ASM International NV (ASM) Earnings Call Transcript & Summary

Good day, and thank you for standing by. Welcome to the ASM International Q2 2022 Earnings Call. [Operator Instructions] I would now like to hand over to the speaker, Mr. Victor Bareno. Please go ahead.

Thank you, operator, and welcome, everyone. I'm joined here today by our CEO, Benjamin Loh; and our CFO, Paulus Verhagen. ASMI issued the second quarter of 2022 results yesterday at 6:00 p.m. Central European Time. The press release is available on our website, asm.com, along with our Q2 investor presentation. As always, we remind you that this conference call may contain information relating to ASM's future business and results. For more information on risk factors related to such forward-looking statements, please refer to our company's press releases and financial reports which are available on our website. And with that, I'll turn the call over to Benjamin Loh CEO of ASMI.

Thank you, Victor, and thanks to everyone for attending our second quarter 2022 results conference call. I hope you're all safe and well. Before moving on to the second quarter results, let me first comment on the takeover of LPE that we announced a couple of days ago. We are very excited that we reached an agreement to acquire this company. LPE has a great team with a strong innovation focus that will allow us to enter the high growth silicon carbide market. LPE was founded some 50 years ago and has always focused on Epi solutions for the power electronics market. Traditionally LPE was mainly active in silicon Epi, but in recent years the company has gained a leading position in the fast growing silicon carbide Epi market. Silicon carbide is a semiconductor material that compared to silicon offers a much higher power efficiency, especially at higher voltages. The benefits of silicon carbide are particularly relevant for power electronics in electric vehicles where silicon carbide devices enable an extended range for the same battery capacity, faster charging times and lower costs. EV is the key end market application driver for the silicon carbide device market, which according to third-party research firms is expected to grow from some 1 billion last year to more than 6 billion by 2027. Epitaxy is an important step in the manufacturing of silicon carbide devices. Based on the strong increase in required wafer capacity, we expect the global demand for silicon carbide Epi tools will increase by a CAGR of more than 25% in the period from 2021 till 2025. LPE is very well positioned to benefit from the expected strong growth in demand. It offers very competitive 150 millimeter silicon carbide tools, which are used in manufacturing by several customers. It has also developed a 200 millimeter tool, which is in evaluations with multiple customers. Driven by strong demand for silicon carbide and successful customer penetrations, LPE is experiencing substantial growth, with expected revenue increasing to more than EUR 100 million in 2023. And with the largest part coming from silicon carbide Epi. The acquisition of LPE fits perfectly with our strategy and positioning. While the advanced [ logic ] part of the market is the main driver for ASM, we have also built a strong position in the niche markets of power, [ analog ] and wafer manufacturers. In Epi for power electronics, we have so far been focused on silicon epitaxy tools, LPE's advanced silicon carbide tools therefore complement our offering very well. We expect to drive further growth and drive value creation in LPE's business by leveraging our strong customer base in power electronics and further differentiate LPE's tools on the back of ASM's strengths in areas such as 200 millimeter platform expertise, process control technology and cost of ownership. We also see opportunities for synergies in leveraging ASM's scale and capabilities in manufacturing, supply chain management and our global service network. In short, with the acquisition of LPE, we expect to add a very attractive and high growth business to our portfolio. Next, I'd like to highlight the appointment of Hichem M'Saad as the third member of the management board, as approved by the shareholders meeting last May. As most of you know, Hichem has been instrumental in many of ASM's successful new product introductions in the last years and in his new role as CTO, he will play an even more important role in the further expansion of our company. The agenda for the rest of today's call is as follows. Paul will discuss the financial details of the LPE acquisition and review our second quarter results. I will then continue with a discussion of the market trends and our outlook followed by Q&A. Paul, over to you.

