Hello, everyone and welcome to the presentation of Siltronic's Q2 2025 Results. Please note that this call is being recorded and streamed on Siltronic's website. The call will also be available as an on-demand version later today. Your participation in this call implies your consent with this. At this time, I would like to turn the conference over to Verena Stutze, Head of Investor Relations and Communications of Siltronic AG.

Thank you, Elaine. Welcome, everybody, to our Q2 2025 results presentation. This call will also be webcast live on siltronic.com. A replay of the call will be available on our website shortly after the end of the call. Our CEO, Michael Heckmeier; and our CFO, Claudia Schmitt, will give you an overview of our financials, the current market developments and our guidance. After the presentation, we will be happy to take your questions. Please note that management's comments during this call will include forward- looking statements that involve risks and uncertainties. For a discussion of risk factors, I encourage you to review the safe harbor statement contained in today's press release and presentation. All documents relating to our Q2 2025 reporting are available on our website. I now turn the call over to Michael for his remarks.

Thank you, Verena. And a warm welcome also from my side. Let's start with the key messages of today's call. Wafer demand remains subdued continuing the soft trend we've observed over recent quarters. Despite this, we achieved our targets for H1 this year and delivered solid results in Q2 2025. Regarding the tariff situation, even though things seem to be clearing up lately, we have to analyze this in detail, including implications for our sector and our business. So far, our overall end market assumption from the beginning of the year is unchanged. The most significant headwind for us is the weakening U.S. dollar against the euro. This is why we've aligned our full year 2025 sales guidance based on these FX developments. Our new FX assumption for the second half of 2025 is a euro-U.S. dollar rate of 1.15 compared to our previous assumption of 1.08. Consequently, we now anticipate that full year sales will be in the mid-single-digit range below 2024 levels. Before we look at the details of our performance in the second quarter, I'd like to take a moment to recap the progress we've made in H1 2025. The key highlight was the successful completion of important prime wafer qualifications on our new fab, paving the way for depreciation to begin in August. At the same time, our cost and cash initiatives are progressing well, further emphasizing our commitment to operational discipline. And last but not least, we remain fully on track to complete the phaseout of the small diameters business in just a few days by the end of July 2025. This marks the successful conclusion of a highly professional and well-executed transition process, thanks to the outstanding collaboration and commitment to everybody involved. Let me give you a broad overview of our development in the second quarter and Claudia will provide a detailed financial overview shortly. Quarter-on-quarter, our sales declined by 5%. This development was in line with our expectations. While we saw an increase in wafer area sold, this was not sufficient to fully offset the negative impact driven by FX effects and to a lesser extent by price effects. Despite the decline in sales, we achieved an improvement in our profitability in Q2. EBITDA reached EUR 86 million in Q2, up from EUR 78 million in Q1. Consequently, the EBITDA margin increased to 26.3% compared to 22.6% in Q1. This positive development was largely driven by nonoperating effects, which Claudia will explain. CapEx came in at EUR 126 million, primarily related to our new fab in Singapore. Consequently and as anticipated, the net cash flow continued to be negative at EUR 83 million. On a positive note, our market share among the major competitors remained stable in the first half of 2025, underscoring our resilience in a challenging environment that continues to affect all wafer manufacturers. Let's move to the financials. Claudia, please.

