Operator: Welcome to the Soitec Half Year Results 2025, 2026 Presentation. Today's conference will be hosted by Pierre Barnabe, Chief Executive Officer; Albin Jacquemont, Chief Financial Officer; Steve Babureck, EVP, Chief Strategy Officer; and Alexandre Petovari, Head of Investor Relations. [Operator Instructions] Now I will hand the conference over to Pierre Barnabe to begin today's conference. Please go ahead.
Pierre Barnabé: Hi, everyone, and welcome to Soitec H1 '26 Results Conference. I'm Pierre Barnabe, Soitec's CEO, and I'm very pleased to be with you today as well as with Albin Jacquemont, our Chief Financial Officer; Steve Babureck, Chief Strategy Officer; and Alexandre Petovari, Head of Investor Relations. Before we begin, please take a moment to read the disclaimer included in this presentation. We have a lot to cover today. But before we start the formal presentation, let me share a few words about the current fiscal year. Fiscal year '26 is a special year for Soitec. As you know, I have decided to leave the company at the end of March after 4 years, and I personally recruited Albin as our new CFO, giving him a clear mandate to strengthen our financial discipline and clean up our balance sheet. This job has already been done and done very well. H1 '26 reflects that discipline and the priorities we set back in May, meaning focus on what we can control, give absolute priority to cash and take deliberate sometimes tough, actions to correct inventories and improve cash conversion. These actions have been fully launched, but the impact has just started to materialize. We are being methodical and sequential, managing our own inventories, optimizing working capital and adjusting our cost structure accordingly, while maintaining selective investments in strategic areas. At the same time, we are progressing on multiple fronts, expanding our product portfolio, preparing for new end markets and rolling out our new client and product-centric organization, which positions us to capture the next phase of growth. Our incubators introduced last May are also delivering promising results. We are identifying significant opportunities as key players explore SOI for advanced computing applications and memory. This is a large and fast-growing market, and we are at the forefront of materials innovation, combining cutting-edge R&D capabilities with the ability to industrialize rapidly and produce at scale, a unique differentiator for Soitec. These initiatives comes with a high risk reward profile. So we will remain prudent in our commitments until we see clear customer engagement. That said, recent developments confirm that our efforts are well targeted and aligned with where the market is heading. As you will see, I have asked the teams to continue executing this disciplined plan, combining financial rigor and strategic focus, so that Soitec emerges stronger and ready for its rebound. Let's begin with the main highlights of our first half -- first year -- first half year. Our H1 '26 performance reflects the actions we have taken to strengthen cash generation with lower production volumes to support the reductions of inventories. Revenue reached EUR 231 million, down 29% organic, compared with last year. Our 34.1% EBITDA margin mainly reflects the smaller revenue base and a temporary increase in inventories supported by continued volume production. Initial cost measures have had a limited impact so far, as expected, given their recent implementation. Their benefits will start to materialize in the coming quarters. Finally, our EUR 26 million operating cash flow reflects our effort to reduce production volumes to correct inventories and a temporary increase in working capital as inventories rose in H1 to support our H2 deliveries, partly offset by lower CapEx. Looking at revenue by quarter. Q2 confirms the expected rebound from our low Q1 '26 with a 47% sequential organic increase. Our first half revenue reflects different dynamics across divisions, strong growth from AI-related products with the Edge and Cloud AI division, up 34% organic year-on-year, excluding the impact of the anticipated major SOI phaseout, offset by continued weakness in mobile and automotive. Let's start with Mobile Communications. H1 revenue reflects the continued inventory correction at certain foundry customers as anticipated. RF-SOI inventories remain high, but they are going in the right direction. We expect further correction in H2 '26 and fiscal year '27. We also continue to expand beyond RF-SOI. POI remains a major growth driver with 11 customers in production and 12 in qualification. While we saw a temporary slowdown in Asia after a very strong initial ramp last year, adoptions continue to expand among leading fabless companies, supported by new design wins for flagship smartphones. Beyond RF-SOI, we continue to make solid progress in next-generation communication product with FD-SOI adoption advancing in Wi-Fi 7 SoCs for premium smartphones, confirming our position in future communication architectures. We are also progressing in our 18-nanometer FD-SOI road map as shown by the announcement on Tuesday of a design win from a key customer. FD-SOI technology brings advanced low-power computing with high level of reliability, which is critical for satellite communications applications. Our Edge and Cloud AI divisions continue to show strong momentum. In the first half, revenue reached EUR 96 