Good morning. I think we can start our presentation. Right. To start with, I'd like to say hello. Hello to you in the room and those who are following at a distance with the webcast. As you will see this year, we're going to have three speakers. Salvador is going to be replacing me in May. He's with us, and he's going to talk us through the nicest part of the presentation. He's going to enjoy it. Of course, Thomas, whom you already know, our CFO. I think some of you have met Salvador before several times. Right. 2025 is a year for resilience. There's nothing much to say about the year. We've seen how the markets reacted with a slowdown in the different cycles. That's something we talked about a year ago.
That's for the EV market, and then we talked about the important impact on solar energy. Well, you will see the figures in a minute, but we fared well with the other activities that we have in the group. By and large. Well, there's not much to say about the tariffs. That's a year ago. We were very busy with tariffs. It was a bit worrying when we saw the tariffs and the rates announced by Trump, but yet the year was good. Right. Let's have a look at our achievements. This is what Salvador will be talking about. We've seen interesting things that'll pay off later on. Let's start with the business. We've had good successes in India for a product that we wouldn't really work on. That is the pantographs.
Usually, we would be in the friction elements, and the unit sales at several tens, dozens or hundreds of euros, now at thousands of euros. There's a program that focuses on revamping and changing the fleets in India, a massive program. We are qualified. If you look at our sales, they were really good this year. Thanks to technological developments, we have gained new markets as well, like in Brazil, for instance. Now, there's something else that's important. First time that we've been nominated by CATL, the biggest battery producers in the world, a Chinese company that has 50% of all batteries manufactured wherever they are, on cars but also in energy storage. So far, they had not turned to us for protection devices, that is fuses.
Then all of a sudden, six months ago, they came to us. They said, "Could you bid on that and cover many applications?" I think there are three platforms for cars. There's stationary batteries. There's a flying vehicle, not a plane, not an aircraft, a flying vehicle that they're going to use, and therefore a huge market for us. A big potential if we're good in the months to come because we're working hard on that at present. We've been qualified as well, and that's a bit surprising, I must say. I didn't really believe in that, but the teams believed that it would be a success, so sometimes you have to follow your teams. As I said, we got a good order from the DLA, a good DLA contract for graphite applications, military applications that we can't say anything about.
We're a French company. We have a good foothold in the U.S., as you probably know, and we signed this contract with the DLA against an American competitor. Good bookings for what we call the grid. HVDC is the name of these contracts, and I think that Salvador will talk about this. There's ENGIE, one of the big players, and the other big player is Siemens. Well, Siemens is bigger than ENGIE. Then the Chinese. That's for the HVDC Chinese lines, but they're signing some export contract, and therefore we are working with them on the export contracts, not the contracts in China with them. That's a good ramp-up for us, very positive. It's connected to PV and also wind farms, offshore wind farms usually. That's for the grid.
Good sales as well, which has been the case for the past years with fuses. Good sales for fuses, and mainly in America. The beginning of the year was very good. We've written down data centers, but it's not just the data centers. The whole business is doing well. Sometimes we get 50% of our bookings for companies that equip data centers. That's really good. As you know, this is a buoyant business, and we know that this business is going to last. We know that it's going to last at least 10 years, and that's aerospace, aeronautics.
We have many certified sites, AS/EN 9100, and we'll have to be certified in the U.S., and we'll have to deliver locally in India in the future, in the near future, and probably in China as well. Now, who are the top managers? Well, Salvador will be heading the whole team, and all of these top managers are ready now. They were nominated. It's been the case since January, as you know. Salvador will be steering this boat, and he'll talk you through the business. You know him. The two segments that we're talking about have two new heads, people who were working for us. They know the house through and through, and therefore we have good Comex members and even a level down the executive committee, the Comex or ExCo.
Now, let me talk very quickly about CSR, corporate social responsibility. We've been hatching many action plans for years, and these are the highlights, if you will. That's for 2025. The emissions have gone down. Well, CO2 emissions, that is, of course. Scope 1, Scope 2, down 50% versus 2022. That's a big drop. Not many companies can boast that they do the same. What's important for us as well is to have a high level of waste recycling. If we have waste, we try and reuse our waste. 73%, that's good. When we started working on that, we were at 40% or 42% recycling rate. Now, the materials segment is working hard on that because we develop our products from A to Z, whereas for electrical distribution, we buy components. We don't produce aluminum, we buy aluminum.
73% is an average. We still have to work on materials. This year, the target is 75%. That's a recycle rate. Always difficult to go up and up. We've worked on Scope 3. Quite a lot of work that we put into that. You have to measure the carbon footprint of your products when you've sold them, when the customers have them. Well, first you have to know how customers use them, what they're going to do with the products, and then what's going to happen, if they're going be recycled, reused, and whatnot. If you look at our figures, you will see that, Scope 3 is an important scope. Usually, there's electricity that goes through our products, and we lose electricity in the wires in your homes, those who sell copper wires as well.