Thank you, Michael. A warm welcome from my side as well. I'm pleased to walk you through our performance over the past quarter and highlight key developments. As Michael mentioned, sales in Q2 developed in line with expectations, totaling EUR 329 million, a 5% decline compared to the previous quarter. While the wafer area sold increased, this was more than offset by headwinds from the U.S. dollar and to a smaller extent, pricing. The product mix remained largely unchanged compared to Q1. On a positive note, EBITDA rose to EUR 86 million in Q2, marking an EUR 8 million increase quarter-over-quarter. In addition to positive contributions from fixed cost dilution, this development was supported by 2 main factors. First, Q1 typically includes seasonal effects such as vacation accruals that are more pronounced at the beginning of the year. Second, Q2 benefited from, I would categorize as nonoperating effects. One the one hand we adjusted the valuation of our spare parts following the phaseout of small diameters. On the other hand, compared to previous quarters, a somewhat higher amount was capitalized for innovation projects that have now entered the development phase. Importantly, our overall R&D spending remains in even under the current cost discipline as we continue to invest in our future. With higher EBITDA despite lower sales, our EBITDA margin improved significantly, reaching 26.3% in Q2, up from 22.6% in Q1. In the comparison of the first half of '25 versus '24, FX effects had only a limited impact on the top line performance. However, we observed noticeable effects within the P&L lines, other operating income and expenses, where FX-related impacts are recorded. In H1 last year, we benefited from positive hedging results, resulting in a net gain of EUR 5 million. In contrast, H1 this year, showed a net expense of nearly EUR 6 million in these line items. While Q1 was negatively affected by hedging losses, Q2 brought a new dynamic, with hedging gains driven by the sharp weakening of the U.S. dollar. However, these gains were more than offset by valuation effects on recorded receivables at the reporting date. To put this into perspective, within just 1 quarter, the U.S. dollar weakened significantly against the euro, moving from 1.08 at the end of March to 1.17 at the end of June compared to 1.04 at the end of last year. As depreciation remained almost unchanged quarter-over- quarter, the positive development in EBITDA also translated into a higher EBIT of EUR 24 million, an increase of EUR 9 million compared to Q1. Looking ahead to Q3, we expect depreciation to rise significantly starting in August as our new fab in Singapore will begin to be depreciated. Turning to the financial result. A slight decline was offset by lower tax expenses. Taking all these factors into account, Q2 net income came in at EUR 15 million. Let's now turn to the key developments on our balance sheet. At the end of June, total assets amounted to EUR 4.9 billion, slightly below the EUR 5.0 billion reported at the end of March. CapEx in Q2 totaled EUR 126 million, once again clearly exceeding depreciation of EUR 63 million. However, the FX valuation effects related to our Singapore entities had a counterbalancing impact, resulting in a slight decline in fixed assets. Inventory saw a slight increase mainly due to lower write-downs on spare parts. Trade receivables declined partly as a result of FX valuation effects at the reporting date. In contrast, other receivables increased, reflecting positive marketing -- market values of our FX hedging positions. Operating cash flow fell short of covering our payments for CapEx, leading to a decline in cash and securities to EUR 535 million. Our equity ratio remained stable at a healthy level of 43%. As communicated in Q1, we drew EUR 53 million of our syndicated loan in Q2, which is reflected in financial liabilities. The increase was partially offset by FX valuation effects related to our Singapore dollar loan at the end of June. Trade payables, primarily those related to investment activities decreased as planned from EUR 280 million at the year-end '24 to EUR 233 million by the end of June. Alongside the refund of prepayments and FX valuation effects, this contributed to a noticeable reduction in liabilities and prepayments. CapEx in the first half of 2025 totaled EUR 222 million. Given our full year CapEx guidance of EUR 350 million to EUR 400 million, investment activities will be lower in the second half of the year. Our strategic focus remains unchanged, the ramp of our new fab in Singapore and the necessary steady-state CapEx level to support our operations. Looking beyond 2025, we still expect some incoming equipment from the initial ordering for our new fab. However, new equipment orders will only be placed if we see clear positive market trends. We continue to provide updates on the structure of our financial liabilities as illustrated on left-hand side of this slide. As of today, our total loan drawdowns amount to EUR 1.42 billion. This includes EUR 53 million from our syndicated loan newly accessed in the second quarter. Our revolving credit facility of EUR 127 million remains entirely undrawn and fully available to us. On the right, you will find the corresponding maturity profile. As previously communicated, repayments will start in '25, beginning with a modest tranche of EUR 65 million in the fourth quarter. Our balance sheet also reflects short-term prepayments of EUR 42 million, which will be reimbursed within the next 12 months. Overall, our liquidity position remains robust with EUR 535 million in cash and securities, excluding the undrawn revolving credit line, which further strengthens our financial flexibility. Turning now to net financial debt. As shown in the bridge, Siltronic closed the year '24 with net financial debt of EUR 734 million. In the first half of 2025, we generated an operating cash flow of EUR 80 million, negatively impacted by working capital effects, prepayment refunds and rising interest payments. As anticipated, CapEx payments in the first half totaled EUR 250 million, exceeding the investment by around EUR 30 million. Consequently, net financial debt increased to EUR 903 million by the end of June. Looking ahead, we expect net cash flow to improve significantly in the second half compared to H1 level, mainly driven by positive working capital effects and an improved cash flow from investing activities. With that, I'll hand it back over to Michael.

