© 2024 Alpha Spread Limited. All Rights Reserved.
www.alphaspread.com is not operated by a broker, a dealer, or a registered investment adviser. Under no circumstances does any information posted on www.alphaspread.com represent a recommendation to buy or sell a security. The information on this site, and in its related newsletters, is not intended to be, nor does it constitute, investment advice or recommendations. The information on this site is in no way guaranteed for completeness, accuracy or in any other way. In no event shall Alpha Spread Limited be liable to any member, guest or third party for any damages of any kind arising out of the use of any content or other material published or available on www.alphaspread.com, or relating to the use of, or inability to use, www.alphaspread.com or any content, including, without limitation, any investment losses, lost profits, lost opportunity, special, incidental, indirect, consequential or punitive damages. Past performance is a poor indicator of future performance.
You don't have any saved screeners yet
Intrinsic Value is all-important and is the only logical way to evaluate the relative attractiveness of investments and businesses.
Good morning, good afternoon, ladies and gentlemen, and welcome to Besi's quarterly conference call and audio webcast to discuss the company's 2020 third quarter results and the joint program with Applied Materials to develop industry's first complete and proven equipment solution for die-based hybrid bonding. You can log into the audio webcast via Besi's website, www.besi.com. Joining us today are Mr. Richard Blickman, Chief Executive Officer; Ruurd Boomsma, Chief Technology Officer; and Ms. Hetwig van Kerkhof, Senior Vice President, Finance. [Operator Instructions] As a reminder, ladies and gentlemen, this conference is being recorded and cannot be reproduced, in whole or in part, without written permission from the company. I would now like to turn the call over to Mr. Richard Blickman. Please go ahead, sir.
Thank you. Thank you all for joining us today. We will begin by making a few comments in connection with the press releases we issued earlier today and then take questions. I would like to remind you that some of the comments made during this call and some of the answers in response to your questions by management may contain forward-looking statements. Such statements may involve uncertainties and risks as described in the earnings release and other reports filed with the AFM. For today's call, we'd like to review the key highlights for the third quarter and the 9 months ended September 30, 2020, and update you on the market, our strategy and outlook. We also refer you to the joint press release issued today with Applied Materials. First, some overall thoughts on the quarter and the 9-month results. Besi reported solid results for the third quarter and the first 9 months of the year. For the quarter, revenue and net income reached EUR 108.3 million and EUR 34 million, respectively, increases of 20.7% and 77.1% versus the third quarter last year. Third quarter orders of EUR 94.9 million grew by 15.5% versus the third quarter last year. Besi's operating profit of EUR 42 million was at the high end of prior guidance as we had better-than-anticipated shipments for mobile applications, maintained gross margins in excess of 60% and exceeded our operating expense reduction target. As a result, Besi recorded a net margin of 31.3% in the third quarter, the second quarter consecutive -- the second consecutive quarter in which profit margins exceeded the annual level achieved during our last 2017 cyclical peak. Results for the first 9 months of this year were also strong, with revenue of EUR 323.9 million, up by 22.8%; and net income of EUR 87.6 million, up by 84% year-over-year. Similarly, orders of EUR 314.8 million grew 26.8% versus the prior year period. Besi's business prospects have improved this year as demand for mobile applications, the U.S. and Asian customers grew significantly in light of new product introduction, expanded 5G capabilities. Growth in mobile and end-user markets has more than compensated the relatively stable demand in computing applications and continued weakness for automotive applications. The execution of strategic initiatives has also benefited our 2020 performance, a 2.1% reduction in fixed headcount levels, increased labor efficiencies at both gross and operating margin levels and proactive supply chain management has limited inventory development and improved cash flow generation. Besi's liquidity position continued to expand during the quarter with growth in cash and deposits reaching EUR 564.5 million at the end of September this year, an increase of 54% versus June 30, 2020. The increased cash levels