Thank you, Benjamin, and also thank you all for joining today's call. As Ben said, for some financial details on the acquisition of LPE, we will pay for the acquisition with a combination of EUR 283 million in cash and approximately 631,000 ASM shares. The cash will be financed out of our cash balance and the share part representing 1.3% of ASM's further share count, consists of 580,000 treasury shares and approximately 51,000 newly issued shares. In addition, we will pay earn outs of up to EUR 100 million depending on certain performance metrics over a 2 year period. The earn outs will be exclusively paid in cash. LPE's revenue for 2023 is expected to be more than EUR 100 million as already mentioned by Benjamin and this represents substantial growth compared to recent years. LPE is profitable and expected to contribute to net earnings from day one. We're not providing details on LPE's margins, but their profitability matches nicely with our midterm guidance for ASM's profitability. Let's now continue with a review of our Q2 results. In the second quarter, revenue increased to EUR 560 million, up 8% from the first quarter. Compared to the second quarter of last year, revenue increased 30% at constant currencies and 36% on a reported basis. Revenue in the quarter was at the higher end of our guidance of EUR 540 million to EUR 570 million. Spares and service revenue grew 14% year-on-year at constant currencies and accounted for 16% of total sales. Our equipment revenue in the second quarter increased 33% driven by strong ALD sales, which continued to account for more than half of our equipment revenue. In terms of customer segments, revenue in the second quarter was led by foundry, which surpassed the record level in Q3 last year. Logic was the second largest segment with revenues strongly up year-on-year and compared to Q1 and the combination of logic and foundry continue to account for the largest part of our sales driven by advanced node spending. Memory sales decreased both compared to Q1 and year-on-year, mainly explained by the phasing of investments. 3D-NAND accounted for the largest part of memory sales in the quarter. Gross margin amounted to 47.5% in Q2, slightly down from 47.8% in the first quarter and 48.1% in the second quarter of last year. The variations in the gross margins are mainly explained by application mix. SG&A expenses increased by 47% year-on-year on a reported basis and 41% at constant currencies. The increase in SG&A in Q2 was mainly driven by employee expenses, including related expenses such as travel. Our recruitment efforts were again successful with the total ASM headcount at the end of June, up 37% year-on-year. We continue to invest to support the growth of our company. The sequential increase in SG&A compared to Q1 was somewhat higher, including the annual compensation adjustments that took place in the beginning of the quarter in April. For the remainder of the year, we expect gradual increase in SG&A from quarter-to-quarter. Net R&D expense increased 51% year-on-year and on a reported basis 46% at constant currencies. Below the operating line results included the currency translation gain of EUR 26 million, mainly reflecting the appreciation of the U.S. dollar. This compares to a translation gain of EUR 9 million in Q1 and a loss of EUR 2 million in the second quarter last year. Let me switch now for a second to ASMPT. Our income from investments in associates, which reflects our 25% share of the net earnings from ASMPT, increased to EUR 23 million in the second quarter, up from EUR 20 million in the first quarter and also up from EUR 16 million in the second quarter of last year. ASMPT reported Q2 sales of EUR 664 million, down 2% compared to Q1 and about unchanged compared to Q2 last year. Bookings amounted to USD 593 million in the quarter, down 34% sequentially, and down 37% year-on-year. Now I'll turn back to ASM's consolidated results. I mean, our new orders in the second quarter hit a record high level of EUR 943 million, up 83% from Q2 last year as reported and up 73% at constant currencies. The order intake was led by record high ALD bookings and we also booked record high orders in Epi. The order intake was driven by broad based demand with the new quarterly record for the logic/foundry segments, with strengthened foundry and also record high bookings in 3D-NAND and in DRAM. Foundry presented, again, the largest segment in the second quarter followed by memory and then logic. Now going to the balance sheet. In Q2 we generated a healthy free cash flow of EUR 121 million including ASMPT dividends of EUR 32 million, which was offset by a cash outflow of EUR 122 million related to the payment of dividends. As a result, we ended the quarter with a largely unchanged cash position of EUR 552 million. Free cash flow for the quarter was supported by continued strong profitability. Cash outflow for working capital increased EUR 43 million mainly due to high inventories. This is in part explained by a policy to maintain higher levels of buffer inventories for critical materials due to supply chain constraints. CapEx in Q2 was EUR 16 million and EUR 43 million in the first half. We still expected for the full year of 2022 to be at the high end of the range of the midterm target of EUR 60 million to EUR 100 million annually. Note that we have not yet started the share buyback that we announced earlier this year in view of cash outflow for the acquisition of Reno in Q1 and the dividend payments in Q2 as well as the announcement of the LPE acquisition. We still intend to execute the program within the previously announced 2022-'23 timeframe. With that, I hand back the call to Benjamin.