million, flattish organic year-on-year, but up 34% when excluding the anticipated Imager-SOI phase-out, reflecting robust demand for AI-related products. The increase was mainly driven by higher Photonics-SOI sales, benefiting from AI-driven investment in data center infrastructures and by strong demand for FD-SOI across both edge and cloud applications. On Photonics-SOI, we are leveraging the AI acceleration across the industry, supported by large-scale CapEx investment. The technology stands out as the most efficient solution for high-speed optical interconnects, including co-package optics, which enable faster, more energy-efficiency and cost-effective data center architectures. Photonics-SOI continues on its fast growth trajectory from a very low point in fiscal year '22 to approaching $100 million in revenue for fiscal year '26. On FD-SOI, our product portfolio continues to expand, supporting new generations of AI computing devices and edge applications with strong customer engagement and committed capacity investments. On Imager-SOI, we completed the phaseout of first-generation product in H1 '26, which represented an impact of around $32 million. Residual purchase order in Q2 '26 generated a few million euro revenue. Let's move to Automotive & Industrial, where market weakness continues to weigh on activity. In a challenging automotive context, we continue to see increasing adoption of our products and rising content per vehicle, driven by infotainment, autonomous driving, functional safety and electrification trends. Power-SOI sales were impacted by inventory adjustment at customers following a strong restocking at the end of last year. We are preparing the transition to 300-millimeter to meet growing demand for battery management systems and vehicle electrification applications. FD-SOI adoption continues to progress, supported by leading foundries and IDMs, developing automotive solutions for ADAS and edge computing in radars, microcontrollers and wireless connectivity. On SmartSiC, we have revised downwards the market perspective set. When Soitec launched the program in 2021, reflecting intensified competition from Chinese monoSiC players. We are continuing to qualify 5 customers. While we are seeing growing interest in SmartSiC's efficiency benefits for next-generation power supply and data center applications, these opportunities are unlikely to materialize in the near term. Let me now say a few words about our new organization, which the entire Executive Committee has been working on for several months. This new client and product-centric structure strengthens Soitec's readiness to expand into new SOI and beyond SOI end markets and applications. It is built around 4 key pillars: one, the acceleration of our product portfolio expansion and diversification, structured around 5 established product lines, already industry standard or on their way to becoming SoCs, FD-SOI, Photonics-SOI, RF-SOI, POI and Power-SOI. Recent progress on the product development front supports our strategy to enter new markets and new applications with SOI and beyond SOI. Two, a more balanced customer, supplier and geographic base, expanding our ecosystem influence. Three, an innovation powerhouse driven by more targeted R&D investments focused on future growth opportunities; and four, agile industrial capacity management, ensuring optimized utilization of our state-of-the-art production tools and greater asset fungibility across sites. Let me now leave the floor to Albin for the financial review. Thank you, Albin.
Albin Jacquemont: Thank you, Pierre, and good morning, everyone. Let me begin with the key financial highlights for the first half, some of which Pierre has already touched upon before taking you through the details of our financial performance. As Pierre mentioned, we have mandated teams across the organization to reinforce financial discipline and accelerate the cleanup of our balance sheet. I will update you on the progress we have made on this front. Our first half results reflect the deliberate actions we have taken to initiate a reduction in inventories in the second half of the year and to strengthen cash generation, all while maintaining close oversight of customer demand and inventory levels. We delivered revenue in line with our first half guidance, although organic revenue declined 29% year-on-year, reflecting continued complexity of the market environment. Our EBITDA margin improvement is largely attributable to a lower revenue base and should be viewed in conjunction with a temporary increase in inventories, supported by ongoing production volumes. Our net result was minus EUR 67 million, primarily reflecting nonrecurring items including the SmartSiC impairment and the one-off noncash foreign exchange conversion loss, recorded in the first quarter. Excluding these nonrecurring effects, current net income was broadly stable at minus EUR 2 million. Free cash flow was minus EUR 31 million, reflecting seasonality, lower revenue and a temporary increase in inventories ahead of second half deliveries, partly offset by lower capital expenditures. Turning to the balance sheet. Our position remains solid. We closed the half year with EUR 483 million in cash and investment, pro forma the repayment of the OCEANE, which took place on October 1 -- October 2025 and with EUR 145 million in net debt. This maintains a robust financial profile with 0.5x EBITDA leverage, including