Welcome to Arkema's third quarter 2025 results and outlook conference call. For your information, this call is being recorded. It will take place in a listen-only mode, and you will have the opportunity to ask questions after the presentation by pressing star and one on your touchtone telephone. I will now hand you over to Thierry Le Hénaff, Chairman and Chief Executive Officer. Sir, please go ahead.
Thank you very much. Good morning, everybody. Welcome to Arkema Q3 2025 results conference call. With me today are Marie-José Donsion, our CFO, and the investor relations team. To support this conference call, we have posted a set of slides which are available on our website. As usual, I will start with some comments on the highlights of the quarter before letting Marie-José go through the financials. At the end of the presentation, we'll be available, as always, to answer your questions. Let's comment first on the economic environment, which remained, as you know, challenging. We noted weaker-than-anticipated trends in the U.S. over the summer. The lower demand is probably a reflection of ongoing uncertainty around the tariffs and frictions in adjusting supply chains. On the other hand, Europe and Asia remain consistent with what we have seen since the start of the year.
Europe at relatively low levels, and Asia still with a positive dynamic, in particular in China. The negative currency impact was also slightly stronger than in Q2. Despite this challenging macro environment, our growth pockets, which are at the heart of Arkema's strategy, delivered substantial growth. As a matter of fact, our sales were up 20% in several key markets, namely batteries, sports, 3D printing, healthcare, and new- generation fluoro specialties with low global warming potential. This positive momentum is also supporting the ramp-up of our major project current on slide seven, but all in all, that was not sufficient, obviously, to offset the strong macro headwinds of the third quarter. With the Q2 results, we shared with you that this major project should bring EUR 50 million additional EBITDA in 2025 versus 2024.
I am happy to convey that we reassessed the progress, and we can list the impact to EUR 60 million. This contribution is essentially supported by the strong momentum in PVDF for batteries, biosourced Pebax in sports, and 1233zd fluoro specialties in building insulation. PIAM has also showed good growth since the start of the year, thanks to new smartphone models and Dow's flexible packaging adhesives starting to contribute. This is certainly less than the initially estimated EUR 100 million attribute to the tough environment. Nevertheless, this project showed good momentum, and the setup for 2025 is encouraging. As you know, these projects are already fully financed and therefore are included in our capital employed, with only limited contribution to the P&L until now.
In this regard, the group will gain around 2.5 points of ROCE from this project over the next years, in addition to the improvement of the cycle, which will benefit everyone. We can mention also the start of two new plants in Q3 in the U.S., both on budget and on schedule. The new 1233zd unit, a fluoro specialty with low emissive impact, used for building insulation or thermal management, and our new DMDS capacity for refining and biofuels, with an impact on earnings still limited in 2025. In addition, the new refined clear transparent polyamide plant reached mechanical completion. This unit, downstream of the polyamide 11 plant in Singapore, is expected to be operational in the first quarter of 2026. Given the tough environment, I'd like to stress that all things are fully mobilized on a daily basis to best manage the current economic and geopolitical context.
We ran a number of cost-cutting initiatives, as shown in slide three, and are on track to deliver the targeted EUR 100 million of fixed and variable cost savings by year-end. The cost alignment will continue, and we strive to, again, offset inflation in 2026. In addition, Arkema stays disciplined in capital allocation. You see that we made progress in working capital management and delivered EUR 200 million recurring cash flow compared to last year, despite lower earnings. This cash generation is fully reflected in the reduced debt, maintaining a robust balance sheet. Arkema will once more reduce CapEx next year to EUR 600 million, while continuing to optimize its working capital. Despite all the efforts of the Arkema teams, EBITDA was down to EUR 310 million. Looking at the results by segment, you could recognize the different profiles of each product line.
Adhesive Solutions and Advanced Materials are more resilient, with earnings affected by lower demand, while net pricing was only slightly down. In contrast, there was more volatility in Coating Solutions linked to the low cycle in upstream acrylics, while the old- generation fuel gases in Intermediates reported a seasonally lower outcome. I already mentioned the ramp-up of our major project, which reflects the execution of our growth strategy, but also our ability to work in parallel on two tasks. We focus on optimizing our operations in the short term, but at the same time, secure our growth potential on the long term. Both prepare us to be ready in the new year when the macro will again be more supportive. As highlighted before, we follow a strategy focused on five identified high-growth markets, where we continuously look for new opportunities.