Thank you, Paul. Let's now look in more detail at the trends in our markets. In the course of the second quarter, parts of the semiconductor market, especially PCs and smartphones started to weaken, in part impacted by rising inflation and interest rates, geopolitical tensions, and concerns about the economic outlook. Growth in other segments, in particular data centers, automotive and industrial remain solid. On balance, growth in the semiconductor market is still expected to be positive this year, but not as strong as expected earlier in the year. Looking at the wafer fab equipment market or WFE, demand remains strong and broad-based in the second quarter. For the full year, WFE is on track to increase by a mid to high teens percentage as we communicated earlier. Although it is now likely that growth will be more towards the low end of this range due to industry wide supply chain constraints. Our company booked record high orders in the second quarter, and as Paul just mentioned, for a big part driven by the logic/foundry segment. The leading foundry and logic customers are currently preparing the ramp of their new technology nodes. This most advanced nodes that we're moving to high volume manufacturing in the second half of this year and into 2023 accounted for the largest part of order intake in logic/foundry in the second quarter. We have secured many new ALD layers and applications in this transition consistent with our previous projections for strong double digit increase. Our customers are investing in leading edge capacity to accommodate new products to be introduced in 2023 in structural growth segments such as high performance computing. We are currently not seeing a change in this trend. Logic/foundry demand in the advanced node remains strong. Looking a bit further out, we remain strongly engaged with leading customers in the development of gate-all-around, ALD and Epi applications. This will be the next major inflection in logic/foundry, which as discussed on previous occasions, will drive a strong increase in our addressable ALD and Epi markets. In DRAM and 3D-NAND, we also booked record high orders in the second quarter. While general visibility in memory spending always tends to be a bit shorter, it is important to note that our strong order momentum is, to a large extent, supported by the contribution of new ALD applications. In DRAM, we continue to benefit from the adoption of high-k metal gate, which is expected to drive further healthy demand. In 3D-NAND, we had a very strong increase in orders for substantial part driven by ALD gap-fill reflecting our recent tool selections with leading customers. In this context, I would like to highlight our announcement of TENZATM ALD for gap-fill and liners at Semi Congress last week. It is a highly innovative process solution that we offer on our QCM platform and provides the best film quality, highest productivity and lowest cost of ownership in this class. And it's especially effective, complex, ultra high aspect ratio structures, such as in the upcoming generation of 3D-NAND devices. At Semi Congress we also announced the launch of our new 300 millimeter vertical furnace, SONORA. It is a completely new platform that offers substantial benefits compared to its predecessor, our A412 vertical furnace, such as 30% higher productivity, low usage of energy and chemicals per wafer and optimal film uniformity. Initially we were targeting with SONORA, our existing segments in the vertical furnace market, such as power analog and some parts of logic. But we believe SONORA also offers compelling advantages in advanced logic/foundry and memory applications. Lastly, looking at the power analog and wafer market segments, demand also continues to be strong. In this market, we are also benefiting from recent successful product introductions, such as our 200 millimeter vertical furnace A400 dual and the Intrepid ESA for 300 millimeter power and wafer maker applications. These segments are the smaller part of our revenue, but on the back of a very robust increase, we expect a meaningful contribution to our total top line growth this year. Let's now have a look at the supply chain situation and the guidance that we provided for the rest of the year. In the second quarter, supply chain conditions continue to be challenging. Our global ops team went again the extra mile to meet customer requirements as best as possible, resulting in our second quarter sales reaching the higher end of our guidance. The improvement in supply chain -- the improvement in supply constraints has been slower than expected. And as a consequence, we expect, again, an impact in the third quarter. While certain areas of the supply chain are gradually improving, for other parts, lead times continue to be extended with limited visibility for improvement in the next few months. We remain focused on mitigating the impact through early ordering and qualification of additional suppliers. We expect an increase in our third quarter sales and our guidance as stated in our press release yesterday is for a range of EUR 570 million to EUR 600 million. Supported by our record high order backlog of EUR 1.4 billion at the end of the second quarter, we still project an improvement in our sales in the second half compared to the first half. We continue to assume some improvement in the supply situation towards the end of the year. And based on that, we expect revenue in the fourth quarter to be higher than in the third quarter. Finally as supply constraints will limit the extent of the increase in our shipments in the second half, we expect our order backlog to remain at an elevated level as we exit 2022. With that, we have finished our introduction. Let us now move on to the Q&A.

We'd like to ask you to please limit your questions to not more than 2 at a time so that everyone will have a chance to ask a question. Operator, we are ready for the first question. Can you please start the Q&A?

[Operator Instructions] The first question comes from the line of Adithya Metuku from Credit Suisse.

Congratulations on the stellar order book. I had a couple of questions. Firstly just on the orders run rate, that was a big surprise, a head scratching surprise. Was there some kind of one-off in terms of the delivery of these orders? Has something changed with the delivery timing? Are you getting orders well in advance of 3 to 6 months delivery times? So any color you can provide there would be helpful. And looking forward, how do you expect the run rate to be for orders going forward? If this is the run rate that we should expect, then your midterm guidance will be -- will start to look very conservative. So any color you can provide around the order book? And any one-offs in the order book would be very helpful. And then secondly, just a quick question for Paul. I think you talked about SG&A trends in the second half of the year. I just wondered if you could also give us some color on what to expect for R&D in the second half of the year. That will be super helpful.