Thank you, Claudia. Turning to the market outlook, even though the situation around the U.S. tariffs begins to clear up, it remains volatile. This makes it difficult to assess the broader impact on global GDP growth and demand for wafers in the end markets. Nevertheless, we would like to share our current view on the end markets. Looking at individual end markets, the smartphone market has softened slightly. Meanwhile, PCs have shown a notable uptick compared to our March assumption, driven by a stronger-than- expected Microsoft Windows 11 impact. Server demand remains robust, fueled by the continued AI momentum. The automotive segment has seen a slight decline, partly due to lower sales of electric vehicles. Overall, our March assumption is still valid and the end markets are expected to grow approximately 7% this year. The majority of this growth comes from content and only smaller share from unit growth. However, elevated inventory level continue to weigh on this positive end market picture, resulting in a significantly lower volume impact for the wafer industry. Our internal market assessment indicates that there was no meaningful inventory digestion in the first quarter as most of Q2 data of our customers are not yet available. Inventory levels across key segments remain elevated, continuing to weigh on wafer demand. Memory inventories are still elevated. Power inventories have reached record levels, further limiting short- term demand and prices. In contrast, logic inventories look the best. Overall, Q2 did little to improve market visibility as elevated inventories will postpone our demand recovery. As you all know, we have so far been calculating with an FX rate of euro-U.S. dollar 1.08 in our guidance. However, due to recent developments in the FX market, we've adjusted this rate to 1.15 for H2 '25 and updated our guidance accordingly. It's therefore worth taking a closer look at this. As you can see on the left side, in 2025, we have a U.S. dollar exposure of more than 80% of our top line, while the majority of our EBITDA costs are euro-based. This makes us sensitive to exchange rate fluctuations. Let me explain our updated U.S. dollar sensitivity based on our 2025 exposure and the revised exchange rate of euro-U.S. dollar 1.15 for H2, a change of USD 0.01, including the highly correlated Singapore dollar would impact our full year sales by approximately EUR 10 million and our EBITDA by around EUR 6 million before hedging. Our hedging strategy remains unchanged. We hedge up to 18 months ahead based on expected foreign exchange net exposure. In closing today's presentation, we would like to share our refined guidance for 2025. Although we are more than confident in the mid- and long-term growth of the silicon wafer market, we expect elevated customer inventory levels and related volume shifts to continue influencing the next quarters. Furthermore, there's clear uncertainty in the market from U.S. tariffs going forward and the continuous change in regulations. We maintain our full year 2025 guidance based on a constant exchange rate of euro-U.S. dollar 1.08. However, as explained before, we have updated our expectations for the second half of the year using a revised exchange rate of euro-U.S. dollar 1.15. With this new assumption, we now expect 2025 sales to be in the mid-single-digit percentage range below 2024. Previously, we had anticipated sales to be roughly in the region of last year. For Q3 '25, we expect sales to come in below the second quarter. This is primarily due to intra- year shifts in delivery volumes with a significant portion now scheduled for Q4, while EBITDA margin forecast remains unchanged at 21% to 25%. We've also refined our depreciation outlook. Depreciation is now expected to range between EUR 340 million to EUR 400 million. This adjustment reflects improved visibility and the impact of a weaker Singapore dollar. Our guidance for CapEx, EBIT and net cash flow remains unchanged. With this, we conclude our Q2 '25 results presentation and Claudia and I are happy to take your questions. Thank you very much for your attention. Elaine, please open the Q&A session.

[Operator Instructions] And the first question comes from Amelia Banks from Bank of America.

I was wondering if I could just ask a question on your margin. Is it possible if you could quantify the decrease in cost of sales from the write-down of the spare parts?

This is Claudia. Yes. Just to give you a ballpark number, this was, let's say, a mid-single-digit amount in Q2, which impacted our EBITDA positively.

And is that quarterly?

That's a onetime effect.

We will take our next question from [ Daniel Jessy from AT Group ].