In this context, I am happy to announce that we will expand our potential in the attractive advanced electronics market by adding a new structured platform dedicated to data centers. You have heard details of it in slide five of our Q3 presentation. By dedicating joint initiatives to this powerful market, we see significant growth prospects, though starting from a new base. Besides, I would like also to emphasize Arkema's success in the battery market in Asia. Our bet on LFP batteries is clearly a winning one, and our strategy to expand our asset base with modest CapEx in China, Europe, and the U.S. turned out to be relevant, putting us in a good position to grow in this dynamic market.
We recently inaugurated a new laboratory dedicated to the next- generation of batteries, using an innovative dry coating process for electrodes that significantly reduces the cost of battery production, while lowering its carbon footprint. This innovation illustrates that PVDF long-term growth potential remains significant, while offering premium margin when you target, as we are doing, the high end of the range. I will now hand it over to Marie-José for more details or a more in-depth look at the financials before we discuss the outlook at the end of the presentation.
Thank you, Thierry, and good morning, everyone. Let's start with Arkema's revenues. At EUR 2.2 billion, Q3 sales were down 8.6% year-on-year. They were impacted by a - 3.9% currency effect, reflecting mainly the weakening of the US dollar against the Euro, but also from other currencies, including the Chinese yuan and the Korean won. Volumes were down 2.5%, reflecting the lower demand observed in the U.S. over the summer and the overall weak demand environment in Europe. On the other hand, we continued to benefit from a positive dynamic in Asia, and more particularly China, mainly driven by high-performance polymers. The price effect was a - 3.7%, impacted essentially by the acrylic cycle and the old generation refrigerant gases. All other activities showed a more limited price decrease of 1.3%, with a slightly negative net pricing. The benefit from lower raw material costs works progressively through the supply chain.
Q3 EBITDA came in at EUR 310 million, the currency effect representing around EUR 15 million negative. Looking at the performance by segment, Adhesives EBITDA reflected the weak demand in industrial adhesives and the disappointing summer in the U.S., notably in flexible packaging and construction. On the other hand, construction business grew in Asia thanks to positive momentum in efficient buildings and remained broadly flat in Europe. The performance of Bostik continues to be supported by our ongoing work on efficiency and our price discipline. Finally, the integration of Dow's adhesives brought a limited contribution this quarter due to the softness in the U.S. market, notably. In Advanced Materials, the EBITDA was essentially affected by the volume decrease in performance adhesives. They were impacted by the weak demand environment in Europe and in the U.S., as well as the reorganization of our Jarrie site in France in hydrogen peroxide.
On the other hand, high-performance polymers volumes were stable, benefiting from strong growth in Asia. The margin of the Advanced Materials segments remained overall at a good level, 18.8%, with HPP maintaining its solid margin level of 20%. In Coatings, EBITDA was essentially impacted by the low cycle conditions in the upstream acrylics. Sales declined in the U.S, in particular in the construction and decorative paints markets. The performance of the segment was therefore significantly lower than last year. Lastly, Intermediate's EBITDA included the usual seasonality of the third quarter, as well as the impact of the evolution of regulations in the U.S. and Europe in refrigerant gases. Depreciation and amortization stood at EUR 168 million. They included the amortization of new production units, which started up in the course of 2024, as well as in 2025.
This led to a recurring EBIT of EUR 142 million and a REBIT margin of 6.5%. Non-recurring items amounted to EUR 48 million. They include the typical EUR 35 million of PPA depreciation and EUR 13 million of one-off charges, notably refreshing costs linked to the reorganization of our hydrogen peroxide in France. Financial expenses stood at EUR 33 million. They increased versus last year, reflecting mainly the increased cost of our bonds, as well as the lower interest of invested cash. Consequently, Q3 adjusted net income stood at EUR 78 million, which corresponds to EUR 1.04 per share. Moving on to cash flow and net debt, Arkema delivered a solid cash flow, as you could see in Q3, with recurring cash flow standing at EUR 207 million. This reflects our continuous initiatives to tightly manage our working capital and integrate also decreasing CapEx versus last year.