Adi, thanks a lot. So maybe I'll answer the first question regarding the orders, the run rate and maybe the outlook. So maybe just for clarification, if you look at the way that we report our orders, we only report orders that we have actual POs and that have to be delivered within the next 12 months. So actually the orders that we get, the significant high order intake of Q2, all of them, if we can deliver, has to be delivered within the next 12 months. Is it a one-off? Actually no. We -- as we have shared in the commentary, we see very strong demand coming in from especially logic and foundry. And this is for the ramp of the next nodes, which is currently ongoing, which is going to happen starting from the second half of this year well into the next year. So all the orders are actually going to fabs to be placed into high-volume manufacturing. Do we see, again, a lot of advanced orders? Again no. All the orders that we have reported, the EUR 943 million have to be delivered within the next 12 months. I mean, the delivery dates requested by the customers are all within the next 12 months. And of course, depending on the supply chain situation, we will endeavor to do as much as we can, although it's a little bit tight now. Looking forward, we do not provide guidance on bookings or orders anymore since the beginning of this year. But probably suffice to say that we still see that order or demand is going to continue to be strong in this quarter.

Just a quick clarification on that. So usually, you used to say the timeline is 3 to 6 months for the orders in your order book. It looks like now you're saying it's 12 months and so that time period seems to have gone up. Am I correct? Or has it always been 12 months?

I think before all of this COVID, this ramp came up, it's correct. Our order lead time or let's say, our equipment lead time was probably in the range of 3 to 6 months. But we have also shared that this has now significantly extended. The other thing is we do not just work with our customers based on actual purchase orders. In fact, for our largest customers, we work with them based on forecast. So the purchase order is just the last part of the process where we actually book and report it. So a lot of the times, we really, for example, get forecast from our customers, and we know what to expect, and we have to actually do some work in advance to ensure that we meet their forecast.

And just on the OpEx, Paul, please.

Yes. Maybe first on SG&A. As you've seen, there was a decent increase in Q2 compared to Q1, and that was to a reasonable extent related to the annual compensation cycle that we have. There is sometimes targeted, let's say, merit increases for special groups, et cetera, based on benchmarks and based on retention and based on other regions. But April 1 is our annual run basically for all of our employees. So that was a large part of the bump you saw in Q2. For Q3 and Q4, we expect a gradual increase. I still expect it to be somewhat higher, but you should not expect a similar increase as you've seen in Q2 compared to Q1. R&D, I would expect to trend up faster than SG&A in Q3 and Q4. We are still investing in R&D to support the growth of the company. It's always a matter of whether or not we can get the people, et cetera, given the war for talent that is ongoing. But I would expect the increase in R&D from quarter-to-quarter, so going into Q3 and then going into Q4 to be somewhat higher than the increase that you most likely will see in SG&A. I hope that gives you enough color for you.

We have the next question coming from the line of Sandeep Deshpande.

My first question is regarding the order book. Congratulations on the very strong order book. How much of these orders are to do with some of these new products that you are or new layers that you have broken into such as gate-all-around, in the logic space or gap-fill, in NAND or high-k in DRAM? So I mean, how much of the order book would you say is from these next-generation technologies? And how much is what was always, I mean, the ongoing ASM business as such really? And then I have one follow-up.

Sure, Sandeep, thanks for calling in. So if we look at the very high orders in the second quarter, a lot of it was driven, first of all, by logic and foundry. And maybe to give you some color, of course, the current ramp in logic and foundry is at the most advanced nodes. So whether it's 3 nanometer, regardless of whether it's FinFET or gate-all-around, that is where it is driving most of the orders today. Now of course, we do see orders also from some of the, let's say, one N-1 node and so on, but it's primarily driven by the latest node that is being ramped up for high-volume manufacturing. On the memory side, I think we have also shared in our commentary that, first of all, we had record high orders in 3D-NAND. That is largely driven by the new customer selection wins that we have for gap-fill, for the ALD gap-fill applications because that's actually driving the need for more and more gap-fill applications. And of course, in DRAM, you're absolutely correct, Sandeep. A large part of it is driven by increasing adoption of high-k metal gate.

Just one clarification there, Benjamin, which is in terms of the -- what you said about next-generation node, next-generation node is 3 nanometers, which is not the gate-all-around node. So are you saying at this point despite this very strong order intake, GAA is not very much significantly there in your order book?

I think, Sandeep, we all know that the main bulk of 3 nanometer is still on FinFET. There's a little bit of gate-all-around that has just started. But I think we all know that the volume is very, very small. So larger part of the drive is still coming from the traditional FinFET 3 nanometer.

One quick financial question. I mean you've had these great orders and great sales growth. I mean, when you look at your sales growth in the first half, I mean, 34% year-on-year sales growth. But I mean, normally, the tech company sales -- such sales growth will mean much more leverage. And unfortunately, when you look at your operating margin, I mean, that's up only 20% year-on-year. So will this -- I mean, clearly, have there been any reasons which you have given in the past why this has happened? Will this change now going into future? Or are you still going to remain at lower leverage as such?