We've also worked on what we call Mersen Care, all types of benefits. Now, I think this is something. We have to make sure that all of the employees who work for Mersen get something from our profit. That's the profit-sharing scheme in France that we call intéressement in French, That's profit sharing. And we have more complicated schemes in the U.S., but it works locally. If the sites do well, if the results are good, we give some of the money back to the employees. 96% of the methanol workers and employees benefit from profit sharing, skills development, 20 hours of training in average per capita.
There is mentoring and many more things for the employees. And for people with disabilities, well, we have a plan. The first objective is to try and, oh, sorry, I'm out of breath this morning, I don't know why. We try to identify people with disabilities in the group and then work on the working conditions for them. So, to improve the working conditions and give them a better type of access to the workstation so that life is simpler for them. Then the energy transition, there's not much to say about this. As you can see on the slide, there's sustainable development sales. Now it's called the energy transition. It's connected to global warming.
Today is a day at 2.6 degrees above the average. So, you see, this is the topic nowadays. We've done the map again. We have 59% connected to the energy transition. There's not much change compared to the percentage we had before. We reclassified a number of products. So, and as you can see, the product delivers on markets that focus on the transition, the energy transition. Then let's have a look at the numbers of 2025 sales to start with. As you know, at DA at EUR 190 million, it's in line with the guidance, 15% of sales, capital expenditure, EUR 129 million.
We have a strong action plan to reduce our capital expenditure. So down to reach EUR 129 million. Our guidance was at a high level. And then what's very positive this year, as you can see, is the free cash flow. We thought that we would meet the target in 2025. EUR 6 million positive at free cash flow. We thought we would reach the target a year later. That's very positive for the year. Now, a few words about SiC semiconductors. There was a slowdown in 2025, which had started in 2024. I think you know that by heart as well. There were delays due to the EV market. Now, with the main clients that we have, we've worked on the PVT technology. That is the normal SiC, the monoblock, the ingots, and we've negotiated.
Our contracts that the clients were not following, well, two of them mainly. As you know, there was a massive impact on last year's EBITDA, which is not going to be the case this year if you look at this year's figures. It was quite okay, and most of our clients managed to stick to their commitments that they had in writing in the contract. Therefore, if you look at the two largest clients, the contracts have been modified and spread over the coming years with quantity outlooks that have been modified depending on the needs. We'll still have these volumes in the future, quite high volumes. Then, as you can see on the slide as well, usually sometimes people mix this with the second topic. We have graphite to produce ingots.
Now, let's have a look at the p-SiC. Now, here again, look at the EV market. Demand is not really that good for this product and for other products, but more specifically for this p-SiC product. We have a production line to produce p-SiC, and now we know that the volumes will not be that good, which was the case as well as I said for PVT. Even though now we're thinking about diversification because, as you know, we've identified other applications to use our production lines, manufacturing lines, but we will have a EUR 37 million non-cash impairment loss this year. Now, given our cash performance, the group level cash performance, we've decided the following policy, a dividend payout of 90 euro cents, that is EUR 0.9 this year.
That means a 39% payout rate by restating the net income. Thomas has some slides on that, and he'll give you more information on that. Now, there we are, as far as I'm concerned.
Um,
Good morning, everyone. I'm going to present 2025, and we've indicated solid results for 2025, starting with sales nearly EUR 1.2 billion. There are a few highlights that I would like to remind you of. We've been impacted by the negative depreciation of the euro versus many currencies, starting with the U.S. dollar, the Chinese renminbi, EUR 40 million. These are mainly conversion or Forex impacts, which does not, of course, mean we have changed anything with respect to our competitiveness. There are some scope effects. We made some acquisitions in 2024. I will go back to the prices as well, which illustrates our pricing power. I have some interesting information to share with you.
The last point, as you know, we have been significantly impacted by solar semiconductors and solar -40%, more than -40% in those two markets. Conversely, we've had very good performance in other markets, electrical distribution, for example, wind energy, rail as well. If you exclude solar, you have growth of 3.5%, which is not neutral at all. It shows that we've actually had a fairly good year. Let's take a look at the company's profitability. EBITDA amounts to EUR 190 million, which is a margin of 16%, which is very much in line with the guidance we gave at the beginning of the year. We said between 16% and 16.5%.
That's a good performance, given the solar market context, which ended up being much lower than expected. Amortizations have increased. Again, this is very much in line with our investment plan. Those investments have been made, and the expected returns have not appeared yet, and this has an impact, about 90 basis points. This has an impact on operating income. Salvador will come back to this. Those investments will continue to increase as we move into 2026, as we move forward in 2026. Sorry, this is the wrong slide. Here we go. Let's now take a closer look at the evolution of operating margins in 2024 and 2025.
I'm going to start with the price and productivity, which offset some of the inflation on raw materials, on labor as well. Labor is a significant line item on this table, 0.4% margin, and this illustrates our pricing power.