So I have 2 questions. The first one being, you're mentioning that you continue to have a stable market share among your main competitors. However, relative to your peers, your dynamics are slightly slower now on customer and product mix. How is then your share stable right now? And in addition to that kind of also we see China now accelerating in 300-millimeter. Sure, it's not leading edge yet. But still do you see any pressure there if you look at the overall share, is it then decreasing? And my second one would be on guidance. Basically, you lowered it on FX reasons with a sequential deterioration in 3Q. Could you elaborate what gives you the confidence in the 4Q order as it was already pushed out twice and now end markets continue to be sluggish and customer inventory is high.

Well, thank you, Daniel. A couple of questions. Let me take the market share question first. Market share, of course, is an average of all the different customers. And indeed, we have some negative mix effect there, which means there are effects that certain customers buy less in certain, let's say, segments. On the other side, this is opposed by market share gains at other customers. So that in a overall average consideration and that's what we're talking about, we can say our share is stable. With regards to the China question, we do not see very specific or very accelerating China dynamics. And when you study, let's say, the wording around this of our peers, they are also, let's say, more optimistic and more pessimistic statements. So we would see ourselves in the middle of this range. So China is on the one side, a very important market for us. China competition is the most serious, the smaller the wafer diameter is. We see 200-millimeter, I would say, as a mixed bag. They are still very attractive segments for us where we can play our technology-leading position. And in 300-millimeter, particularly in the higher specs, as you mentioned, leading edge, there's still a huge technical gap between the Chinese players and the 3 major players that are able to serve leading edge. And with regards to your guidance question and the patterns in Q4, so these are particular and individual customer postponements of volumes. So we have allocated those volumes now to Q4. And I can say we have reasonable confidence that the customers will execute and we are, in a way, positive that this will happen as predicted in our guidance.

We will now move to our next question from Constantin Hesse from Jefferies.

Also for me, if we could talk a little bit about, Michael, the outlook overall. I mean, clearly, this bottom has just been dragging on and on and on and it just feels like it's almost never ending. Inventories, if I look at the SUMCO charts, are still extremely high, both in -- I mean, logic, obviously a little bit better. But memory, of course, pretty bad, power getting only worse. What I'm trying to figure out now is, I think even if we want to be conservative on the top line and the P&L going into '26, how should we think about CapEx in '26? Now I think you did say during the call that you will not invest in additional equipment unless you see a sustainable improvement in markets, which, I guess, implies that CapEx will obviously come down next year.. But I think your maintenance CapEx is currently running at about EUR 200 million. Is there any CapEx that you already have committed -- expansion CapEx beyond maintenance CapEx that you already have committed that you have to spend in '26? I mean it would be just pretty good to have a -- if we could have a clearer picture of how we should think about investments in '26 because clearly, your balance sheet is -- it's levering quite significantly. So it would be good to have a bit of a picture there.

Thank you, Constantin. The outlook, I mean, you're absolutely right. This inventory situation is hanging on for quite a while. We see the good progress in logic. Memory, I mean, the overall picture isn't good. Still individual customers are on the right track on memory. Also there's a huge variety of different statements from different customers. And power, indeed from all what we see is further creeping up and this might be related to the very, let's say, poor overall automotive sentiment and industry environment. On the other side, we have this end market growth of 7%. So I mean, it's very difficult to predict but it should be and will be, hopefully, a question of time until those 7% are making its way through this inventory -- through the supply chain and also eventually landing at the wafer industry. But it's a way to go and it's very difficult to predict how long this will still need. I mean your '26 question is a bit a difficult one. Today, we don't guide any '26 numbers. However, maybe around your specific CapEx point, let me emphasize a couple of things. So we communicated the steady-state CapEx of around EUR 200 million but we also said it's an average number. So there could be years below, there could be years above. So that's maybe something you might want to have in mind. And indeed, there is, let's say, a certain CapEx hangover still from our fab in Singapore that will trail into '26. So overall, we would see CapEx coming down and we will specify this more concretely when we come closer to the '26 and the related guidance time frame.

I mean, that's very, very helpful. And on the next question, if I may, I think part of the depreciation adjustment was driven by a improved knowledge of equipment activation. So I'm just wondering, does this at all change that, all the legacy outlook that you used to have on depreciation of EUR 500 million plus a year, once Singapore is being fully depreciated, just to have a rough idea.