The working capital ratio on annualized sales stood at 17.3%, and total capital expenditure amounted to EUR 131 million, in line with our objective of EUR 650 million for the full year 2025. Net debt amounts to EUR 3.4 billion, including EUR 1.1 billion of hybrid bonds. The net debt to last 12 months EBITDA stands at around 2.6x . Note also that we continue to refinance our 2026 bond maturities, with the issuance in September of a EUR 500 million green bond, with an eight-year maturity and an annual coupon of 3.5%. This has also enabled us to extend our debt maturity now to 4.6 years. Thank you for your attention, and I'll now hand it over back to Thierry for the outlook.
Thank you, Marie-José. Going into Q4, the macroeconomic environment remains challenging, marked by low visibility and weak demand in the U.S. and Europe, so no surprise there. In this context, as already said, optimizing the short term by working on fixed costs, CapEx, working capital remains among our top priorities. We are on track, as mentioned, to achieve around EUR 100 million of fixed and variable cost savings in 2025. More specifically, our numerous initiatives on fixed costs will enable us to offset inflation in both 2025 and 2026. At the same time, we continue to build Arkema for the future. It is very important. By maintaining our efforts in R&D as well as in sales and marketing, focus on the key attractive markets identified by the group, supported by the major projects that you are well aware of.
Taking into account the macro environment that remains challenging and the softer than expected demand in the U.S., for the time being at least, we aim at delivering an EBITDA of between EUR 1.25 billion and EUR 1.3 billion in 2025, with a midpoint globally consistent with the consensus and a recurring cash flow of around EUR 300 million. I thank you now very much for your attention. Together with Marie-José, we are certainly ready to answer any of your questions.
Thank you. If you wish to ask a question over the phone, please press star and one on your telephone keypad. If you wish to withdraw your question, please press star and two. Again, please press star and one on your telephone to enter the question queue and wait for your name to be announced. Our first question is from Martin Roediger, Kepler Cheuvreux. Please go ahead.
Yes, hello and good morning. I have three questions, if I may. The first question is for Thierry. Regarding the reason for this additional guidance cut this year, it seems you are getting more concerned about the U.S. market, especially about the weaker construction market in the U.S. How do you see the near-term prospects in the U.S. construction market? The other two questions are for Marie-José. Other chemical companies have released bonus provisions in Q3, some already in Q2. Have you done that as well, and do you plan that in Q4? Is that part of the EUR 100 million cost savings this year, which tackles both fixed costs and variable costs? We know that bonus belongs to variable costs. Finally, on cost savings impact on your P&L, your SG&A costs increased in Q3 quarter on quarter and year over year.
Why do we not see the cost savings in your SG&A line? Thank you.
Okay, Martin, so on the first, thank you for your question. First, on the first one, no. What we see today is that, yes, in the summer, the U.S. was weaker than expected. Not necessarily weaker than Europe, because Europe was already weak, but does that mean necessarily that we are more pessimistic on the U.S. going midterm? I think U.S., and we have a long experience there because it's 35% of our total sales. Our experience in the U.S. is that things are moving quite fast. This country is very agile, so you can have a quarter which is disappointing, and two quarters after, it goes in the other direction. I will be cautious on the answer. We just comment on what we see in Q3.
We think Q4 will be in the same vein, but I will be cautious in extrapolating what it means for the following quarters, because just it's a nature of the U.S. to be more reactive, and we have a little bit more volatility than we can have in other parts of the world. Marie-José?
Regarding cost saving, maybe more broadly, just to remind what we aim at doing, basically, the objective is to deliver EUR 100 million cost saving. I would say 2/3 fixed cost, 1/3 variable cost. When we look at the fixed cost component, for sure, they incorporate the bonuses. This has been adjusted progressively as we progressed in the year based on, let's say, the revised guidance we gave to you. There is no last-minute, I would say, effect that I expect in Q4. It has been progressively factored in the publications. When you look at the evolution year- on- year, basically, you see a slight increase, which means right now we are not fully offsetting inflation, but we are actually quite largely offsetting inflation. I consider inflation amounts to roughly EUR 60 million-EUR 70 million a year on fixed costs for Arkema.
Clearly, we are limiting the effect of increased fixed costs right now at, I would say, only a total of EUR 15 million. Clearly, we are producing or delivering the effort to massively compensate this inflation effect. I'm not sure why you think it's not visible, but it's definitely visible internally when we look at our metrics. No particular effect of provisions, which, when you look at the balance sheet, are rather stable over the period. Hope it answers your question.