The inflation also includes customs duties, tariffs. We've said this time and again, this is not something new, but we have an objective. We define objectives, and we've reached the objective. We have the volume effect, which is negative. Some sites specialize more in solar semiconductors than others. They've been impacted. Again, this was offset by adaptation plans which partially offset these unfavorable effects, 160 basis points. Let's move to some of the adaptation measures that were taken. You may remember that we launched an adaptation plan in 2024. We said we were going to close some small operations, transfer productions. We said we were going to optimize and restructure some businesses, some operations, and in order to reduce headcounts in some locations.
This cost us a total of EUR 23 million. The bulk of this was in 2024, some of it in 2025. I think EUR 17 million in 2024. Of course, this generated significant gains, EUR 20 million in the two years, 2024 and 2025, mainly in 2025. As a matter of fact, exceeding our expectations. This is quite a positive point. The return on investment was rapid. The plan cost us EUR 15 million in cash. In H2 2024, we said we were going to reduce stocks by EUR -30 million within one to two years. We're at EUR -63 million in a year and a half, 18 months. This shows that we have been very focused on this matter in a quite successful way.
EUR 63 million once again in 2024 and then in 2025 with working capital requirement which has become quite low. Let's now take a look at margins for each of our divisions, starting with Advanced Materials, which is on the left. You can see that EBITDA is at EUR 111 million with margin of 18.2%. This is down versus 2024, but it's a moderate drop given the significant drop of SiC and solar. SiC only impacts Advanced Materials, and solar represents the bulk of Advanced Materials. This division had some significant market headwind, and it lost some of its operating margin as well because of amortizations. Investment plans were made, really made for Advanced Materials. Margins shrunk.
Conversely, when volumes will come back, there will be a real lever effect, which means that they will significantly offset things and that this will be quite considerable. Moving to Electrical Power represents half of our total business. You can see that margin. Sorry, that was my phone. The margin went up 300 basis points, which is considerable because of the price effect. Thanks to some competitiveness and competitively measures which we undertook, and they have significantly contributed to this division. The Electrical Power, you can see that margins are up in similar proportions, which is positive. Advanced Materials went down, but there is this lever effect. Electrical Power actually improved and became a more competitive division.
Looking at the overall income, net income is EUR 14 million, and this includes EUR 53 million in non-current. This includes the EUR 37 million mentioned by Luc earlier for p-SiC. This is p-SiC depreciations. We have nearly EUR 8 million in additional asset depreciations. You know, some of them are so-called underused assets in solar. Then we have the remainder which is linked to the competitiveness issues. The average debt has slightly increased. We have taxes as well. The EUR 18 million represents nearly a 25% tax rate if we restate some of the asset depreciations which could not be recovered. Moving on to the next slide. In terms of cash, we've posted a very good performance for 2025.
The operating cash flow was already very good in 2024. As you can see, we had heavily invested, but before investments, we had generated a lot of cash flow because we already started to reduce stock EUR -63 million once again, and the bulk of that was in H2 2024 and about EUR 30 million in 2025. We've had generated very good cash flow. This gives us a positive free cash flow, as Luc said, one year ahead of schedule, one year ahead of our plans. Thanks to a contraction of capital expenditure, we set EUR 160 million or EUR 170 million, or we're at EUR 129 million, and we continued to further reduce stock. Working capital requirement is 18%. It has never been so low.
This takes us to an increase of debt. We have, of course, after the free cash flow, we have the dividends. The debt is slightly going up 3% and our financial structure remains extremely robust. We have a leverage ratio of 2.2, which is very much in line with our policy between 1.5 and 2.5 leverage ratio, which is well below our bank covenants. We have significant cash reserves, thanks to the renegotiation of a USPP in two segments in 2032 and 2035. They generated some cash. We have a syndicated loan as well.
In 2025, we have a deadline when there will be no problem to reimburse it with the cash reserves. We may also use some of the syndicated credits. The average debt maturity is five years with fixed rate debt, which is 78%, so very good coverage. In summary, this gives us a fairly good balance sheet. Good resilience of our margins, a good cash, good financial structure. I'm going to give the floor to Salvador for what is ahead of us, mainly, with the focus on 2026. Thank you.
Merci, Thomas.
Thank you, Thomas. I'd like to come back to some of the key points of the company to show our strengths in the current market environment and better help you understand our perspectives moving forward. First of all, we're number one or number two worldwide in our product categories with market share ranging from 15%-30%. We are a recognized and well-established company in those, in these markets. There are several types of entry barriers. First of all, from an industrial standpoint with unmatched graphite production capabilities worldwide. Another barrier is the technical with the design of specific fuses or the graphite manufacturing procedures. These are just two examples. We're also a unique supplier of a wide range of passive components that are used in power conversion.
Last but not least, we have a unique global footprint. This footprint is a strength. First of all, it allows us to have genuine proximity with our customers, which is very important in designing customers. It also limits the tariffs. We've really seen an increase in tariffs. More than 80% of the products we manufacture are manufactured locally and delivered locally in the same geography. Last but not least, it allows us to reduce intercontinental shipments or trade, which is really an advantage in the current context. I would also like to say that our suppliers are largely local as well, which means, again, that we source primarily in the same geographies. The company operates in markets that are structurally growing midterm and long-term.