leases recorded under IFRS 16 and provides us with ample flexibility to support our strategic and financial priorities. Pierre already addressed the revenue performance, so let me move directly to the P&L. As you heard from Pierre, reducing working capital and reinforcing cash generation are top priorities, and we have advanced on these fronts. First, we actively managed fab utilization to better align production with planned deliveries, thereby paving the way for a reduction of our own inventories in the coming months. Second, we launched a comprehensive cost reduction program, addressing our major cost drivers. Third, we scaled back capital expenditures. These actions are all aligned with our objective to enhance cash generation, improve operational efficiency and secure lasting savings across the company while preserving our technological capabilities. The key message I would like to leave you with is that while these actions will take a few months to translate into meaningful results, we will remain relentless, systematic and disciplined in the execution. Gross margin declined 490 bps year-on-year, driven by 3 factors: the disposal of Dolphin Design, representing 120 basis points; lower fab loading as an initial step towards reducing inventories and an unfavorable mix price effect. Going into H2 2026, do expect a significantly lower loading of our fabs and that will weigh obviously on our gross profit. Net R&D expenses decreased by EUR 23 million year-on-year, reflecting the disposal of Dolphin Design, a favorable phasing of public funding and lower material purchases linked to reduced use of pilot lines. Excluding the effects of the Dolphin Design disposal and the timing of public funding, gross R&D spend was broadly stable year-on-year, underscoring our continued commitment to technology leadership. SG&A expenses declined by EUR 6 million compared with the prior year, driven by lower compensation-related expenses, tighter control of discretionary spending and the disposal of Dolphin Design. Other operating expenses totaled EUR 46 million and include a EUR 41 million impairment loss on SmartSiC, noncurrent assets following a downwards revision of business prospects as a result of increasing competition from Chinese players and EUR 3 million downward adjustment to the earn-outs related to the disposal of Dolphin Design. As a reminder, the SmartSiC program was launched well before 2022 at a time when prices for alternative competing products were significantly higher than they are today. For context, Dolphin Design acquired in 2018 generated EUR 40 million of operating losses over the period since its acquisition. We also incurred a EUR 17 million one-off noncash foreign exchange conversion loss in Q1 of our financial year. This loss results from the reevaluation of balance sheet, foreign exchange exposures following the depreciation of the U.S. dollar against the euro with the euro-USD moving from 1.08 at the end of March 2025 to 1.18 at the end of June 2025. As background, in 2021, the company began contracting euro-denominated loans at the level of our affiliate in Singapore, whose accounts are kept in U.S. dollars. Converting euro-denominated debt into U.S. dollars had been beneficial to our financial results as long as the U.S. dollar was appreciating against the euro, and we repeatedly recorded foreign exchange gains. However, in Q1 of our 2026 fiscal year, the situation reversed, leading to the foreign exchange loss recorded this quarter. Because experiencing significant foreign exchange volatility on our results is clearly not in line with our standards, we took action. As a first step, we implemented appropriate hedging instruments to prevent foreign exchange movements from impacting our financial results. This is now in place. In addition to that, we engaged external advisers to conduct a comprehensive review of our foreign exchange risk management framework, and this review is now well advanced. Now moving to the free cash flow. First, let me note that we have aligned our definition of free cash flow with prevailing market practices. The updated definition incorporates 3 key changes. First, all tangible and intangible capital expenditures are included in the free cash flow calculation regardless of how they are financed. Capital expenditures that were previously funded through finance leases and therefore, excluded from the CapEx base are now fully taken into account. Second, free cash flow now includes both interest received and interest paid as well as other financial expenses. Previously, only interest received was taken into account in the free cash flow calculation. Lastly, the free cash flow definition now excludes inorganic CapEx, which incidentally was nil over the period. Operating cash flow was EUR 26 million for the period, down EUR 103 million year-on-year mainly, reflecting lower EBITDA and an increase in working capital driven by higher inventories built ahead of deliveries scheduled in the second half of the year. Working capital resulted in a cash outflow of EUR 57 million compared with an inflow in the prior year. This primarily reflects the seasonal buildup of inventories to support second half deliveries and the reduction in trade payables, partly offset by a decrease in trade receivables following the strong fourth quarter 2025 activity. Capital expenditures were largely directed towards industrial investments, including manufacturing