Yes. You're absolutely right. With the progress of investment projects, you gain a better visibility on the time line. And we had to do the adjustment due to FX. So -- but we stick to our communication that depreciation will be above EUR 500 million, not this year, definitely not because we have the guidance out. But once FabNext is in full depreciation, going forward, we will see the EUR 500 million plus in depreciation.

Okay. This is great. And then lastly, on the long-term agreements, I think, Michael, last time, I think in Q1, you said there are no major long-term agreements expiring either this year or next year. If you could just confirm that? And then second of all, is there any reason to believe that the bargaining power of Siltronic, SUMCO, Shin-Etsu, may be at a place where if you do have to renegotiate these long-term agreements, say, in 2, 3 years, given the current market conditions, could you potentially experience significant pricing pressure here? Or is there an understanding between the semi players and the wafer manufacturers and I don't want to sound naive here; but would there be an understanding that they do understand how much money you just invested into these plants and there will be some kind of in the core that pricing would probably be under pressure but it wouldn't be a huge amount. So I'm just trying to get an idea of the potential negotiation dynamics of yourself and your customers given the current market environment?

Yes. Thanks, Constantin. First of all, let me confirm indeed no major LTAs expiring this year or 2026. And the major LTAs, which we concluded in the framework of our new fab in Singapore have a very long duration. So we talk and almost like 2028 or even 2030 there, yes. So there's a very, let's say, overall robust situation for us around the LTAs. And further LTAs, yes, we also concluded smaller ones. And there are -- here there are segments that are still very significant in demand and that is also reflected in pricing. On the other side, of course, we will not conclude major LTAs in this current market environment for sure. But customers have a rather strategic perspective on this year and some talked about LTAs of a couple of years for certain products even in the current environment. So it's really kind of portfolio. You have to imagine one small one can go away and another one is coming in but the major ones are fixed for a long period. And with regards to pricing, I can also confirm that all the LTAs are adhered to as in contracting price. And the price effect we were describing is happening outside LTAs and that is around 1/3 of our business.

We will now take our next question from Harry Blaiklock from UBS.

My first one is just around, I guess, I mean, you kind of answered a bit but why you think inventories are taking so long to come down. And in particular, with Asian manufacturers having a higher portion of LTAs than you do, quite significantly higher, do you think you might be suffering because they continue to deliver contracted volumes despite higher inventories? And then I guess to follow up on that, if you and competitors continue delivering LTAs, how can we expect inventories to come down, especially when they haven't budged even when you've been saying that end demand is kind of greater than what you've been shipping and what wafer demand is.

Yes. Why is this hanging on for so long time, is indeed in a way the key question here. I think it's still the situation that the chip manufacturing is in a way bifocal. There are a small number of companies that are centered around AI chip developments. And they're doing extremely well and you can study the quarterly and monthly numbers, which are going up significantly. However, when we analyze this in more detail, we see that this is primarily value-driven today with technical developments in a position to charge much more for their chips than they used to be for previous generations. At the same time, the amount of wafer area sitting underneath these chips is not growing significantly, or only moderately. So we have those decoupling of volume and value growth at certain customers. And then, of course, we have others, particularly those in the power segment, whose business is not, let's say, developing pretty aggressive. So that means their volume demand and their wafer each situation is still tuned down for quite a bit. And this in a way, totals to the overall picture we're describing here, healthy or reasonable end market demand of around 7% but not landing yet vastly and some parts are progressing as we talked about logic. But vastly in its totality, not landing at the wafer industry. And your question was whether Asian peers are insisting on LTAs or whatever. Everybody tries to insist on LTAs as much as possible. And I think the clear evidence that this is affecting everybody to the same extent, is our market share statement. So we are stable. So there is no indication that we are doing significantly better or worse than our peers. It's really an industry-wide phenomenon and affects all of the wafer manufacturers.

Got it. And then on the ramp of FabNext, I know you said it kind of depends on market conditions. But based on your current view of that and how the market will develop into '26. Can you give some color around how much capacity you think will be added in Singapore this year and next?

I mean we talked about the initial ramp capacity. It was 100,000 per month by end of last year. So now that is further build up. We're reluctant to talk about very specific numbers in relation also to equipment, et cetera, that would be a highly competitive information. So therefore, yes, this will be further ramped but we will be also very careful with the ordering of new tools, as we explained. And that we will only trigger once we have the indication that markets are really ticking upwards again. So for the time being, we work with what we have. We have certain hangover of CapEx, which means there are still some effects trending CapEx-wise into next year, as we said already. And then there will be a decision point when we have market, let's say, clearer indications to further bring capacity into this fab. So detailed numbers is not what we want to communicate for competitive reasons.