Thanks for the question, Sandeep. Indeed, I actually 100% agree with you that with the growth that you see, we would maybe expect in normal circumstances a larger expansion of margin. We are actually investing still in growth, which is also supported, of course, by the order book you've seen. What I can say is, of course, that our midterm guidance stays. So still definitely our plan with the growth that we anticipate for the years to come to get to high single digits for SG&A, which was around 12% in Q2 and to get to high single to low double for R&D. That didn't change. So the benefit, in particular, in SG&A are still to come. But yes, as we say in the Netherlands, the [Foreign Language]. So the cost is ahead of the benefit. I don't know how you say it in English, that's what you see reflected in our results, but it will come.

We have the next question coming from the line of Marc Hesselink from ING.

First question is on the supply chain. In the supply chain, do you see particular bottlenecks? Or is it more broad-based? And if it's broad-based, how much is in your control to improve it over the coming quarters? And second question is, after the acquisition, WHAT'S now your appetite for additional M&A? Is it more you're looking to digest it first? Or is it something that now that you make these steps, you can do more of it in the future if you find the right match?

Thank you, Marc. So first of all, on the supply chain, I think we are still seeing the same bottlenecks. I would say the particular one is in chips. And because of the shortage of certain chips, it actually impacts the, what we call, our sub-tier suppliers. So some of our module suppliers, for example, aren't able to supply because they couldn't get the chips. Now in terms of control, I would say we have very little control, although we are doing everything that we can. Some of the manufacturers of the chips, we really -- we know who they are, and they are our customers, and we have also gone, maybe for lack of a better description, begging for chips for allocation. So we do whatever we can. And we just want to make sure that we are able to get as much as possible so that we can meet customer requirements. On M&A, we have always had the same strategy that we will look at M&A, but it will be restricted or limited to what is our core strategy of deposition. And we will continue to do that. LPE of course, falls very nicely into the same area in 2 ways. One, it is in deposition. 2, it actually allows us to add value with a couple of things; our epitaxy process, control technologies. It also allows us to potentially leverage on the wider sales and service network that we have [Audio Gap] supply chain capabilities, manufacturing capabilities. So in terms of a good add to our portfolio, I would say, that's probably an option that we came across until LPE, which we have signed the agreement. So we are actually very excited with this. And hopefully we get to close the deal as soon as possible.

Maybe shortly on the supply chain, you're assuming that it will get a bit better over the 2 quarters, the coming quarters. How strong is your visibility on that? Because I think this is what we've just seen in the second quarter is actually it got quite a bit worse. So how do you feel about visibility on that?

And that is where we are trying to be a little bit cautious here. Some areas, visibility is still very limited. And without the clear visibility, it's difficult for us to try to make further quantification of how much revenue we'll be able to deliver. But you are correct, Marc, some areas the visibility is still bad. We do think that some areas have improved. And hopefully, those areas which still have limited visibility, over the next couple of months, it will get better and help us to actually convert more of our backlog into revenue in the fourth quarter.

We have the next question coming from Robert Sanders from Deutsche Bank.

Congrats on the quarter. I just wanted to ask a few questions about LPE. So when we speak to LPE's customers, they say that the single chamber -- single-wafer approach is preferable for uniformity. But there are players doing planetary reactors that have big customers as well. So are you basically placing a bet on their approach on the single chamber or single wafer approach? That's the first question. The second question would be, should we assume that because you've done this deal that they are very well positioned at the top 5 players in silicon carbide for 8 inch? Obviously they have STMicro, they have X-FAB, but how are they positioned at the top 5 players today? And then the last question would just be, when you think about this market in silicon carbide, are you not a little bit concerned that there could be a China player that enters the market like happened with AMEC, which obviously really hurt the LED market back 5 or 6 years ago. I was just wondering because LPE does have a lot of customers in China, I think.