were primarily due to Besi's issuance of EUR 150 million and 0.75% coupon convertible notes in August and strong cash flow from operations, EUR 60.9 million, generated during the quarter. Similarly, net cash and deposits grew to EUR 158.7 million at quarter end, an increase of 69.6% versus the end of the second quarter of this year. Of note, we will cancel 1.5 million shares held in treasury in Q4 this year so we have sufficient room within our current authorization to increase quarterly share repurchases from approximately EUR 3 million a quarter to EUR 10 million a quarter. Next, I'd like to speak a little bit about the current market environment. The trajectory of semiconductor equipment recovery, which began in Q4 last year was significantly altered by the onset of the global pandemic. As you can see in VLSI's most recent climate index report, the dramatic decline experienced in the second quarter was followed by a retracement back to peak pandemic levels. Principal question now is, is the slope of the index trajectory in the near term given renewed COVID-19 restrictions in selected parts of the world? Hence, October update VLSI revisited its 2020 assembly equipment forecast to grow by 4.5%. They now assume the market growth over the next 3 years, reaching $4.4 billion by 2023. We think the upturn this year was fueled primarily by renewed mobile and Chinese capacity investment. More robust growth in the upcoming years will be driven primarily by 5G network expansion and feature functionality upgrades, continued investment in cloud infrastructure and artificial intelligence application, spurred by the work-at-home economy and significant investment by the Chinese government to build its semiconductor production capacity. In addition, we see IDMs more actively engaged in the development of next-generation processes and in the last cycle. Further, we've seen more focus by memory manufacturers on high-speed flip chip production versus traditional wire bond solutions and increased engagement on the topic of hybrid bonding for the development of next-generation applications. Towards this end, Besi and Applied Materials today announced an agreement to develop the industry's first complete and proven equipment solution for die-based hybrid bonding. This process technology represents a major improvement versus conventional chip interconnection because it increases the density and shortens the length of the interconnect wiring between chiplets. As a result, it improves overall performance, power, efficiency and cost. Besi's collaboration with Applied Materials, harnesses each firm's respective expertise in front and back-end process technology for next-generation applications such as high-performance computing, artificial intelligence, 5G mobile, data storage and automotive. Parties will establish a hybrid bonding Center of Excellence at Applied's Advanced Packaging Development Center in Singapore, which is located across the street from Besi's Asia Pacific headquarters. The Applied facility is one of the industry's most advanced wafer-level packaging labs and contains 17,300 square feet of Class 10 cleanroom space with full lines of wafer-level packaging equipment. The Center of Excellence will be outfitted currently towards the end of the fourth quarter this year with customer engagement at the facility to begin in Q1 next year. Now a few words about our fourth quarter guidance. Looking ahead, we estimate that the fourth quarter of this year, revenue will be flat to down 15% due to typical seasonal influences, lower demand for mobile applications post the capacity built this year and some concerns as to the development of COVID-19 pandemic. Besi's gross margin is estimated to range between 58% and 60% based on the forecasted product mix. Operating expenses are expected to increase between flat and 5% versus the third quarter. Assuming the midpoint of guidance, we can forecast full year revenue, gross margins, operating profit of approximately EUR 424 million revenue, 60% gross margin and EUR 144 million, respectively, increases of 19%, 4.2 points and 56% versus 2019. Longer term, we are encouraged about Besi's prospects in the next investment cycle, given our strong performance during the last industry downturn and the current pandemic by strong secular growth drivers. This chip functionality, complexity and density increase and geometry shrink Besi's advanced packaging solutions are ever more important to customers. As such, we are increasing our engagement with leading mobile memory logic players to expand our addressable market. That ends my prepared remarks. I would like to open the call for some questions. Operator?
[Operator Instructions] And the first question is coming from Peter Olofsen, Kepler Cheuvreux.