Renewable energies is one example, and we have some additional example here of growth rates renewable energy between now and 2034, for example, power transmission, power distribution with infrastructures that require new operations or new applications like energy and power storage, all the way to other applications such as data centers, for example, the transport sector, railway, electric vehicles, aerospace as well, which is increasingly electrifying as well. All these markets are connected, and Mersen has become an absolutely necessary, a much-needed player for those industries.
Well, these markets will be drivers in the long run for Mersen, and therefore, we'll be able to reach 65% of our total sales connected to the energy transition by 2030. Also, in the short term, as we know today, this is going to be a winding road, which is what we saw with solar markets and with the SiC markets. Now this year, the solar energy market is at a low level, and the turnaround point that we expected for EVs, that is SiC, is something that we expect at the end of 2027. We shouldn't forget that we are strong because we operate on different markets, though. Even though the solar and SiC markets are at a lower level than expected, other markets will be growing.
It's the case for aeronautics and the rail transport business, plus the energy transition, and that includes data centers. Which is what we can see on this slide, the following page, where renewable energies will be going down due to the fact that we'll have less on the solar front, but the other markets connected to the energy transition will be growing. On the right-hand side, we have the chemical industry markets and the process industries that will follow global economy and the global trends. All in all, in 2026, what we expect is an organic growth in our sales in between 2%-6%, with more growth during the H2 versus the H1 of 2026. The expected EBITDA margin will be at the level of 2025 that we expected, that is 16% ±50 bps.
The operating margin before non-recurring items will be impacted by the increase in amortization, which is what Thomas said earlier on, and therefore the margin will be at 8.5% ±50 basis points. The industrial capital expenditure, that'll be in between EUR 90 million and EUR 100 million, therefore a sharp decrease compared with 2025. We took a number of assumptions for raw material prices, copper and silver, more particularly, that we use in our products. The assumption is that we'll maintain a higher level of copper and silver price, but we'll offset the impact thanks to the prices that will increase that the clients will pay. What's also important to say is that given the general backdrop and the situation in the Middle East. By the way, we are following very carefully what's happening in the Middle East.
As we speak, we have not identified any significant direct impact on our activities. Now, of course, as you know, the situation keeps on changing day in, day out, so we'll keep a close eye on the developments in the Middle East. If we look at capital allocation, we're going to continue and invest to maintain the level of capacity, but also to deliver on some growth projects. The normative level will be 6.5%, CapEx on sales, of sales after 2026. We'll also continue our dividend policy, return to shareholders with a dividend that's going to be paid, which is what we've shown you. Finally, we're going to continue and use the M&A approach, our lever, which is very important for us.
By the way, if you look at our pipeline, we have quite a lot of active projects, very interesting projects for what we call the bolt-on acquisitions. In the midterm, that is by 2029, we can confirm our objectives with sales nearing EUR 1.7 billion, EBITDA margin before non-recurring items 19% ±50 basis points, operating margin before non-recurring items 12% ±50 basis points, and ROCE 13% ±50 basis points. These objectives were defined in February 2023 with the numbers that you can see on the screen. Those have to be remembered. Thank you very much for your attention. Now, Luc, Thomas, and myself are here to answer all of your questions, should you have any questions. Hello. Julien Onillon, Marex. I'd like to start with the short-term outlook.
The growth that you mentioned, which is rather easy to understand, but back to the margin levels. Okay, there's organic growth, so we might expect that this would have an impact, a lever effect on the margins in addition to or going beyond the increase in amortization. Now, have you been conservative or not? Why are margins not improving better given the sales that's going to improve? We could talk about the EBITDA margin, but we could look at operations as well, because in theory, there's the lever effect. Have you factored in a number of things like the increase in the price of raw materials? Do you think that there's going to be some type of squeezing there? Do you think that the price of energy will go up? Do you have some hedging?
Because you use energy in a way or another with coke and other elements. Could you perhaps give us more color on that? Why have margins not improved that much? Is it due to the commodity effect or raw materials?
"I'll start," says Thomas. Now, you know, as Luc said earlier on, we renegotiated our SiC contracts. We had some positive one-offs. That is, we managed to offset the situation with zero cost, and that is equal to 0.5 margin points. The second effect that's important is that we've seen a sharp increase in the price of silver and copper. Silver, well, the price increased twofold versus last year's. Therefore, if you look at our assumptions, working assumptions, well, of course, we'll try and offset part of these increases, but we can't offset all of these price increases.
