Yeah, you're welcome, Tammy.
We will take our next question from Mr. Robert Sanders from Deutsche Bank. Please go ahead.
Yeah. Hi. I have two questions. Firstly, on EUV at Foundries. Clearly, TSMC is a large customer. They are adopting EUV or trying to in a big way at five nanometer. How is that impacting your single wafer ALD patterning business with them? Have you seen an impact already? Are they reusing tools more? I'd love to get an idea of, have you seen any impact and how you see that going forward. Just a kind of naughty question. Probably this is a stupid question, but I'd love to know a bit more about the mechanism behind this capital return. What does it mean for our model, et cetera? Did the share count go down? I'd love to learn a bit more. Frankly, not seen a lot of these in my career. Thanks.
Okay. First of all, yeah, EUV. Don't want to speak too much about specific customers, but I think in general, with the introduction of EUV in general, we really embrace it. First of all, we expect our ALD patterning business to remain a really steady business for the foreseeable future. As we said before, percentage-wise, in the growth to EUR 1.5 billion, likely, we did not foresee the patterning business percentage-wise to significantly grow, in all our projections. That has not changed. The absolute contribution will stay a steady one, regardless of perceived introductions or planned introductions that have been announced recently. Yeah. In general, within the foundry space, we see tremendous opportunity to increase our share of wallet, going from N7 to N5 in that industry segment in general, Rob. Okay, so that's one, Robert?
Sorry, could you clarify a little bit more about that? It does seem like they're not going to be adopting Gate-All-Around. You're not saying the non-patterning driver at TSMC is not the transistor gate type stuff? It's something else. It's a replacement of CVD or something else.
Yeah, we have multiple areas. Multiple areas, multiple product engagements with foundry to increase our share of wallet going from seven to five.
Okay.
Across the board, we are working with them.
Okay.
Across the board, multiple products to increase our presence in the foundry space. Yeah, going to the next technology node, those corporations are very, very intense from both sides. One last thing, because basically your question mixed two things. EUV in general, I think, EUV will be successful in general. We cannot, let's say repeat that enough. It will enable the industry to further continue Moore's Law. That is, we view that since we are an innovative company that tries to be at the leading edge of the deposition landscape with many of our technologies, many of our equipments, many of our materials, chemistries. It opens up more opportunities for us to grow our business down the road. Okay? That's the answer to that question. Peter, will you take it?
Yeah, let me try to answer your question with regard to the capital return. It's rather a technical issue. First of all, what we are doing on this one, we have, of course, nominal capital, and we have paid up capital in excess of that nominal capital. We're following basically a two-step approach, which has also been approved by the AGM in May. The first step is in principle that we add the paid-up capital in excess of nominal capital towards the nominal capital. You increase in principle the nominal value per share. Then in the second step, you reduce that nominal capital again for the same amount as what you have done, that you have increased it out of the paid-up capital in excess of the nominal capital. By doing so, you are in principle returning capital towards your shareholders.
You can do that tax-free. The only thing, what you have is a waiting period of slightly above the two months because creditors of the company could disagree with that. That waiting period ends early August, and directly after that period, we will pay out that EUR 4 to the outstanding shares in the company.
Got it. From a modeling point of view, is that basically equivalent to the share count going down accordingly by that amount and a cash out of that amount? Just to ask a stupid question.
The share count will remain the same.
Okay. It's modeled like a special div, basically.
It's a principle that the combination of the nominal value of the capital and the paid-up capital in excess of the nominal capital. When you add those numbers together on this moment in the books.
That will reduce towards simply the nominal capital per share again.
Got it. Okay. I'll follow up maybe offline. Thanks.
Thank you. We will now take our next question from Edwin de Jong from NIBC. Please go ahead.
Good afternoon, gentlemen. A couple of questions left. If we take your guidance of above market growth for the full year, and then look at the Q3 guidance as well. Q4 should be very strong, well above EUR 200 million quarter again. Could you maybe explain a little bit what is driving that, and can we maybe also have some comments on what order intake looks like at the moment?