Yes. My first question is on the growth of the company this year. If I look at the guidance for Q4, you will enter this year with growth in the high teens, which is clearly better than the latest VLSI forecast for the industry. I understand that most of the growth as you've seen today is in mobile applications. So the growth in mobile, is that a combination of -- yes, the transition to 5G and related investments as well as market share gains? And in case you see market share gains in mobile, where do you see these? Is it for a specific product? Is it for specific applications, specific type of chips? A bit more color on that would be helpful. And then I have 2 questions on the announcement you made together with Applied Materials. First, will this cooperation affect the R&D spending of Besi? Also, do we assume that you were going to continue to gradually increase over time? And could you shed some light on how we should look at volumes for hybrid bonders? How that will develop? Is it fair to assume that in the next 1 or 2 years, it's mainly a couple of development tools and that it might take still a few years before you will see larger volumes linked to high volume manufacturing? Or how do you see those volumes developing in the coming years?
Excellent. Thanks, Peter. Your first question, the revenue growth this year, up in the high teens. Yes, certainly, mobile, but don't forget also the server, high-end server processor. We had a significant growth in both segments, which offset the, let's say, standstill in automotive, apart from some development. On the mobile front, many new features, features requiring new equipment with different process solutions. On the 5G front, we did gain market share also in both applications. So the iOS world and also the Android world and multiple Android high-end smartphone manufacturers. Further details, we don't disclose. But on average, you can say, we have continued to increase our share of wallet with existing customers and added a broader market position. On the AMAT front, hybrids, and that's also in our strategy, which we have updated every year, the progress and also what it would mean to R&D spending. Like, for instance, our last analyst update presentations from June of this year. Gradually, R&D spending will increase simply because of the increased complexity. With the support of AMAT and the acceleration due to our corporation, as we expected, it could very well be that it would help us in bringing our products faster to market. In your assessment of market growth. If you remember, flip chip, which was introduced in the mid-'90s, later part of the '90s, now 25 years ago, and forecasted to take the majority share of interconnect of high-end devices in the next 10 years, we're now 25 years ahead, and it's still less than 30% of the overall interconnect. If that repeats itself, hybrid is certainly focused on the highest end and the most complex devices. And in -- yes, the next year, certainly, a lot of development. But there are already major customers who forecast production facilities to be installed in the course of next year and certainly, 2022. So the visibility for mainstream applications is certainly there. So we should hurry up and do our homework right. Maybe, Ruurd, you have something to add because this is a very important question on the corporation with AMAT.
Yes. I think Richard already explained it quite extensively. Of course, it's a big step for us because we see in this hybrid bonding, you come in a new class of applications with very small dimensions where particle, where placement accuracy, all bring us into a new field, so to speak. And to do these bonding successfully, you need more than only the bond, you need also the processes that come in front of that. And there we have with AMAT and an absolute perfect partner to develop these steps further. And yes, we are very, very excited about this possibility. And the big advantage is also their fab for this work is really opposite of our building in Singapore. So that's a very natural fit also. And what we have also seen over the past, I would say, 6 months is a serious uptake in interest for this hybrid bonding as the next level of technology. And we predicted already some numbers during our analyst conference in June, and we think we are -- yes, so it can become a serious market. But as Richard said, also in the flip chip, it took a long time to get to high volumes.
Maybe as a brief follow-up, are there already customers that have some of your hybrid building tools at that facilities for development work? Or...
Yes. Remember, we mentioned that -- I think it's already 2 years. The major customer in this universe is developing on our equipment. These processes in different applications with accuracies. Now, Ruurd, we are at 100 to 200 nanometers?
Yes. We're getting in the 120-nanometer range, and you have to think about this is back-end world. So we've made big steps forward and that at speeds where we can reach about 2,000 dies per hour, which is really starting to get in a production range.
So we have been developing in this arena, I would guess already for 4 years in. And then 2 years ago, installed the first equipment at a major customer. Another system was installed last quarter, but all still in the development phase?
Yes.
So a lot is happening. And as stated in the press release, the intention is that it accelerates. The industry, let's say, acceptance and use of this technology simply by offering production capabilities in the not-too-distant future.