tools for SOI and POI products in Bernin and Singapore, upgrades to our industrial facilities and targeted IT investments to enhance operational efficiency. This results in minus EUR 31 million of free cash flow under the new definition. We maintained a moderate leverage ratio with net debt to EBITDA at 0.5x EBITDA at the end of H1 2026. Let me conclude my remarks with a few comments on the balance sheet. As part of the financial discipline mandate issued by Pierre to the teams, we have carried out a restatement of the prior year account. In accordance with IAS 8, we have retrospectively restated consigned raw materials as inventories with a corresponding amount recorded as trade payables to reflect the transfer of control upon receipts at our sites. This restatement has no impact on the group consolidated income, EBITDA, working capital, free cash flow or equity. As a result, EUR 37 million of additional inventories and trade payables were recognized as of March 2025. For context, this compares with EUR 31 million of consigned inventories as of September 30, 2025. As of September 30, 2025, cash stood at EUR 808 million, reflecting a temporarily high level of liquidity ahead of the repayment of the EUR 325 million OCEANE 2025 bonds, which took place on October 1, 2025. Post OCEANE repayment, net cash was EUR 483 million at closing. As of September 30, 2025, we are undrawn on the maximum EUR 150 million use of proceeds loans secured from EIB. Our available liquidity post OCEANE repayment, including our undrawn confirmed revolvers was EUR 603 million. Financial debt totaled EUR 953 million, up from EUR 782 million at the end of March 2025, reflecting the new EUR 200 million Schuldschein loans and prior to the OCEANE repayment. This brings my prepared remarks to a close. As you can see, we did not shy away from making tough decisions. As I mentioned, our first half performance reflects the mandate to take decisive actions on inventories, strengthen balance sheet discipline, reduce costs and improve cash conversion. While the initial measures were implemented in the first half, we will accelerate and amplify these actions in the second half and expect to see inventory improvement by year-end. Our focus remains laser sharp on generating positive free cash flow under the new definition by the end of the fiscal year. At this point, let me pass you on to Pierre to take you through our strategic priorities and guidance.
Pierre Barnabé: Thank you, Albin. That's very clear. Let's now have a word on our guidance. On revenue, we expect Q3 '26 organic growth in the mid- to high single-digit range sequentially. For the end of this year -- fiscal year, we expect continued undershipment in RF-SOI as we pursue the inventory correction, persistent weakness in automotive and a strong momentum in Edge and Cloud AI, supported by sustained demand for Photonics-SOI and FD-SOI. On capital allocation, we remain fully committed to a disciplined and agile investment strategy. We now expect fiscal year '26 CapEx of around EUR 140 million, down from the EUR 150 million previously indicated and well below the EUR 230 million spent in fiscal year '25, reflecting our selective approach and focus on cash generation. We continue to leverage the fungibility of our industrial footprint to optimize asset utilization. On financing, we redeemed those EUR 325 million OCEANE 2025 convertible bonds on October 1 and successfully secured new funding. Finally, on profitability, our H2 gross margin will reflect fab unloading with headwinds from mix price, FX and lower volumes. Some important data points for profitability, a 10% decline in fab loading resulted in a 300 bps negative impact on gross margin. The 5% -- $0.05 removed in the euro-dollar rates represent roughly 150 basis points impact on EBITDA and EBIT margin. And our net exposure is about 95% hedged at around 1.10; as well deliberately reducing production to bring inventories down year-on-year by the end of fiscal year '26, while moderating our CapEx profile. All actions remain focused on securing a positive free cash flow for the full year. To conclude, H1 '26 was in line with our expectations. We are focusing on what we can control, taking deliberate actions to reduce inventories and applying strict financial discipline to improve our cash generation. Our new client and product-centric organization ensures tighter alignment between innovation, product road maps and customer needs, reinforcing Soitec's readiness to expand into new SOI and beyond SOI markets and applications. We are also progressing on our incubators with strong traction in advanced computing and memory. A large and fast-growing market, where Soitec materials innovation and ability to industrialize rapidly give us a unique position. I have decided to leave the company at the end of March, but I remain fully committed until then to deploying our new organization, accelerating the company's shift towards AI-driven markets and application and preparing Soitec for its rebound. We are strengthening the foundation of the company. Our growth potential remains intact in an addressable market set to expand at a double-digit pace. I have full confidence in Soitec's ability to deliver meaningful value. This concludes our remarks. Thank you for your attention. Now let's please move to the Q&A.
Operator: [Operator Instructions] The next question comes from Aleksander Peterc from Bernstein.