Got it. And then just a quick one on Q3. Are you able to give a bit more quantification on kind of below the level? Is it slightly below the level, mid- to low single digits, secondly below the level of Q2? Any color would be helpful.

So I mean, you can do a lot of maths, right? You have our full year guidance, half year is over. We gave some indication where Q3 will sit and that Q4 should be further elevated. Giving more details would be very difficult as we have those quarterly phasings and quarterly closing effects typically letting in a few millions left or right in the end of the quarter. So therefore, I think all you can do is to do the calculations, which should be then already pretty precise.

Got it. No worries. And one last quick one just on -- I'm not sure whether you disclosed in the past but in terms of the debt that you have, are there any covenants that we should be aware of?

You are asking for our covenants, right? Yes, have covenants in place but we haven't communicated so far any contract details and we will stick to that. So we won't disclose any details here.

We will now take our next question from Florian Treisch.

Yes. Just, I mean, a quick follow-up on my end. It is mainly around phasing Q3, Q4. I mean you mentioned that end of July, the smaller diameter part will go out of the equation. Can you maybe quantify the impact in Q3, Q4? And I mean, in general, I mean you have pitched the story that the next -- or the quarter after the next quarter will be better. I mean, you mentioned you have a decent confidence that the shifts from Q3 are going into Q4? Is there something different in the structure of it compared to the last couple of quarters? Or is it the same crystal ball you like to mention in the quarterly calls?

So, thank you, Florian and I think very good question. So #1, the small diameters will be closed as scheduled by end of this month. The revenue effect from that is very small. For total year, it was mid-single-digit percentage of total revenue. And now let's have this, it's only 7 months. So it's really small. So we will not see any significant effect from this in the second half. The crystal ball is still there. And let's say, the overall view on the future is still for. I think we explained that in great detail. Nevertheless, the particular phasing between Q2, Q3 and Q4 this year is something different as it's the result of individual customer discussions and allocations with different customers and in consents with different customers to the different quarters according to their demand needs. Therefore, there is, I think, a different level of confidence of our quarterly phasing, our quarterly statements we made for this year. The overall situation is still a bit in the cloud. And I think I would really like to differentiate those 2 in terms of how large our trust and confidence is.

We will now take our next question from Martin Jung from BNP Paribas.

I just have 2 questions. The first one is on the pricing impact that you've seen in the second half. The guidance implies now relatively flat sales development in the second half. So it's right to assume that the small diameter wind down, which has seen a price pressure and also the ramp of FabNext volumes that typically or should carry higher prices should need to at least flat pricing development in the second half? That's the first question.

So I mean, overall, we don't give quarterly or half year pricing statements or comments. But the effects you were mentioning like a small diameter and ramping of the fab should not affect the pricing at all because the pricing is more market thing where we clearly can confirm that around 2/3 of our business are in LTAs where the prices are up as contracted. And the other 1/3, which is maybe a bit more smaller diameter and 200-millimeter is more under price, let's say, discussions, which leads then to our overall price situation as described. So we do not see a very particular quarterly pricing effect and it should not be affected by the SD business closure neither and nor the Singapore [indiscernible].

Okay. And then just a follow-up on this cost stuff that you had in the second quarter. I mean you mentioned the mid-single-digit number from the write-downs. You also had a positive effect from capitalized R&D, negative effect from the revaluation of receivables. Can you just disclose what the total number and the cost base was considered one-off and what we can extrapolate into the third quarter? And then just on hedging, I mean, would you expect to realize hedging gains now in the second half at current FX rates?

So yes, we disclosed -- we just disclosed the rough amount of the valuation effect, mid-single digit. Regarding R&D, you should have a picture, if you look at our P&L, where you see the development quarter-over-quarter in R&D costs. That gives you an idea and this will also translate into the next quarter. So R&D costs will stay on roughly that level because we will continue to capitalize those activities that we started right now to capitalize.

Let me emphasize, this is not affecting our R&D activities and our operative R&D works, as we didn't cut there at all. So the activities on the ground, working on R&D and tech development are unchanged.