And the next question is coming from Mr. Wim Gille, ABN AMRO.
Yes. Can you hear me?
Yes.
Very good. Wim Gille from ABN AMRO. I saw in the press release that you mentioned specifically that you also see kind of initial orders from global memory producers for high-volume flip chip systems versus traditional wire bonding solutions. Can you give us a reminder of, let's say, what your position in the memory market is? And what your served addressable market is today? And how you expect this to grow into the coming cycle as apparently they are also moving to flip chip? And then the second question is in relation to the announcement with AMAT. Can you give us a bit more idea on what the split is that you have in terms of your, let's say, deal with AMAT in terms of revenues and costs, how this is organized? Is it organized in a joint venture? Or is it, like that you said, sell 2 separate machines? Or is it one machine? Just to give me a bit more clarity on kind of how exactly the cooperation with AMAT is formalized? And then maybe as a last question, why specifically AMAT and not any of the other players?
Excellent. First question, flip chip for Besi was never more than 15% of revenue at most. The reason is simply formulated that the complexity required for interconnect was such that we would not really differentiate ourselves compared to our competitors. So we focused broad brush mostly on logic, high-end logic and also interconnect features, and that split has been for many, many years. With the intended move from wire bond to flip chip, that has created, and this -- we also coming in the past 5 years, an opportunity to apply our leading position in flip chip applications for many, many years in the memory applications. But at the same time because memory is high volume, we were successful in developing a next-generation flip chip machine, which can do more or less 4x the output, more than, and let's say, conventional, if you can call that flip chip for logic system and produce. And that has been tested thoroughly against the competitors in that field. And as we announced in the press release, we successfully received orders for several high-volume systems for this application. So that's an entry for us for the first time in many years in the memory market. That can be a very interesting, sizable next business line in our flip chip market position. On your second and third question, on AMAT, I'll ask Ruurd to answer the second question. I can share some light on the third, why AMAT? AMAT has the most complete suite or portfolio of processes, which are needed in achieving a wafer-level interconnect process. So in customer, in time-to-market requirements in developing of new processes, AMAT is by far the unique partner to develop the leading edge in this new world. And as Ruurd said in the earlier comments, Besi is extremely proud that AMAT has selected Besi to develop this new world in this joint operation. So it's the start. So it's very nice always at the beginning but we have to deliver. The industry has high expectations. So a very positive challenge, and we look very much forward to that. Ruurd, maybe you can shed some more light on how the arrangement is organized?
Yes, yes. Let me give a few comments on that is -- and the focus is, of course, on establishing the first the Center of Excellence with Applied and Besi to co-optimize this integrated hybrid bonding flow and the equipment solutions around that. And we will work jointly with the customers on the technology development, and we also will be able in the center to have our own customized test vehicles. That's always very difficult for Besi to do that alone, you need more for that. And that will allow us to provide a more integrated solution and work on different hybrid bonding integration schemes. And then, yes, the commercialization of this is basically the traditional equipment sales model. And if you then look at what the different roles will be is that Applied is, of course, and it was a little bit your third question, why AMAT, they have, of course, all the base processes like: you need etching capabilities; you need deposition of oxides; you need, yes, wafer plating-type of technologies. You have to understand these technologies to come to a well-prepared surface for doing the actual bonding. So that's the part that Applied will do. So the semiconductor process expertise, the equipment for the etching, the planarization, a singulation, dielectric deposition, metal deposition, the wet [ films ], the plasma activation, all the things you need to come to that level. And then, of course, they have also an excellent suite of metrology tools. In this lab to do the measurements you have to do with that. And we will provide the die placement process expertise, the equipment around that, how you do that in a very particle-free environment and also the deeper chip assembly process expertise. So that's a little bit the split of the work. And yes, we are currently rolling in equipment there. And goal is that early next quarter, we start also with deeper development together with customers. Does that cover it a little bit?