Aleksander Peterc: I have 3 to start with. So the first one is, if you could explain what drove the relative resilience of your gross margin and your EBIT margin in the first half? And how we should think about margins sequentially. You do cite headwinds into the second half on unloading charges. Could you quantify them? And when you say you have headwinds, is that implying a sequential decline in EBITDA margins in the second half? Second question is on your investment in SmartSiC. Is that now a total write-off? Or do you still keep assets on your balance sheet that are attached to SmartSiC? And can you also tell us what you're going to do now with Bernin 4. Is it going to be repurposed? What are you going to do aboutit? And then final question on your situation with your former licensee GlobalWafers. Now your press release suggests that the cross-licensing agreement between you and them has been terminated. But it seems to me that they continue to manufacture SOI wafers, they have a big fab being built in Missouri in the U.S., 12-inch fab that they described in their third quarter results as an SOI facility. So does this mean that they have a work around your patents? And if they do it, will others follow?
Pierre Barnabé: Thank you, Alex. What I propose is the question one regarding the EBITDA H1 and H2 to be treated by Albin as well as the first part of your second question of SmartSiC, I will take over the B4 fulfillment as well as your third question on GlobalWafers, then Albin, please.
Albin Jacquemont: Yes, sure. Alex, of course, what we are doing will have a meaningful impact sequentially and on the full year on the profitability very clearly. Like Pierre said, the priority for us is to reduce our inventories. In a context -- in market context where revenue is under pressure, it is probably not very easy. And what we are doing to achieve our objective is to significantly reduce the loading of our fabs. It's quite mechanical. When revenue is stable or under pressure, if you don't reduce the loading of our fabs, then inventories do not decline, and that wouldn't be consistent with our objectives. Now when you look at our profitability and our gross profit, I think that the 2 main key drivers of our gross profit are: first, fab loading, which is very important; and second, the mix and price. So what we will see in H2 is significantly lower fab loading, whereas typically, the loading of the fab would be higher in the second half of the year for the company. That will not be the case this year. We will see lower process costs, and we will see a much lower absorption of these process costs in the inventories. To put things -- to take -- answer your question with a different angle, idle costs or underutilization costs will be significant in H2. And that will weigh on the profitability. So yes, you should expect a sequential decline of gross profit, EBIT and EBITDA in H2. As for your second question, Pierre said that we review downwards prospects on SiC, but still, the product is -- has great technical capabilities, and we are in the process of qualifying with some customers. So we do expect some business, and the impairment that we took against our assets reflects our expectations. And overall, what we are doing is maximize the fungibility of our assets to minimize the financial consequences of these prospects being revised downwards.
Pierre Barnabé: Then I will take over, if you don't mind, Albin, on the Bernin 4 capacity. First of all, we need to keep -- we need to keep in mind that we're going to continue and we'll have to continue to produce SmartSiC. First of all, because we have 5 customers today under qualification and pre-advanced qualification, and that is progressing on automotive applications. Second, we are working on new prospects for new applications in the data centers areas and the lower efficiency, let's say, management. And third, SmartSiC as a road map going to evolve and going to improve because this product is recognized and has an excellent product. That's the reason why we believe that despite the postponement in the business plan, this product is going to find several market applications in the future. And we need to keep Bernin 4 product line for SmartSiC. That said, as you know already, B4 is busy with other applications in our process because close to 1/3 of the footprint of B4 is dedicated for refresh, SOI refresh that is, as you know, a very important piece in our process, a very important milestone in our process. B4 is already partially used for refresh SOI. And talking about rebounds, B4 going to be used for the rest of the footprint for any other application, particularly around SOI, if necessary, to produce more of the product, we will have to deliver to our customers in the coming years. And we are not concerned by the ability to use as it is already the case, B4 for SmartSiC, refresh-SOI and other SOI production in the near future. Regarding your third question on GlobalWafers, then it has been said that we terminated our license agreement with GlobalWafers. We are today in a transition period of this application. Of course, you can't make SOI without using SmartCut and our dedicated and patented processes. Then we're going to observe in the future, in the near future the way GlobalWafers going to produce SOI product in their brand-new U.S. factories. And we are today in transition and in observation.
Operator: The next question comes from Sébastien Sztabowicz from Kepler Cheuvreux.