Okay. And then just on hedging...

And regarding hedging, yes, in total, we had a positive hedging result in Q2 but it was counterbalanced by those valuation effects. Nevertheless, for Q3, we have some beneficial, let's say, hedging positions in place for the second half of the year. So there's a strong movement always in the U.S. dollar, so you can't really make a math on valuation effect. But given that the U.S. dollar would stay at 1.15, like it's today. We should have a positive effect from hedging in the second half of the year. I would like to add something which you haven't mentioned but which is important for the second half of the year and this is the ramp cost. As we start to depreciate FabNext in August, there's also an effect in ramp costs. We will see them in the P&L from August on. Until now, we capitalize them on CapEx.

Great. Can you just remind us how much it roughly is in the second half?

Yes. We don't disclose a number here. On a full year basis, the effect is not so big but on a quarterly or half year base you will definitely see it.

We will take our next question from Jimmy Huang from JPMorgan.

First, I would like to ask about how do you think about potential impact of Chinese competition within the China market and in international markets? Because we see that 5 Chinese silicon wafer suppliers, they reported their 12-inch revenue was around USD 100 million in 2021 but it has increased to about USD 1.1 billion in '24, only for 12-inch silicon wafers. So I still would like to ask about your views on potential Chinese competition for international suppliers.

Yes. thank you very much. I mean Chinese competition is definitely on the ground and they're progressing and developing. So we see their capability strongest in the smaller diameter place and in 300-millimeter, as you highlighted, they're starting with some lower specs, some test wafer activities. We do not see a lot of activities outside China. And here, we think geopolitics could even be contained and locally, as you know, the tariff discussion is hot between China and the U.S. and some other countries. On the other side, we also have, let's say, good business in China. A lot of customers appreciate our quality, our technology and our, let's say, technical support and capability there. So yes, there is a Chinese dynamics and China's progress. But I think it's a very long time to go until they would be in a situation to deliver premium products, particularly such as leading edge or high-bandwidth memory or something like that.

Yes. And second, would like to ask about the leading edge logic silicon wafer because previously, people think there are 3 major suppliers, including you and Japanese companies. But we heard that another international silicon wafer supplier from Taiwan is also relatively confident about their progress within the leading edge logic, i.e., from their U.S. fab. So how do you think about potential like more competition from this international supplier in the leading edge markets?

So thank you very much. I mean I must not and cannot comment on individual competitors. However, when you talk about the new fab in U.S., of course, we have to take in mind that qualifying a new fab and we experienced ourselves in the last couple of years, I can say, it's a very, let's say, long and difficult process with customers. You might remember when we had the first wafer ceremony in late '23 and now we're starting depreciation here in August this year with being fully qualified with the prime products with our customers. And if we had no leading edge experience, it would have taken even much longer time. So therefore, yes, everybody is ambitious to go into the leading edge segment and provide premium products. However, we know there's a clear entry barrier and it's, let's say, long and very difficult path to go there. And customers feel quite happy with 3 volume suppliers of leading edge. So there's not, let's say, a very aggressive need for another one.

[Operator Instructions] We will now take our next question from Robert Sanders from Deutsche Bank.

I just was hoping to ask about the FabNext scale that will give you accretion on the EBITDA margin level. Is there like a [ 100,000 or 200,000 ] level where you think that could -- that ramp could start to be accretive rather than dilutive at the EBITDA margin level? Second question would be about what cash flow burn you expect in 2025? And then lastly, if you could just -- although you're not willing to disclose covenants, I think it would definitely benefit investors to understand if you're at least close to breaching covenants. And so if you could just confirm that you're not actually close to breaching covenants.

Thank you, Rob. And I take your first question about the new fab and then Claudia will talk about the other elements of your questions. I mean the good news is, now FabNext is qualified in major products. And with that, we have, for the first time, the opportunity to ramp volumes, to balance volumes between our global fabs' footprint. So that means we can actively steer now volume developments in the fab. And we will do that through, let's say, an optimum manner. On the one side, there is an interest to load the fab, let's say, as quickly as possible to come into the volume area where it becomes margin accretive, which will take some time. And of course, the target here is unchangedly to deliver those EBITDA margins and also 50% out of that fab, which only will happen, with what we called a significant loading, as you might remember. So that's the clear path going there. On the other side, we must not underutilize our existing footprint in Germany to avoid too heavy idle cost. And it's, let's say, the ongoing task of our global operations team to steer this mix in a optimum manner. But for sure, FabNext will see volume increase now because the way is clearly paved with prime [indiscernible] we've done, particularly achieved in the last couple of months, where we now say the fab is ready and that's also the reason why it's depreciated starting in August. For the other elements of your questions, I hand over to Claudia.