The impact is a big impact, EUR 10 million-EUR 15 million. It'll vary, of course. There's fluctuation, so we'll have to have some type of hedging, or we'll need to offset that. By the way, our hedging will have bigger impacts during the H2 versus H1, so the margins will be more positive for H2 versus the H1 . The other reason as well is that we have more volumes during H2, the H2 of the year. It has to do with the phasing in of contracts, and as you know, with EVs and ACC, we have more volumes during the H2 . That's the reason. That's why the margins are not increasing that much. You know, we've given you a figure 16%, and the operating margin will do the math. It's due to amortization, the equipment that we commission, et cetera.
Therefore, all of these impacts. To answer your question about energy, it's 3% of our sales. There's electricity. No real impact. Gas is 1% of sales, more or less. Yeah, there could be some effect here, and plus transport. It's not easy to do any simulation with this, but we think we have pricing power. Then there are the indirect impacts. Frankly, we don't know how to quantify this, like the suppliers who themselves would be impacted. Very difficult to quantify this. But as we speak, well, there will be some impacts here and there that we'll probably or most certainly be able to offset completely or partly thanks to our pricing power, the price increase as we speak. Question in the room, off microphone. Jean-François Granjon.
I couldn't understand the 16% where there was no change, even though there's a growth between 2% and 6%, because apart from that, you have this operating lever that you could use. If we go to the upper part of the bracket, it's surprising that we stick to the margin that's not going to change. That's for EBITDA. Could you tell us more about this? I have a second question to ask. Very good performance if we look at Electrical Power. The margin is 13%. Is that something that could last given the general context? Thirdly, you've confirmed your objective for 2029 in your roadmap, but it's a major leap forward, I'd say, if you look at the 2026 guidance going from an 8.5 margin to more or less 12%, 10%, 12%.
Do you trust you can meet the objectives in the roadmap? What are the main drivers, according to you, to justify this improvement of your margin in only three years? Salvador. I'll try and answer. I'll answer the question on the roadmap, your third question. Well, that's true. It's a leap forward, as you said, if you look at what we're saying for 2026. Today, we know that there are M&A opportunities. We've not yet finished with the M&A pipeline. In the plan, we factored in EUR 100 million. That's M&A for 2027. By the way, this is not fully realized, and our plans now stretch to 2029, which means that we can still factor in more bolt-on acquisitions, M&As. That's my first answer. Also, we factored in an improvement in the volumes for the EV markets.
The EV market will turn around, and there's a direct impact on our products for the EV cars, but also for the SiC markets. Important levers, which means more volumes and improvements of margins, significant improvement. To answer your question about the electrical segment, you're right. The progress we made was considerable if you look at the EBITDA margin and the EP segment margin. We looked around, we looked at our peers, and our results are rather good, and yet slightly below the average of the large corporates, the very large companies. It's not easy to do this type of benchmark because they have a portfolio of products that's broader than ours.
Yet we think that this level is a level that we can sustain, and we can still improve in the years to come. Right? Thomas says, "I think I've answered your question, Jean-François." If you set aside the contract effect, we have an improvement of 50 basis points, and this notwithstanding the increase in the price of commodities. If you look at our working consumptions, you might call them conservative or not, but we can't really offset all of these increases, even though this is still our objective. I'd like to come back to solar. You said quite clearly that this year, 2026 will be complicated again. Looking forward, however, 2027 and beyond, don't you think there might be a risk that things don't change or that the future market remains sluggish? You're producing primarily out of China for the solar segment.
The market today is much, much bigger in terms of size than what it used to be. Your sales went up. You reached EUR 100 million. It then dropped. It went back to EUR 100 million, but with a much bigger market, which means that indirectly, you've actually lost market share, significant market share. It's 4x bigger now than it was in 2011, and your market share hasn't really changed. The entry barriers are less significant. The temperature required to make those products is lower than six, so you do have a real competitive edge. Here, compared to Chinese players, this is not the case, and the Chinese have been very, very aggressive on prices. How do you see the market in three, four years from now?
Do you really expect the market to pick up, and do you really expect to grow, or do you consider you may not be as competitive as Chinese players? In other words, you've depreciated some assets in China. I'd like to hear you on the Chinese market in three to four years. Again, a very complex market. Well, I guess I'm the one who should answer with 15 years of experience in this area. We've said for a while now that we focus on niche markets, and you're right, we have small market share. Chinese players are very active and are producing ingots extensively. We also work on PECVD coatings in which we have a very strong footprint.
This is a market where our competitors are not as active. 600 GW were produced in China in 2025. Capacity is 1,000. The Chinese authorities are pushing them to reduce their production. Everybody wants a piece of the cake, of course, needless to say. They wanted 1,000 two years ago. Everybody's losing money today, including our competitors with stock problems. Of course, we consider we cannot lose money for decades, but we are convinced that eventually things will come back to a more normal situation in all segments. You referred to 2011. There were 350 players in China who were supplying the solar market. The market consolidated. Today, there are five or six companies that make ingots, and maybe one or two too many.