Sébastien Sztabowicz: One on the inventory situation, both on RF-SOI at your large foundry customers, and also, on the other hand, on auto and industrial market. Where are you standing right now in terms of level of inventory in mobile? Are you still in RF-SOI, are you still expecting to go back to 11 months of inventory in December or you are a little bit trending behind this target? And the second question is on -- a follow-up on the first one on fab loading. What was the fab loading in H1 exactly? And where do you see the loading trending in H2 based on the Q3 guide and what you can expect for Q4? And the last one is on the OpEx trend for H2 because in H1, you had big subsidies impact or tax credit, I don't really know, but the OpEx were quite low. How do you see OpEx moving to the back half of the year?
Pierre Barnabé: Hello Seb, then I propose to take the first question, and I will ask Albin to relay me on the fab loading and the OpEx evolution. Then regarding inventories, what we said in July is that the equivalent 8 inches, 200-millimeter wafers in our customers' inventories, Soitec's inventories by our customers, were around 2.5 million wafers units. What we are measuring today, what we are observing is around 2.3 million, then we are clearly in a trend of depletion, that's a fact. And we want to focus on what we can measure, and this one is measurable. Then what we see is a continuous depletion of these inventories, meaning that some of the customers going to go for a minimum level of inventories. We can translate it in the fact and this is what we already said in July, and we repeat it right now is that H2 is going to be another semester of depletion to reduce the level of 8 inches wafers inventories equivalent by our customers. But we think also that the year 2026 going to see even further depletion to reach a level that's going to be pre-COVID. But what we believe is really to give you a maximum of information on what we can measure. And what we can measure is the evolution of the overall inventories by our customers. And 2.5 million equivalent 8 inches July, 2.3 million equivalent 8 inches September, and we're going to give you another point for the next call in February. Albin, for fab loading.
Albin Jacquemont: Sébastien, yes.
Sébastien Sztabowicz: In automotive, Pierre, where do you see the inventory?
Pierre Barnabé: Sorry, I missed the automotive. For the automotive, of course, we are not a major player with Power-SOI, but we see what is observed everywhere in the automotive world, meaning that there are quite big inventories still, then Power-SOI is experiencing a big drop this year. We don't see, let's say, a clear rebound for the year after for calendar '26. It seems that what we have today engaged with some customers going to be sufficient for this year and next year, provided the execution of the existing LTAs. And we don't see, let's say, a clear rebound in automotive before 2027 so far on Power-SOI, linked to, of course, a low level of electrification volumes in units of cars that is not very high. The good news is that we see a shift with more and more qualification by new customers on 300-millimeter platforms that's going to give another [ double ] of oxygen to develop Power-SOI beyond 200 millimeter. Albin?
Albin Jacquemont: On fab loading, Sébastien, look, traditionally, our loading is much higher in H2 than what it is in H1. For instance, last year, in H1, our loading was slightly below 60% in H1, and it was in excess of 80% in H2, that was for fiscal year 2025. In the current year, in H1, our loading was a little bit lower than what it was in H1 2025. And that was because the actions -- the decision, the actions Pierre, referred to have been taken in July. And there is a time line for us to -- a time lag, sorry, to adjust the production planning. So the decision we have taken will take full effect in H2. And for H2, I'm not going to give you -- I'm not going to share our forecast fab loading. But think of it as being slightly lower than what it was in H1 of the current fiscal year. So you see that compared to the fab loading in H2 2026 compared to the fab loading in H2 2025, it is much, much lower, and that's a very significant shift, which, like I said, we will have significant impact, temporary impact on our gross profit in H2 2026 fiscal year. Very atypical for Soitec. As for the OpEx, the first thing to say is that while we are telling you that our profitability will be will be lower, we do not stay idle, and we have engaged a comprehensive program to reduce our cost drivers. Our OpEx line is not very easy to comprehend in our net income because it is OpEx, what you see is a net of our gross OpEx, net of what has been charged to the process cost. So of course, what we are doing on the production has an impact on the OpEx level that you are seeing. On H1, as you could see, OpEx are down EUR 5 million. And for the full year, OpEx will be down as well. And in fact, the financial impact of what we are doing on our SG&A is higher than the EUR 5 million that you are seeing, but not fully visible on the P&L because of this charge-out framework that you see -- that we have in place.
Operator: The next question comes from Francois Bouvignies from UBS.