Okay. Rob, I'll start with the cash burn. Yes, our guidance for net cash flow this year is significantly negative but still better compared to previous year. And I can only add that, yes, we had a cash burn in H1. And let me put it like that, that was the majority of our cash burn this year. We will see definitely an improvement of net cash flow in the second half of the year, mainly driven by, as I mentioned, positive working capital effects and then improved cash flow from investing activities. We will drive down our investing activities in the second half of the year. And regarding the questions, financial covenants, I can only underline that as of today, we expect to stay within the financial covenants. So unfortunately, no more details on that.

[Operator Instructions]

[indiscernible] I hope you hear me. A few questions. The first one on the inventories. I mean I'm not sure if I got that correct. So your inventory analysis is basically based on customer inventories from the first quarter. Is that correct? And then related to this, particularly in the memory business, where do you see the crossover for the inventories where customers should start to see higher order activity or start with higher order activity? I mean, we have seen Micronet and SK Hynix, where inventory days in Q2 came down much stronger. And related to this also is in memory, maybe your customer mix rather a headwind than really the overall demand.

Yes. Thank you very much for your questions. And indeed, as you mentioned, the Q2 picture is not fully on the table yet there. So a couple of other statements are really, you could say, outdated or, let's say, based on Q1 data where we have the full picture of the inventory for all the 3 segments, logic, memory and power. You're right and I said it already, within memory, there are some differences between different customers, some reported a bit, let's say, progressing data. On the other side, the question might imply, do we have a negative exposure or customer mix, which I can deny clearly because it's again coming back to the overall market share. We have the very reliable data from this independent semi organization. So if that was a significant influence, we would see market share differences, which we don't see. So I would say all the wafer manufacturers are exposed to different customers and as everybody is at every customer, of course, with different market shares. On average, those major players are affected similarly by this inventory picture. So there's nobody really better off than the others.

Got it. And then on the customer pushout, you mentioned from Q3 to Q4. In previous calls, you mentioned that customers are pushing volumes to the second half of 2025. And now in this call, you mentioned especially this one customer -- is that really related to this customer? Or is the big picture still a lot of customers are pushing out volumes, sticking to the LTAs but pushing out volumes? And are you willing to share what type of customer or which end market this pushout comes mainly?

So thank you very much for this request. I mean, what we see as, let's say, the quarterly pattern of volume and of our sales development is, the total of a couple of, let's say, individual customer discussions and conclusions. So some of them wanted to have more in Q2, others wanted to have more in Q4, et cetera. So it's really a pattern of many customers. We're not in a situation to disclose individual customer patterns or individual segments there. But what we see is now the consolidated picture. And as you know, we have a couple of huge customers and some of them were also, let's say, demanding specific allocations.

And then final one, if I may. On the smaller diameters, the question was asked previously and you mentioned the impact is, yes, small. I understand that on the full year basis, however, if the smaller diameter is 5% of the sales, which basically means from last year sales is roughly around EUR 70 million headwind. And that would be probably 10%, 11% of your second half sales. So it would be, I think, helpful if you could give a little bit color around what this is going to shape out between Q3 and Q4, the phasing out?

The phasing out is ongoing already for more than 1 year. Yes, so we were in that business already on a kind of phaseout level. And when we say the overall market share is stable, then, of course, you could conclude and that's definitely true that some other segments, we did overcompensate the decline in SD. So I would not -- and we don't see any particular further effects in the second half. It's all been worked into our guidance already.

At this time, I would like to turn the conference back over to Verena Stutze, Head of Investor Relations and Communications at Siltronic AG.

Thank you, Elaine. This concludes our Q&A session. Thank you for joining us today. We will release our Q3 2025 figures on October 28. On this slide, you can also see our next IR events. Thank you. Bye.

This concludes today's call. Thank you for your participation. You may now disconnect.