That's true. Well -- so I'll try to answer the questions one at a time. So I think the first question was single wafer versus planetary reactors. Based on very extensive studies that we have done, I think our planetary reactors was more than sufficient, actually good, very good for high-volume manufacturing up to 150 millimeter form factors. As the industry is looking to transition to 200 millimeter form factor, what we understand is that or what we have found out is that the planetary reactors or what you call the batch type of approach is losing steam because their performance is not good enough. And you'll find that even some of the competitors of LPE, they are developing single-wafer reactors at this moment. So it looks like that is going to be the trend going forward, especially with the transition to 200 millimeter form factor. So that's one. In terms of the engagement, the 200 millimeter, I would say that LPE is very well engaged. They have -- we will not go into a specific customer names, but they are in, I would say, a lot of engagements with quite a number of customers on either joint development or evaluations or doing demos on 200 millimeter form factor. So we are quite confident that these customers will also find that the performance of their single-wafer reactor will be good, and we'll hopefully adopt them for high-volume manufacturing when the transition happens. China. China is interesting. And of course, when you look at the history of LED, that's exactly what I think you are referring to, Rob, where the Chinese players basically dominated a large part of the LED, let's say, manufacturing equipment market. I think silicon carbide is a very much more different animal in the sense that it is a very difficult material. The process is so not easy. So to try to copy is very difficult. And I do believe that the -- not just LPE, but maybe some of the other foreign players or peers, they have good technology, which they have developed over many, many years. It's not something which a Chinese competitor can just copy over a short period of time. I hope that answers your questions, Rob.

We have the next question from Maarten Verbeek from the Idea.

It's Maarten Verbeek from the Idea. Firstly question concerning your SG&A. When I analyzed your SG&A cost and relate that to your midpoint of your 2025 sales target, then more or less, you are already at your high single-digit target for 2025 SG&A related to revenue. Since I don't project that SG&A will really come down, does this imply that you're now betting on the high end of your target for 2025?

Thanks for the question. The SG&A cost indeed is not likely to come down. I agree with that part of the analysis. With regard to the revenue, I stick obviously to the things that we have previously announced, which is the range of 2.8 to 3.4. You can do the math. Our target is still to reach at a high single-digit left -- high single-digit level of SG&A within that targeted midterm range that we have stated. Maybe to note, there's a few one-off costs. It's not the bulk of the increases to be clear, acquisition related costs, obviously, a few more to come in Q3 most likely. But still the bulk is driven, but I communicated earlier by the growth in headcount and employee related to expenses like IT, travel, et cetera. So yes, that's where we stand on SG&A.

Okay. But if that math doesn't work out, so 1 of the 3 has to change in future, has to be different. Then secondly, assuming that you will conclude the LPE acquisition. And you have mentioned it will generate more than 100 million in 2023 and the growth is some 25% going forward. That would imply that by 2025, it will have sales well ahead of the 150 million. When I would relate that to your estimates for the epitaxy market by 2025 of about 1.5 billion to 1.8 billion. Then LPE will have a market share of, let's say, more or less 10%. Does it also imply that your target for 2025 would be a 40% market share in this market?

I think there's probably some confusion here. So when we shared with everybody about our market share projections for epitaxy growing to between USD 1.5 billion to USD 1.8 billion by 2025. That was only referring to silicon epitaxy. So silicon carbide, because we were not a player at that point in time, last year when we did our Investor Day, it's not included. Now if you look at silicon carbide epitaxy, of course, our projection is that it's going to grow at about 25% CAGR over the next couple of years. So we do expect that the market or the size of the market is going to grow fairly nicely. I think we will probably want to do more work before we want to share projections or market share up to 2025. And in any case, we'll probably do this after we have closed the deal and have more -- much more insights, much more information. But we do expect that it is going to be a nice growth business for us. First of all, next year, getting to more than EUR 100 million and hopefully allowing us to grow that at least the same kind of growth that the equipment market is going to grow, if not even higher until 2025. So we will provide more details on that.

We have the next question coming from the line of Janardan Menon from Jefferies.

Just 2 for me. One is when you say that you will -- you have a 12 month period to ship the kind of orders you're taking in right now at more than EUR 900 million. And clearly the component shortages are not allowing you to go even to EUR 600 million at this point in time. Does that mean that you expect a significant improvement in the component supply situation, say, in the first half of next year, which would sort of take your sales much higher than the current run rate, say, to the EUR 700 million, EUR 800 million kind of range so that you can meet some of these obligations within the next 12 months or so? That would be my first question. And the second question is, I agree that your run rate may not continue at EUR 900 million, but you're suggesting that it could stay strong. That sort of run rate, as the previous question was alluding to, is almost getting you to your 2025 target, annual targets, if not going higher than that. I was just wondering, from your own capacity standpoint, does this big inflection in orders mean that you have to reevaluate your capacity? You're bringing on the second floor in Singapore next year. Would that -- is that coming on fast enough? And do you need to already think about another floor there to accommodate these kind of growth rates, assuming that you're going to address a growing market even beyond '23 into '24, '25, et cetera? I just want to know your thoughts on that as well.