Yes. To maybe -- to kind of to summarize it. Basically, you guys have your expertise, and they have their expertise. And the Center of Excellence that you're setting up is to have a joint, let's say, go-to-market strategy, whereby, you optimize the design of both, let's say, part of the system to work together, but you will still kind of be selling your own part of the system?
Yes. That is also, of course, developing over time. But you have to imagine at this moment for this type of development, we see this even with the biggest customers, they have to send a material tool. That's the location, then to that location, that location that works, first of all, very inefficiently, but it also creates more, I would say, process issues. And by bringing this on the one roof where you can do all the steps in one line and develop the exact setting, all this customer material is always slightly different. This customer has that, that customer has that. And that's the unique possibility we have in that location. And yes, that's the goal, of course, to optimize that we can say, hey, in this flow, the whole process will work excellently.
Yes. Understood. And the 2 machines that you already have installed in -- that one was installed 2 years ago, and then an R&D setting, the other one, I think, last quarter, you said. Were you already working together with AMAT on those? Or did you have a different setup there?
No. This was our own work, and we started with hybrid bonding, I would say, more than 4 years ago. We already identified this as a key possibility. And in about 2 years ago, I started discussions with AMAT on, hey, we have this knowledge, you have that knowledge. This grew in a very positive way, and we start to see now the advantages of that. But the first ones were done, I would say, on our own. And more is to come. We're going to ship more tools in the next couple of months. And there, we will start working together more on it.
Very, very clear. And then I have a last follow-up question, and it's basically regarding Apple. With the ongoing miniaturization at Apple, you basically keep expanding your share of wallet with each iteration of the lineup with Apple. Is there more to go for? Or are we already getting closer to kind of a natural ceiling for the products and the position that you have there?
Well, we've said on several occasions that the ultimate objective of the smartphone world, but basically of every computer is that you end up with 1 device. So hybrid is a key enabler for that. So if you simply look at the road maps of all these devices, and it's mentioned also in the press release, why this all makes sense is the more you can design onto one single device, ultimately, the faster the speed, the lower the energy and the lower the cost. So we are, by far, not at the end. It's still -- all these units, if you would open them up, it's crammed with individual components of all different sorts and sizes and complexities. And gradually, generation after generation, more integration, and that means more complexity and more opportunities for us.
And the next question is coming from Mr. Marc Hesselink, ING.
Three questions. First question, maybe to start, also on the AMAT partnership. I was also wondering, how is the distribution of their products are going to work? Are you are going to combine the sales channels? In other words, are you going to benefit from the large scale of AMAT?
Well, first of all, in this hybrid world, you can count the customers and institutes developing that basically on 2 hands, [ AB 3 ]. And we all know where they are. There are many publications. There are constant exchanges in conferences. So finding customers is not that difficult. What is key is to exchange with those customers, ongoing process development data and also to organize on an ongoing basis, new programs of specific applications. And that dialogue, and we explained that already, covers many processes in such a new device or a new design. And that interface needs experts. So you need experts on what can you place and how can you place that? And at the same time for etching and all the other processes, as mentioned, in a similar way. So there are customer road maps and the leading ones in this industry. And there's an ongoing dialogue on a continuous development exchange. And if you can imagine what in the future we can deliver in Singapore is practically complete process solutions. And that will work extremely well for customers having an additional capability outside their own facilities to increase, and we call that an accelerate, the adoption of this technology for many applications. So that's done jointly. That's already today. There's very close contact, and customers also organize that contact. But that's not much different from today's world. All the interfaces in today's products with different suppliers, mostly are driven by customers because many specifications, material specifications are, let's say, for many of the processes. So it's not a new world we have to develop.
Okay. And maybe then as a follow-up on that one. Can you share how this came together to get to this partnership? Is it something that was initiated by your customers or by one or two companies?