I think we can reasonably expect the market to further consolidate, which means more simplification of the market. I don't think our competitors are willing to continue and produce while making losses. Salvador earlier presented the 2029 plan. We've remained quite modest when it comes to the contribution of solar. We have the other applications, power conversion, for example, which will contribute. We deliberately moved out of some cheap value add or low value-added markets. We remain quite active in critical components. Some of those markets are a bit less active than what they used to be. We have some other projects in China. We mentioned them last year. We also have two projects in India, Radhi and Adani, and Corning in the U.S. as well.
On those markets, we don't really have the same competition, and therefore more market share. Those markets are more appealing for us. We'll see what happens in terms of sales in both 2026 and 2027. We have some questions through the webcast. Giovanni from Berenberg. In your presentation, you mentioned an increase of prices with a more significant impact on H2. Does that mean H1 profitability will be below 8.5%, and will that increase in H2? As far as the divisions are concerned, 8.5% you said, does that reflect a deterioration of the two divisions or only of the Advanced Materials division? I can take those questions. The Advanced Materials division is most impacted because amortizations were primarily involved in the Advanced Materials divisions.
As far as operating margins, I'm not going to talk about operating margin, but EBITDA margin rather, because it really depends on the timing and the phasing. In terms of EBITDA margin, I think H2 could be better than H1 for the reasons I mentioned earlier, greater volumes and the price effect in H2 more than in H1. The currency effect and high prices are already in our P&L. Question two on utility prices, energy prices. What is included in your 2026 guidance? Is the guidance based on an increase or not? The energy costs are not very significant, as we said earlier, and I would like to repeat once again that we have good pricing power.
That's really all I can say at this point. Right now, we're looking at the indirect impact. It also very much depends on how much time all this is going to last, more so than the direct impact. A question from Mr. Dubois. In p-SiC, in your press release, you talked about diversification opportunities for new applications and new customers, which could partially offset the slowdown of the EV market. Could you clarify what specific markets you're referring to and what could be the impact on sales in 2026?
Uh, [inaudible]
Well, we're looking at multiple diversification opportunities for p-SiC. Most of those applications would be industrial, residential/building or construction. Not so much to be expected in 2026. We're looking at product qualifications for 2027. It is a bit early at this point to develop. We have questions from Gilles Chauffour, who is very passionate about electrical distribution, as you probably know. Could you help me quantify the chipset of Mersen for the 800 V direct current of Mersen versus the current applications, current infrastructure applications? In order to benefit from this very rapidly growing market between now and 2027, would it be foreseeable to have partnerships for Mersen?
Do you confirm a three chipset, which is 3x higher than 800 V DC versus 48, the current 48 V DC used in the protection of data centers? You wanna go for it? Go ahead. Well, I think what's important first is to look at how data centers operate. Is it alternating current AC or not? The fact is that there are multiple conversions, which means significant electrical or power losses. Data centers require more and more power. When you don't have enough with 220 V at home, you switch to 380 V. They're going to move to 800 V DC delivered to the data center. You need a converter to do this.
It captures alternating current, converts it into 800 DC. ST says they're going to make those chips. They're going to require I don't know how many diodes for a data center. In a car, you need 34. Not one, 34, depending on the phases. Some say there could be some interesting volumes. Right now, we're sort of waiting and, but we'll see what happens, down the road, 2027, 2028, even if the EV market remains sluggish. Coming back to data centers, however, once you have your 800 V direct current, you need converters, GaN converters. GaN converters allow you to bring that to 1.2 V, 1.5 V, for on chipsets or on microprocessors, which means much more simplicity, much more efficiency.
NVIDIA has made some recent announcements as well. They sell chips that are increasingly powerful, but they of course want to know what kind of power supply will be delivered. To answer the question, indirectly, the SiC market is going to increase a bit more, which is good news, mainly for PVT technologies. We'll see what happens with volumes, but our assumptions are quite positive. On fuses and fuse architecture, I think this should also serve. Right. The 800 V architecture is used in different applications by different customers. Each customer has their own architecture with 800 V. I recently traveled to the U.S. Everybody there is interested in data centers. There are massive investments being made, and we're seeing a move towards 800 V.
What we're seeing is that there is a race against time to build new data centers. With the current infrastructure, with the current supply chain, whatever is widely available to beat competitors. This is really what we're seeing in the U.S. The electrical architecture, including fuses, are quite positive for us because we make fuses. In data centers, you have controlled areas, but you need to supply them. You need UPSs. UPS ensure a good power supply throughout the entire life cycle of a data center. Of course, you need to supply data centers through the grid. Right now, this is a problem. There is not enough power, and this is why we're working on small nuclear reactor projects that will supply data centers.
As far as chipsets are concerned, well, we have a range of applications. It's easier to find chipsets on planes than for data centers. It's not negative. No, it's not negative at all. Next question on investments. You made fewer investments than initially planned. Does that mean you've discontinued or postponed some projects? If yes, what decisions were made? Well, as we said, 2027, 6.5% of investments, 6.5% of sales. We could do less. It depends on growth. We've somewhat reduced, but there are potential opportunities to turn this around. An easy question for Thomas: What's the average cost of your debt for 2026? That's a lot easier than chipsets. In 2025, it was 4.20. Hedging is 80%.