Sure, Janardan. So maybe on the 12 months on how we report or how we book and report orders. Yes, the requirement is that if we were to book a PO, it should be that the delivery of that equipment has to be within the next 12 months. Now we would love to be able to ship everything. Unfortunately with the supply chain constraints, we probably will have some which will not be possible. Now having said that, it all depends on how the supply chain constrain situation develops and we are optimistically or we are cautiously optimistic that as we get into the next year, hopefully things will improve significantly. That will allow us to convert more of the backlog that we have into revenue. So whether we see a significantly higher first half in 2023, I think some of it, you can probably do the math because of the high order backlog that we are now reaching, it is possible. But at this moment, given the lack of visibility that far into the supply chain, we are not able to quantify that. Secondly, on the run rate, we, again, hope that if this run rate is going to continue like this for the next couple of years, of course, we are going to run out of space. But having said that, we have already taken a very proactive step in already expanding the second floor of our Singapore facility. With that expansion, we basically triple our manufacturing capacity compared to 2020. And with that capacity, we are actually confident that we will be able to meet our revenue, our midterm revenue target of 2.8 billion to 3.4 billion. And if we have to look at -- unless the things grow so fast, much faster than we expect, then maybe we need to look at other options. But at this moment, we are confident that based on the midterm revenue target of 2.8 to 3.4, we have enough space, and we are fast enough because we have already taken probably a good step to fit out the second floor, which will be completed by the beginning of next year.

The next question comes from the line of Didier Scemama from Bank of America.

First of all, congratulations on this extraordinary order intake and also congratulations on the acquisition of LPE. It looks like a very smart acquisition. 2 questions, if I may, a lot of the good ones have been already asked. I just wondered if you could go back to the order intake, if you could deconstruct that a little bit. I just wanted to understand if you could give us a bit more color on the ALD capital intensity at the 3 nanometer node versus the 5 nanometer node. Is that the major driver for the order intake beat? And then secondly, it sounds like you are like sort of reaching new levels of order intake. And I wondered how much of that is driven by 3D-NAND and high-k metal gate? And I guess related to that question is, I mean, should we start to think about ASMI also as a memory player? Or is it too early to say, is still foundry/logic, building a small position for the time being in memory? And I've got a quick follow-up for Paul.

Didier, thanks for the congratulatory message. On the first question of the ALD capital or ALD intensity. I think what we have always shared, of course, is that when you move from one node to a smaller one, we do see a double-digit percentage increase. And as we have also mentioned in our prepared remarks, that is one of the factors that is driving the, of course, the higher bookings that we are seeing. Now of course, there's also the level of investment, which is very high now in logic and foundry. And again, alluding to our previous comments, you see a lot of investments now into, for example, 3 nanometer, which is the node that is being ramped in preparation also for next year's consumer end products. So I think that is what is really driving it. ALD intensity, of course, is increasing, but it's also the volume of investments. On maybe, let's say, the breakdown and so on, we will not give that by -- specifically by markets. But generally we have very broad-based good order intake. And I think we also mentioned in the prepared remarks that we saw record, for example, order intake for 3D-NAND. And I think this is due to the wins that we have had recently, especially in the 3D-NAND for gap-fill applications, which is becoming more and more important as the number of layers or number of pairs goes up. DRAM, of course, with the increasing adoption of high-K metal gate. Now having said that, I think we are doing what we have set out to do in terms of our strategy, and that is to slowly but surely increase our share in memory. But having said that, the bigger part of our play is still in logic and foundry. That is still our main market. So I think we'll leave it to you to figure out or to decide whether you want to consider us as a memory player. We will definitely continue to grow our memory business, but the larger part of our business is still in logic and foundry.

And I think Janardan tried to have a go, but I'll have a go as well. If you had all the capacity that you wanted, can you say what your orders might have been, A? And then, B, what your revenues might have been in Q4?

We will not quantify this. But what we can say, of course, is if we were to assume a hypothetical situation where we did not have supply chain issues, revenues could be significantly higher. But how much? We will not quantify that because, as I said, we still have difficulty with limited visibility in some parts of the supply chain. What we can say for sure is that supply chain constraint is, in fact, limiting our ability to convert more of our backlog into revenue.

We have the next question coming from the line of Timm Schulze-Melander from Redburn.

I had 2, maybe one for Ben and one for Paul. Maybe just starting where sort of Didier left off. Did supply chain constraints limit the amount of orders that you recognized in 2Q?

No. I think if you look at, for example, maybe compared to some of our peers, our lead times are still relatively short. And in that respect, most of the orders that we get actually for fairly near-term, in fact, for delivery within the next 12 months. So it was not like we were getting orders that were for delivery 1.5 or 2 or 3 years out and we were not getting those orders. So the supply chain is really impacting near-term delivery capabilities. And that is the reason why you see that revenue-wise, we have only been able to increase Q3 or third quarter, not as much as we would have loved to. But we do think that with improving supply chain conditions, we should be able to have more revenue in the fourth quarter compared to what we have guided now for the third quarter.