Ruurd already mentioned that. It was from both angles. First of all, and we have shared for many, many years, the next step for Besi is to move into the wafer-level world. So on an ongoing basis, we map the different processes and also with our existing products. And there are 2, it's die placement and molding. Our wafer molding system is also installed at more customers than the die placement as such. So there's interaction with those leaders in many years. And what we call the missing links are also very clear and ever more clear. So the question is how can we improve our position in that emerging market? And that's partly by our own development of systems, organic growth and also by partnerships. And to say that, again, the simple fact that discussing at some point, the progress with people from Applied Materials has brought us closer together. And out of those exchanges, ideas have been formed and now that -- in an official partnership.
Yes. Yes. Maybe some comments from my side. It has been a very natural process. We -- of course, you publish on scientific conferences where you're working on. And then, yes, you start talking together, and then you see, hey, you have that and that, we have that and that. And yes, at a certain moment, you start, hey, if we combine this, this gives us a much stronger position. And we, of course, check that also with a number of customers, how they saw that and they also agree that, that would be a big step forward. If you just visualize this hybrid bonding, currently, the most advanced bonding is in the TCB area, and you have pitches of about 50-micrometer between the contacts. And that means if you have, in such a case, you have about 400 contacts per square millimeter. If you go to hybrid bonding, you may go to a pitch of 10-micrometer, and then you have about 10,000 contacts. So you come in -- it's a completely new class. And the pitch development from 100-micrometer to 50-micrometer that went from about 100 contacts to about 400 contacts per square millimeter, and now you'll make an enormous jump. And by also eliminating the space between the dies, the speed of the data transfer increase, the term of performance goes up. So there's many advantages to this process. But it works on an almost atomic flat surface. And that means that, yes, you have to do this surface preparation very, very carefully, and that is one of the areas where Applied excels. I hope that explains -- gives a little bit more light on this thing.
Yes, clear. Is it in practice and also that it's more or less exclusive and because you're going to be so close to each other that it will be difficult to have -- to split it? Or...
We don't disclose more details than the partnership as we have as we have announced today. But time will tell how things develop in the future.
Okay. And then the final question was, in previous calls, you said that the lead times of visibility is still relatively short given the COVID-19 [indiscernible] rate until really the last moment. Did that change? And maybe also there, in that respect, in the automotive space, I mean, early sign that automotive is slightly recovering. What's your view there?
First of all, COVID has not changed, sadly. It is becoming more and more clear that life is not COVID-free at all. Maybe even more complex because of that -- of the second round. So short visibility will remain, that's also why we are careful. Although order intake is still very decent, and that's why the outlook is with a somewhat wider range because you don't know what happens because of the uncertainty. But on the second question, please help me again. What was it exactly?
Yes. In relation to that, in the automotive space, I'm seeing a clearly -- what...
Yes. Yes, automotive. That's very interesting. You can easily find that the European -- the major European auto suppliers, their demand has dropped by more than 25%. So although signs are clearly -- that bottom has been reached and there are encouraging signs of improvement, but before additional capacity is needed, in general, still more recovery has to happen. So that's why we are still careful. As I said, there are new devices. And of course, tooling, it needs to be set up for new production lines. But overall, the volume has not reached the previous peak yet. Maybe, Ruurd, you can also comment a bit on that.
No. There are, of course -- the big issue is that basically the car production dropped substantially because nobody bought new cars. And that you see, that means that production capacity is very sufficient for the moment. And that will be probably somewhere Q2, Q3, maybe next year, things will get better. But we are involved in a number of new developments in the industry because with the ongoing electrification that you see, you see a lot of new power devices getting in the automotive space. And there, we're heavily involved with the top players. So that, I would say, opens up possibilities in the next round. And the other part, we see, of course, that they're slowly, but certainly progressing the automatic driving. You call these levels, L0, L1, L2 and so on, is also progressing. And there, we see a lot of development in more sensors, specific also in the, say, in the camera space. We also see that a product that is very well-known in the normal PCB space, the QFN, is getting more and more certified for automotive applications. And there also, we're developing some new steps to reach this very high reliability level that you need. So we are very much involved in new products in the automotive industry. But the volume market at this moment is, of course, low. And that -- but that will come back.