With the current situation, the residual 20%, I don't know. It's a bit like foreign exchange. I don't have a crystal ball, so I don't know. But if you, if you're at 420, I think 420 is a pretty good assumption. Okay. Here is a more general question. Who are our current competitors? Who are our future competitors? Well, I guess in graphite and materials, we have SGL, Schunk Group in Germany and Japanese players, Toyo Tanso, Tokai, Ibiden, and Nippon Carbon. Sorry, the question was asked from someone who does not have a microphone. They are not really in graphite. They're in ceramics. They have a small business unit. I mean, EUR 300 million. It's a fairly small operation. And then power transmission, pantographs. For the electric segment, we have a lot of different competitors.
Eaton Bussmann. In Europe, there are companies like SIBA, ETI, and many Chinese players, Sinofuse, and mainly Sinofuse. In other areas, we have additional competitors like Auxel. There are a lot of competitors. I won't, you know, give you the full list. I think that is why we intend to go for more mergers and acquisitions. A lot of competitors because we have a lot of product lines, but for each individual product line, we don't have that many competitors. We have market share that ranges between 15%-30%, depending on what the segment.
In the graphite segment, we are the only fully integrated player on the U.S. market, which means we are on site. SGL. They operate in the U.S. and Germany, but they cannot produce everything out of the U.S., and the same thing is true for the Japanese. In materials, there are no U.S. competitors.
Now back to nuclear. Here's the question. There's a German competitor, that's SGL Carbon, that says that there are opportunities to develop SMRs for graphite. Do you have the products for that? And do you have contracts that you're currently discussing? Look, well, there are many projects, and there's a lot of technology, and many use graphite in the reactors. Graphite is used as a moderator. That is, thanks to graphite, we reduce the level of chain reactions. There are other devices, of course, depending on the reactors, the HTR, high temperature reactors, et cetera. In the core of the reactor, there are different types of graphites. SGL Carbon say that they have so-called extruded graphite. 20 years ago, that was, quote-unquote, qualified. That's for the moderator.
We have a Japanese competitor, Toyo, which is still delivering isostatic graphite for a small HTR that they've had in Japan over the past 25 years, and they delivered the first inserts for the first two HTRs in China. The Chinese have been using the two technologies. SGL has a product that we don't really have in our catalog. There's not much I can say. I'd say, "Well, good, they have an interesting offer." Now, Toyo is unbeatable because the product is being used, so it's qualified. They're ahead of us. We have some products that are, quote-unquote, qualified in the U.S. Not much has happened in Europe. I could say that, yeah, that's true. We target the same projects, but we cover other parts of the reactor, and we have four or five other companies that are doing well as well.
Salvador says that, well, read the papers. There's a client who talked about the fact that we were chosen for graphite products on these markets. It's easy. Have a look at the internet. That's all we can say. We're making progress with Terra Innovatum. We can give you the name. It's one of the projects that we're working on. We'll talk about this later on, as you will see. Right. We have another question. You cover the SiC manufacturing cycle for furnaces, but you also cover the GaN manufacturing. Answer: Less. Well, there are two answers to the question. GaN can be placed on a SiC substrate for some applications. It's what you do for a car. We'll do the same. GaN is placed on the substrate with machines, MOCVD machines that we also supply.
When it's on a SiC substrate, it's the equivalent for us and very interesting. Most of these GaN devices usually use silicon. We do this as well to some extent, but to a lower level. GaN on SiC is good. I have a question about two markets and more specific contracts. The epi-SiC, I know that you've done your depreciation, so you think the market is not going to turn around. You're trying to diversify your business for the next two to three years. Yes. Now there were potentially good contracts that have suffered, ACC with the batteries, and what about this? What about the outlook for the market? There's something more positive, as you said as well, which is India. What about your sales level today? You probably have interesting contracts in the rail transport business.
What is the sales level that you expect perhaps in 2026, 2027, 2028, 2029, if we look at the longer run objectives compared to your net sales today in India in the rail business? The second interesting topic that you've mentioned is new contracts gained with CATL, and the Chinese player. Do you have estimates of potential sales with CATL? Are these busbars or fuses with your busbar technology that you hoped you would do good things with ACC? What is the future like? What about these two new opening markets for you? Could you give us a flavor of that, a color of that? What's your sales level in one, two, three, five years from now?
I'll start with ACC, says Salvador.
Well, as you've seen, they've had some difficulties in the ramp up to produce the cells, and that is something that happened throughout 2025. Look at the quantities delivered in 2025. Well, the quantities will increase threefold in 2026. During the H2 of the year mainly. We'll have the same level of activity versus what we did at the end of 2025 during the Q1 of the year, a slight growth during the Q2 , and then a strong increase during the H2 , which is why we will have more sales during the H2 than H1. The volumes, of course, are still below the levels that we had set in the plan, but we have other projects to offset this, and to reach the target for 2029. CATL. Well, these are fuses.