And just on those orders that you just got in, are there -- are you in any way restricted to developing second sources? Are there any conditions in those orders that limit you to your existing supply chain?

I think we have been doing that and in cooperation with also customers. I think customers are actually today very much more understanding and helpful. So of course, we have to work with our suppliers, but a large part of the work that we have to do is also with our customers and getting them to help us qualify and approve alternative suppliers. I think if that was not the case, things would actually be even more difficult. But I think most of our customers today are willing to lend a helping hand and willing to offer us this help. So I don't see any bottlenecks or any constraints in that area.

And then just a quick follow-up for Paul, please, 2 sort of housekeeping questions. The first is, you obviously have visibility a quarter ahead in terms of what you ship. Could you just let us know the application mix relative to 2Q? Will that be flat or improve or erode in 3Q? And then your revenue guidance, could you share with us what U.S. dollar exchange rate that was struck at, please?

Yes. Maybe last one, one to start with. Basically every time when we provide guidance, we use the latest rate because we're not going to speculate on a lower, higher, stronger dollar euro rate or whatever. So we typically base it on the current exchange rate, just to be clear. Then coming back on the application mix. Obviously we have a detailed plan and orders by customer, by product. As you know, I'm not going to guide on mix because indirectly I would be guiding on margin. It's also difficult because there are at times still push out, pull in, so things can still change within the quarter. So it's not like that it's already a given what the margin will be because of these type of dynamics. There can be unexpected decommitments from suppliers that can have, again, a change in the mix as a consequence thereof. But what we try to do, of course, is stay within our guidance as we say every quarter. That might not help you a lot, this answer, but that's actually the best I can give you. There is a lot of dynamics going on. And if I would now say to make it the same or better, then basically, I'm saying the margin will be the same or better. And that I simply cannot do given all the dynamics that we have to deal with.

We have the next question coming from Adithya Metuku from Credit Suisse.

Just 2 questions. Just on the acquisition of LPE. Have you looked into the probability of regulatory approvals? I understand LPE, somebody tried to take it over previously, it wasn't domiciled in the E.U. and that acquisition was rejected. Have you looked into that? That's the first question. And secondly, just on the supply constraints, are they -- these chip supply constraints, are they -- do they have anything to do with any discontinued products? Is that why you're struggling to find supply in the market? Or is there some other thing that we need to be aware of? I'm just a bit surprised that these constraints are not easing given the weakness we're seeing in the consumer end market and given these are meant to be on these chips are meant to be on lagging edge nodes? So any color you can provide there would be helpful.

Sure. So first of all, on the acquisition of LPE and the regulatory approvals, of course, we have dived very much into that. We will, of course, need FDI approval from the Italian government, and we will need antitrust approval from some other countries. But we are actually confident that we will be able to get this given the relatively limited impact that this is going to have. So we are just working through the process and the process of getting the approvals have already started. So we are already engaging that. On the chips, I don't think it is necessarily discontinued products. But if you look at some of the older legacy chips, some of the manufacturers have for some time already been producing less and less of them. And because of this less and less of them, there is all of a sudden also having to meet demands from other areas, whether it's from consumer or industrial and even to some extent, maybe automotive, that has created the shortage there. I agree with you, Adi, that hopefully, we have some of the easing in some of the markets for, let's say, demand for legacy chips that hopefully, this is going to result in, for example, not just ourselves, but also some of our peers having a larger share of the older chips that we need actually for our equipment.

We have the next question coming from the line of Nigel van Putten from Kempen.

Again on the order intake because I understand visibility is still limited in some parts of the market, and that may sort of limit your willingness to talk about the current situation, especially relative to your medium-term guidance, I sort of sense that there is not a connect there yet. But let's assume then that spend on foundry is unchanged next year. Should you see orders similar more to this quarter or last quarter because it's quite a step change? So basically, is the current situation still exceptional and not to be repeated? Or is this more in line with sort of the current focus of the market into the next node?

Nigel, I think it will really depend on how much -- the main logic and foundry players are going to invest in terms of adding capacity at the leading edge. That's one. We also see a lot of, what you call, technical buys happening for development. So if we see that both of these are going to happen, continue to happen at a high level, we do expect that potentially orders from the logic and foundry sector is going to continue to be strong.

There are no further questions at this time. I will now hand back the call to Mr. Loh for final remarks.

Sorry. And thank you for calling in today. I was just trying to work through the notes. Thank you again. Thanks for calling in. Stay safe and goodbye to everybody.

Ladies and gentlemen, that does conclude today's conference. Thank you for participating. You may now disconnect your lines.