Francois-Xavier Bouvignies: I have a couple of quick questions. So Albin, maybe on this cost saving program that you are also delivering on top. Can you quantify a bit the cost saving program that benefit you could bring maybe for '27 once it's fully implemented or a bit more time line and details around this cost savings at the group level? And then the gross margin, I mean, what kind of gross margin do you target in terms of range post all these adjustments and inventory correction. How should we think about Soitec's gross margin in the medium term? And then my last question, if I may, is on POI. I mean it seems that you have good prospects and the penetration is increasing, but we feel like now your revenues are under pressure because of Asian customers. How should we think about POI trajectory from here? Is it going to recover like in the second half of the year? Or would that take longer?
Albin Jacquemont: Yes. On cost savings, we did tell you that we did not shy away from transactions. To give a little bit of color, we have agreed with the unions about a thorough being put in place across the company. That does include absolutely everyone, not only the production, but also the innovation and also the G&A and functional departments. So it shows, it testimonies how resolved we are at driving costs out of the system. First -- that was the first thing. Second thing is that incentive related and the profit sharing related expenses will be lower and will reflect our results. So the impact of this is significant -- is significant and goes -- is above EUR 5 million savings that you see in H1. The third things that comes on top of that is that there is a frugality which is being put in place under the direction of Pierre across the company. And all in all, we expect the savings to be significant on the year. As for fiscal year 2027, we will see lasting impact of these savings, but we don't want to hamper the potential of the company for a recovery by doing inappropriate cuts into our resources. It's too early for me to -- for us to guide on 2027, obviously.
Pierre Barnabé: Francois, on POI, to take your question. POI experienced a very, very intense growth last year to equip our Asian, particularly Chinese customers, and that really made our POI as a standard in this area. And we see clearly a relay coming from Western world customers for flagship smartphones. This year being, as we already said, a transition between these 2, let's say, these 2 moves and to shift of, I would like to say, continents. But we clearly see and we continue to see a 20% growth CAGR on POI because we see more and more adoption, more and more customers than plus 50% may be flattish for this fiscal year. And -- but we see, over the years, clearly, a strong growth, around 20% CAGR and more and more adoption.
Francois-Xavier Bouvignies: And on the gross -- and on the gross margin potential for the company, I mean what kind of target do you -- anything you can add here?
Pierre Barnabé: I'm trying to -- I will try to answer this question and help with modeling. The first thing is that we don't want any more to guide on the gross margin. And the reason for that is not because we don't have visibility on what the gross margin will be, but we don't want our action on the inventory reduction to be hampered or constrained by the guidance we gave to the market on the gross margin. So we don't guide. Nevertheless, you need to model what our gross margin will be and we'll help you in so far as we can. So a few things for you to know. When you look at our costs, of course, raw material cost is 100% variable. So needless to elaborate on this. When you look at our direct manufacturing costs and indirect manufacturing cost, I would say that around 30% are variable, 70% are fixed. So if you combine these manufacturing costs with the raw material, it's 1/3 fixed, 2/3 variable. And to go a bit further helping you with the modeling, when you look at our cost split in terms of cash and noncash that may be useful because we fully realize how difficult it is to model, I would say that our manufacturing costs are approximately 70% -- 30% minus cash and 30% plus noncash. And I hope that helps.
Operator: [Operator Instructions] The next question comes from Jakob Bluestone from BNP Paribas Exane.
Jakob Bluestone: Firstly, you made a comment, I think it was part of the SmartSiC write-down around increasing competition from Chinese players. And I just wanted to clarify, is that just purely for SmartSiC? Or are you seeing broader competition emerge from Chinese players? If you could maybe just clarify that comment. And then just secondly, I think you mentioned further inventory destocking in FY '27. So maybe if you can maybe put that in the context of where do you see the EUR 2.3 million kind of going next year? And is that your expectation that, that's the bottom? Or just kind of what are your thoughts around FY '27 in the context of the comments you made on inventories?