And the next question is coming from Mr. Nigel Van Putten from Kempen & Co.
My understanding is that with regard to the work you're doing with Applied Materials tends to be sort of the 3D fabric that we read about. That seems to be -- so my question is the back end typically is much more bespoke work because a lot of -- I mean, every chip and every application has its own package or the front end is a lot more standardized. So how should we look at this? Will -- are you developing this with the foundry partner? Or are you developing individual processes with established customers as well?
It's always an end customer who is developing the road map. So it's the end application, which determines the requirements for the device, design and that then requires the manufacturing process and ultimately, back end required for that. And whether that's done by an IDM or whether that's done through a foundry, but you're absolutely spot on that the amount of variety and different designs for that matter is only increasing. So the complexity and the requirements for systems design, flexibility, so accommodating different families is a major positive challenge in the years to come.
Yes. Maybe, Nigel, a little bit more comments from my side. If you follow the real big IDMs at this moment, they put a lot of additional effort in developing packaging technologies because they have started to see that this is an essential part of their forward-looking road map, and the investments there are really increasing. And the OSATs, of course, follow that also. But the IDMs are clearly here, yes, sitting behind the reader, so to speak. And yes, that's what -- our objective or our task is, of course, what Richard is saying, the variety of packages only increases. And we have been in the past, for example, with our EVO tool, very successful in developing a tool that can handle a very wide variety of packages so that you can, I would say, standardize a tool and only have minor modifications and do a lot of products with it. And in this more advanced area, we're also looking for that type of solution where you can cover with, say, a one machine type, quite a wide variety of packages there also because you cannot build a separate machine for every package type.
And in addition, which we have not highlighted yet, all these processes in the hybrid world are in cleanroom environment. So that requires a different level of machine design. It will also significantly increase the ASPs and also the -- yes, for instance, automatic tool change, is one of the interesting next steps on our first production tool delivered to the industry. So yes, just to give another example, that characteristic will not change. It will only increase.
Yes. Clear. Then a follow-up, something completely different. You've issued a convertible in the quarter with one of the more near-term maturities trading well into the money. What is your strategy there with regard to now 3 convertibles outstanding? Will you try to take out one? Or will you leave them all there? Just curious.
Well, so far, the thoughts are if they knock on our door, and 2 have knocked on our door in April time frame, so the principal amount reduced by EUR 7 million from EUR 125 million to EUR 118 million, probably more will knock on our door. So a natural process as opposed to why accelerating debt.
And the next question is coming from Mr. Robert Sanders, Deutsche Bank.
On the AMAT deal, I just noticed that a company called Xperi has a licensing model for hybrid bonding in imaging and memory. I was just wondering if you considered creating a licensing model for a complete process flow for the logic market. I'm just thinking that the other JDAs have had third-party licensees in the past like IBM-JDA and IMEC also has that. I was just wondering if that was something you considers, and then I got a follow-up.
Ruurd?
Yes. That -- we know this. We know Xperi, of course, well. If you follow their publications, they even do some work on our equipment. So that's well known. But our model is more the classical as we have been used -- making equipment. And then generally, these licensing models are not always appreciated. So it's not the direction we're thinking in.
Okay. And just on China, I think you had some pretty good sales in the first half. I was just wondering how that's panning out as you look into the second half, have you seen any sign of a slowdown? Or is that continuing to keep going at the current levels? I saw your OSAT orders were pretty good, but just interested to see, given all the U.S.-China trade war, et cetera, whether there's been any kind of change in dynamic there?
No change. The expansion, China for China, is still intact. And you can read that also in many other publications. So that has not come to an end yet.
There are no further questions. Please continue.
Well, then we would like to thank everyone for participating and your good questions. And if there are any more then don't hesitate to contact us. You're always welcome. Thank you. Bye-bye.
Ladies and gentlemen, this concludes the Besi event call. You may now disconnect your line. Thank you very much.