Okay. Thank you very much.
Maybe to add also to Marie-José, when we say that we remove inflation for next year, this means that we take into account the fact that, as Marie-José said, that part of the savings this year are linked to bonuses, and we know that we will need to offset that next year, but it is part of the game, and we take the challenge.
The next question is from Tom Wrigglesworth, Morgan Stanley. Please go ahead.
Morning, everybody. Thanks for the presentation. Two questions, if I may. First one is around the data center presentation that you made, and specifically the refrigerant gases. You talk about foraying for chillers. What does that do to the market? Will that rapidly tighten the current offering of HFCs and HFOs? Secondly, with regards to the immersive heaters, is that going to be a higher margin product than your current emissive refrigerant gas business? That is a couple of questions there. Secondly, really on a kind of higher level, obviously, if I look at the bridges for all of this year, pricing is going down faster than raw materials. Is that a function of mix, i.e., you are losing higher value products more so than you are retaining raw material gains? Or are you having to give back price to hold on to volume? Thank you.
Okay. Thank you, Tom, for this question. Very different in nature. With regard to data center first, as you could see, you have a presentation page five. It's a first, I would say, exchange with you on data centers. It does not include, for the sake of your knowledge, it does not include what we do in battery for data centers. This means what this stationary battery is, it's in the battery platform. This means that when we say more than EUR 100 million of sales in 2030, starting from a base which is today, our first estimate is a bit more than EUR 10 million, so a significant increase, it does not include what we are doing in battery. As you mentioned, chillers will be a key point on new generation of refrigerant.
Today, the sales there are very, very limited, but we see strong potential for low GWP fuel gas in chillers. It depends, as you know, on the technology of data centers, but we see very good growth with good margins, but it's only one point among six or seven. We see also growth for high-performance polymers. It can be PVDF, it can be polyamide 11 for wire and cable. It can be also ballistics in flooring or waterproofing. It can be even for direct-to-chip cooling, kind of PVDF, etc. We have plenty of applications there, but as you mentioned, HFO is certainly one element. With regard to pricing, in fact, it's interesting to see more in detail what is happening in pricing. In fact, we separate acrylics monomers, and fuel gas, which obviously are affected significantly in pricing.
The first one because of the cycle, the second one because of the evolution of generation. It's not you don't compare apple to apple because of the quota mechanism. Clearly, Intermediates is a big it certainly is a big impact on pricing. Outside of Intermediates, it's far less. Outside of acrylic monomers, it's even far less. For Adhesive Solutions and Advanced Materials, the net pricing is close to neutral, slightly negative, but close to neutral. The pricing itself is around -1%. Now, with regard to raw material, what is happening is that, as you know, the supply chain, you have some stock, supply chain along, and it takes time to work through for the raw material to work through the P&L. You have a little bit of a time lag between the pricing effect and the raw material.
We are not worried at all on, I would say, Adhesive Solutions and Advanced Materials. You have some examples where we are a bit more, especially in China, under pressure in pricing there, but on the other side, we have positive pricing in some other areas, including our new business development with a high price. For us, the question of pricing, but since the start of the year, is really concerning refrigerant for the old generation and acrylic monomers.
Thanks for that, Thierry. Very helpful.
Thank you.
The next question is from Matthew Yates, Bank of America. Please go ahead.
Hey, good morning, everyone. A couple of questions, if I may, just like to follow up on Tom's one around the data center. As you say, it's the first time you've really mentioned this. I'm just trying to educate myself a little bit. Where are you in the commerciality of some of these products? Are they technically developed? Are they qualified with customers? Are they already generating sales? Or is this more of an ambition, and there's a lot of work to do over the next few years to actually get these products to market and generate some revenue? That's the first question. The second one, I'm not sure if it's for you, Thierry or Marie-José, but I wanted to ask about the dividend. We know that traditionally, Arkema had sort of other strategic priorities, and the dividend payout ratio was relatively low.
Compared to the amount of free cash you are now generating of EUR 300 million, it is effectively 100%, and you have got your leverage creeping up to three times. Can you give us an idea as to how important and how sustainable that dividend is when you get to the end of the year and you debate that with the board? Has it got to the point where if we do not see a macro improvement, paying out so much is going to infringe on your strategic flexibility and whether you want to do CapEx projects or M&A, whatever it is? I would just like to hear your sort of philosophy around the dividend. Thank you.