As Luc was saying, CATL is the biggest battery manufacturer in the world. They produce cells, batteries assembled in groups of batteries that'll be used for different market segments. First, the EVs, you have different packs together. You have to protect them. For this you need fuses for the safety of the operations of these packs. The packs can be used for energy storage. For instance, container systems where you need protection as well, given that there's electricity and current with different ranges of products. We're developing a specific range for CATL, which is one of our strengths, as I said before. Storage of energy, EVs, passengers, plus, industrial EV applications and the rail business as well. There are batteries on trains and also what we call the flying taxis that we might have seen during the Olympic Games, but they were not approved.
They have them in China. They've been certified, and it's going to be going very quickly. If you travel to China, you'll probably have these flying taxis from A to B with no driver or pilot. Now, we use battery packs as well on these vehicles, and potentially there will be fuses. As I said, there is the rail business. It means we're competitive in China with the products designed and made in China. That's a major breakthrough. Good wins with these customers and therefore later on, once we walk the talk, once we've delivered on our promise, because, you know, the Chinese market is very competitive, not just the price levels I mean, but the quality of products, but also the deliveries.
Once you've managed to show that you do a good job. We have the teams to do this. We trust we can do this in 2026. Then this would be a door opener. We could sell other products such as bus bars for the interconnection of batteries. That's my answer for CATL. We can't give you any numbers for 2026, no. This is not going to have a major impact on the sales of the group. Okay, we're nominated for a platform. What is important is to be nominated, and then we have to do our job, and you will see. Then we'll reap the volumes that they're going to sell. India, who's going to answer that? Okay, let's say more or less EUR 50 million. Nice growth in India in the recent years with a pantograph contract.
We work well with Alstom as well as Siemens, GE to some extent as well. They said that we should produce locally. That's for converters, electric trains, not just electric trains. What we're looking at several million each year. We have a subsidiary. We have six industrial buildings. That's the subsidiary. It's rather good success. We're targeting 100, I don't know when, in 2028 maybe. We're growing. It's been so hard for the past 10 or 15 years. Sales were going down, but now we're going up. There are projects for manufacturing electronic devices and to subcontract locally a number of parts that'll be used on some French aircraft that have been sold. You can see from the industrial point of view, there's some type of buoyancy.
There aren't that many competitors either and therefore, they're trying not to buy too many Chinese products at a low cost. That's not really what they want. They'd rather do the job themselves, so it's better for us. Another question in Paris. A question about the data centers. A buzzword. Now, you know that there are many companies that are trying to quantify this business in their global business, overall business. Schneider, Legrand, that's a KPI that's very important for them. Could you quantify this perhaps for Mersen? And also, we've seen massive redevelopment of data centers in the U.S. plus capital expenditure. There will be soon investments made in Europe as well in data centers. Are you ready? Will you follow the main players? Will you work with them generally in France or in Europe?
Do you think there's going to be a tidal wave in two or three years?
Well, I'll start and answer your first question on data centers. It's difficult for us to quantify how many of our products will be used in the data centers. Well, depends how you sell them, which is the marketing channel. Usually, we work with distributors. It's usually the retailers, the distributors that include these devices and products in their architectures that are developed by those who sell the data centers. I can't give you any number. I don't have them. But we're working on that so that we have more visibility of what's going to happen on the markets. We'll tell you more about this in the months to come. Your point about the U.S., yeah, we have a good positioning with our products and devices in the U.S., and in Europe we're ready as well. I think the teams are ready.
As you know, it starts with the use of existing products on existing platforms or in existing families of products, and we're good at developing specific products for future architectures. My answer is yes, we're ready for this wave. Will it be a tidal wave? When will it come? We don't know. We don't really know. We don't have the numbers, and we're seeing that there are changes. Yeah. Electrical Power, that's good for Electrical Power, EP. We've seen some interesting orders. We have more repeat orders and bigger orders, but we don't really have any numbers to give you, so we're not going to do that. Just back to some questions we received from the webcast. I will read this question in English.
You start creating 3D bipolar plate this year.
This year?
This year.
These are plates for fuel cells, so this question probably comes from an expert. We're not a big player in this specific industry. Some are made with graphite. I guess graphite or Bakelite. It's resin, a mix of resin and powder. Not really something we do. We make some for more sophisticated industrial applications in a few countries, but this is it. I guess maybe in 3D printing. 3D. That's probably what the question is referring to, and the answer is no. When you make a part with graphite, you don't do 3D printing. The last question on human robots or humanoid robots. In Mersen's portfolio, are there any products that could be used to manufacture robots? Fuses, possibly.
If you see those human-like robots, yes, some may use fuses indeed. We also sell products for AGVs, so they're automatic vehicles. Not robots, not necessarily humanoid or human-like robots, but robots. Some robots look like humans, but they're robots. We have no additional questions on the webcast. Are there any final questions in the room, please? I think there are no more questions in the room. In that case, many thanks. Bye-bye.