Pierre Barnabé: Yes, Jakob, on the first question, yes, very fierce competition coming from monoSiC has impacted our SmartSiC, let's say, initial plan that was built in 2021. And we revised it to really bring SmartSiC as a premium product that fits with the 5 customers' expectations we have today under qualification. And of course, competition is everywhere. But so far, so quite good because we are maintaining our market shares everywhere, including in China. As you know, if we look at just RF-SOI that used to be our mono product a few years ago, and we're still, of course, observing cautiously this market, we are proceeding with many tier downs of smartphones coming from any integrators, any OEM, and we clearly see that SOI is a standard, first for particularly radio front end and more and more for filters with our POI solutions. And in many cases, it is Soitec solution. Then our market share today are, let's say, protected because we continue to innovate because we are able to deliver in terms of quality to our customers wherever they are, including in China. But that said, we are not naive, and we are very picky in the observations on any players trying to promote SOI-like solution. But today's impact, we have clearly disclosed that concerns the monoSiC particular competition that is changing the trajectory of SmartSiC. If we look at the evolution of the 2.3 million 8 inches equivalent inventory Soitec by customers, we clearly -- we're going to continue to undership. It means that the trend of depletions going to continue semester after semester for sure. Then in H2, we're going to undership. And we do believe that even further, we're going to undership in the year '26. We don't know, at the end of the day, what's going to be the objective of each of our customers. It's going to depend on their strategy of inventories, but we are taking cautious assumptions, meaning that we do believe that many of them, a large part of our customers, going to reach the pre-COVID level of inventories, and we prefer to have this cautiousness than under shipment, depletion of the inventories, and it is highly probable the 2.3 million to decline semester after semester at minimum end of year '26.
Operator: The next question comes from Emmanuel Matot from ODDO BHF.
Emmanuel Matot: I have 2 remaining questions, maybe for Albin. First, what do you target in terms of inventories in the balance sheet, what normative level would be growing given the current visibility you have for the business? I mean maybe an indication as a percentage of your sales revenue would make sense. And second, has there been any progress in the tax adjustment procedure with the French authorities?
Albin Jacquemont: Very good question, Emmanuel. The company said in the previous year, I wasn't there, that the normal working capital requirements as a percentage of revenue should be within a range of 30% to 40%. And I fully subscribe to what has been said. We will not guide for now in terms of components of working capital requirements, but I do confirm that 30% to 40% is something, which is a level, which is adequate, which should lead you to the conclusion that there is much cash generation opportunities ahead of us. Because obviously, as things stand today, we are at a much higher level. So that's for the working capital requirements level. As for the tax reassessment. So we are done, as you know, on responding to the request of the tax authorities. We did not receive an answer to our response yet. We will -- I will not forget. In the meantime, we have strengthened our defense, worked a lot with independent experts. I could not say more than that, but it should bring you some level of confidence on this.
Operator: The next question comes from Craig Mcdowell from JPMorgan.
Craig Mcdowell: Just the first one on pricing. You've talked about pricing in the context of gross margin. I'm wondering whether you can translate that back to what that's doing to your top line? And is this pricing pressure, more negative pricing pressure on particular products? Or is it particular customers? Any commentary probably on pricing impacting your sales would be helpful. The second question, I just appreciate your thoughts at this early stage on the merger between Qorvo and Skyworks, I understand both of them are your customers. Just if you could give any thoughts on what that might mean for both pricing, but also how that might impact your sort of channel inventory worked on, presumably, they'll look at some kind of synergy on inventory if they do merge.
Pierre Barnabé: On the pricing than what we can see that, of course, there are continuous pressure depending on the product. But overall, we are in low single-digit decrease and quite limited because we are also promoting more and more high and added value product. We are also -- we are having road maps. There are projects that is increasing the overall value. And we are protecting, thanks to a good mix and a better mix, the level of prices to compensate the pressure we are getting. And of course, as a kind of parallel of this pressure that we are limiting at the end of the day, we are benefiting from lower prices from bulk providers, thanks to the fierce competitions we have today. Of course, we continue to have this discipline and to invest in our existing road maps to invest in our innovation and R&D to keep higher the value and the prices of our product and putting in the market new products and new features. Just I will not comment that much on the Qorvo-Skyworks, let's say, ongoing project of merger. What we can tell you is that these 2 companies are customers, but complementary customers. We don't see any de-synergies and potentially rather synergies, but it's going to be a long process, and we'll have to -- I'm sure, to comment later when they're going to be completed.
Operator: This concludes the question-and-answer session. I'd like to hand the program back to Pierre Barnabe for any closing comments.
Pierre Barnabé: Thank you all for following our H1 '26 analyst call and for the quality of your questions this morning. The next date in our agenda will be the release of our Q3 '26 revenue on February 4 after market close. In the meantime, we'll be very happy to host you in Shanghai on November 25, where we'll be hosting our first Soitec China Day with key industry leaders and customers from the China SOI ecosystem. This ends our call for today.
Operator: Ladies and gentlemen, the conference is now over. Thank you for participating in the conference. You may now disconnect your lines. Goodbye.