Your first question was what the drivers are for this ALD market to keep growing?
No, just the general sales guidance. You say you're going to grow above market for the year, so let's say 8%, 9% or so. You have a Q3 guidance of around EUR 190 million. If we combine that, Q4 should be very strong.
Yeah. Okay.
In terms of revenue.
You're talking through the remainder of this year. Okay. First of all, you are maybe combining a few facts, which I understand from your point of view. We have not given specific sales guidance for the full year of 2018. I would like to make that clear. Your question basically is whether we anticipate a strong fourth quarter based on our guidance for Q3 and for the full year.
Yes.
Well, we can confirm that we do expect a strong fourth quarter, as of course implied by our guidance. If you look at the backlog of June, it was lower sequentially, but still very healthy. If you combine that with our guidance on the Q3 bookings, we would end up with, of course, a very healthy backlog at the end of Q3. Then you can understand what that could mean for, let's say, Q4 revenue. If you look at the drivers for our Q3 order intake, logic is foreseen to be a strong contributor. Analog is foreseen also to be a healthy contributor. Also foundry to a certain extent, Q on Q, we think will step up. Just to provide you some color on what we took into account in getting to our current guidance for Q3. It's just not a number.
It's based on certain trends that we experience in specific industry segments of the environment that we work in. Yeah?
Okay.
So-
Sorry.
No, that was my answer. Any follow-up?
Follow-up will be on the Synergies. Do you already have some first shipments for the new tool? Can you maybe give some idea of the specific markets you're targeting with the new tool and then how it's being received by the market? Maybe some general comments.
Yeah, general comments. Of course, we always have had with our Pulsar and EmerALD reactors, we really focus on this market. So far, we have had four reactor platforms that we used for that market. We really worked very hard over the last two, three years to do two things. To repackage all the know-how we have on the ALD aspects that are incorporated in these reactors in an eight-tool, eight-reactor configuration to really improve, to make the cost of ownership more attractive. Secondly, take all the learnings from all the process and tool learnings that we built during those ALD engagement with customers that we've had over the last few years to also incorporate those in this tool. We think that for specific application, not for every application, for certain applications the current configurations are still very well valid.
Are maybe more suitable, depends also on the application itself, the speed of the process, wafers per hour, and the process itself without going into too much detail. We basically, by bringing this configuration to the market, we really have a tremendous opportunity to increase our served available market within the ALD space. That's what we said all along in the last few quarters, that we are aiming to increase our served available market within that position. With EPI, with, to a certain extent, with PECVD, to also a certain extent with vertical furnace. PECVD and vertical furnace in a more modest way than in EPI, but also in ALD. Regardless of the fact that we are the leader, still, we said many times, we don't address all the areas of the business.
We don't address all the applications in logic, don't address all the applications in foundry. We have significant areas of opportunity in memory. This is an important step, not the only step, but an important step to increase that served available market, to protect our leadership in ALD. We see a tremendous interest among multiple customers. We aim to go into high volume with this tool next year. We aim to go into high volume with this tool next year. Yeah. That's our aim, again. So that's our ambition.
Okay. That's mostly targeting memory market, I think. Or am I wrong there?
Well, it will be gradually. It will gradually be, and likely the initial traction will be in the logic foundry space. Then subsequently, in memory, it may take a little bit more time, but we have engagements ongoing there. Our belief is that likely we will build first traction the most in logic foundry space.
Okay. Thank you.
You're welcome.
Thank you. Our next question is from Charles Elliott from Inflection Point Investments. Please go ahead.
Hi. A quick one. We're hearing a story that your Japanese competitors producing CVD are not getting enough aluminum capacitors, and that in turn may be responsible for the pushback in DRAM investment, in planned DRAM CapEx spend into next year by one of the major Asian clients. Could you give any comment? Hello? Did that get through?
Yeah. I have difficulty giving comments to this question. Maybe offline, we can sort out if we can respond to this question better. Maybe let's have additional dialogue outside the meeting with Victor, because I do not recognize this comment immediately, so I'm not able to comment on it. I can speculate, but I don't think that will help the audience today. Sorry about that.
That's fine, Chuck. Thanks.
Okay.
Okay.
Thank you for bringing it up. Let's please follow up offline with Victor. Yeah.
Our next question is from David O'Connor from Exane BNP.
Great. Good afternoon. Thanks for taking my question. Maybe just a clarification, Chuck, on one of your comments in your prepared remarks, or to one of the questions. You mentioned the advanced logic step up in 2019 after the issues and that ramp-up this year. Is that your current expectation? Or is that just more of a general comment? Thanks, then a follow-up.
There's absolutely no confirmation in that respect from the customer base. We do know that this technology in advanced logic was announced in 2016, and there should come a time when it will go into HVM. I think together with us, more peers are at least commenting that there's a reasonable possibility. There's no confirmation from the customer in that respect in any way, if that's what you would like to know. Okay?
Okay. Thanks for that. Maybe just then, I'm not sure if you responded to already, the comment on some uncertainty around the timing of tools in Q3. Can you give us a little bit more detail on what exactly you're talking about there? Thanks.
It's just lead times. It's just related to lead times and the exact capabilities to align customer request date with the capabilities of our supply chain. Peter, anything you want to add to that?
Yeah, I think that's the reason why we already have given now for a few quarters a broader range. There are a lot of requirements of customers. Those requirements, due to the pressure that everyone has in increasing its capacity, might lead to, for certain periods, for certain days or certain weeks sometimes to an uneven distribution of the products that you need to have. That leads to last-minute changes, and that leads to sometimes slight delays in delivering the products. We have seen that in the second quarter. That was the reason why we said, okay, we have a range of EUR 200 million to EUR 230 million. While the analyst community, of course, came to a midpoint, which is EUR 250 million, where you see then that one or two tools are not going out at the last minute.
This is a principle, some uncertainty that we see at this moment in the value chain, also for the third quarter.
Got it. Thank you.
You're welcome, David.
Thank you. We will take our final question from Peter Olafsson from Kepler Cheuvreux. Your line is open. Please go ahead.
Yes, thank you. I had a question on spares and services sales, which came in pretty strong in the second quarter. Is there any link with equipment sales in the quarter, or is this spares and services sales mainly driven by your installed base business? What kind of growth rates should we model for spares and services for the coming years?
Let me try to answer that, Peter. I think there is no direct link between the spares and services business within the quarter in delivering of tools. Sometimes you deliver towards customer a few spare tools, but that is not always the case the moment that you deliver the equipment. You have to make a distinction also between the spares, which are for normal usage. The things that are used up during a certain period, and that will increase as a function of the installed base over the longer term. The other thing are spares, where you simply have to replace a part of the tool, which is on a certain moment not functioning any longer. That has a more ad hoc consequence and ad hoc sales levels that have increased.
We expect that while our installed base is growing, that our service business will grow with the equipment installations that we are doing, be it at a delayed time level.
What kind of growth rates you then think of? Is it maybe mid-high single digits or potentially double digits?
That depends a little bit. We expect that to grow with the installed base that we have. When we have a sales level of increase of 10%, then you might highly expect that one or two years later, that also will lead to additional spares and service business.
Okay. Thank you.
Yeah.
Thank you. That concludes today's question and answer session. I'd like to turn the call back to Mr. Chuck del Prado for any additional or closing remarks.
Okay. I would like to thank you all for your questions, especially on such a warm day with regard to all the people in Europe, that you still had the energy and stamina to attend this Q2 conference call. Very much appreciated. Let's really stay in touch in the coming months, and looking forward to meet some of you in one of the upcoming road shows after this midsummer period. Thank you again. Have a very nice rest of your day. Thank you. Bye-bye.
This concludes today's call. Thank you for your participation. You may now disconnect.