Thank you, Matthew, for this important question, clearly. With regard to data center, as I mentioned, we have around a bit more than EUR 10 million of sales, which means that we have already a commerciality of this product and that most of the products we are talking about have already been developed. What takes time, and this is why we have a ramp-up up until 2030 to be above EUR 100 million. It does not, as I mentioned, take into account the PVDF for batteries in stationary which go to data center, which is not insignificant, which is already fully commercial. On what is outside, I would say we have already a certain maturity of the technologies themselves. It is more a matter of developing applications with some qualification, etc.
It will take the time it takes, knowing that in data centers, the technology is really continuing to evolve, as you may know, significantly. Even for you would have already matured sales in certain applications. Anyway, you need to work on the new applications, the new technology, and to adapt your products and your new business development to this. To answer your question, we are in the middle of your question, I would say. Some commerciality, mature technologies, but applications are not fully mature. We need to ramp up, and it will take a certain number of years. I think it was mature enough to share the data center call with you and will have the opportunity to come back in far more detail next year on this topic of data center, which is, as I mentioned, joining the advanced electronics platform.
With regard to the dividend, to a certain extent, the answer is in the question. The good thing first is that we are in absolutely perfect conditions in specialty chemicals and in chemicals overall. In spite of that, with a free cash flow, we cover the dividend. I think it's good news. You know that the dividend return to shareholder is a very important policy for Arkema. The idea is to make it sustainable. It has been in the past. You have not a year which is looking like the other one. To make it sustainable, beyond what I've just said, you could see that next year, CapEx will be at EUR 600 million versus EUR 650 million this year. It's a difference. Our project will also ramp up. I think we stay with the same idea of sustainable dividend.
Obviously, it has to be decided by the board. We will have this discussion normally before the presentation of full year results. This is the philosophy of Arkema.
Okay. Thank you, Thierry.
The next question is from Georgina Fraser, Goldman Sachs. Please go ahead.
Hi. Good morning. Thanks so much for taking my questions. I've got two. Nice to speak to you, Thierry and Marie-José. First question is on HPP. We've seen a decent amount of CapEx and also the acquisition of PIAM go into this division. I just wanted to hear from you, how has the performance of this division been versus your expectations? Should we think about the fact that we're maybe at low utilization rates at the moment? I just want to try and gauge what kind of upside potential there is here and maybe why the performance has been a bit lackluster. My second question, a little bit of a follow-up on the dividend or cash flow outlook question. EUR 600 million in CapEx next year is still quite high.
Could you give us a sense of what your maintenance CapEx is and how much flexibility you have there if it was needed? Thank you.
Yes. Okay. Thank you, Georgina. Hello. On the first one, first of all, if you look at HPP, we have four lines which are very interesting. We have PVDF, specialty fuel gases, polyamide 11, and we have PIAM. I would say all these lines are growing lines on which we have, as you mentioned, spent a high amount of money for development, for growth, for CapEx, and for acquisition. Clearly, this year is a challenging year for all chemicals companies. HPP is not immune. As you could see, and it is part of the performance of advanced materials, HPP has resisted correctly compared to what we can see outside, but below our expectations because of the macro. The projects themselves are ramping up okay, but they are not immune to the macro. We, at the beginning of the year, were expecting more from HPP.
The good thing is that we think that all the strategic moves that we have done with HPP were the right ones, that the line is certainly one of the most resilient inside Arkema. The prospect of growth, even if there have been delays to a certain extent because of the macro, the prospect of growth remains quite significant. With regard to utilization rate, it's clearly that in consistency with what we see from the macro, they are more on the low side. Do not forget that we are also optimizing our inventory as everybody is doing on the supply chain from customers down to our suppliers. Because of that, we have, let's say, also adapted the utilization rate of our site. We consider that HPP will continue to remain a bright spot of Arkema in the coming periods.
With regard to CapEx, I do not share with you the fact that CapEx is still quite high at EUR 600 million. And to share with you, because we have discussions with all of you, including investors, some are telling us the contrary that maybe it is a bit too low. I think for us, we think it is reasonable in this kind of environment to take down the CapEx to something which remains relatively okay, not to jeopardize our mid and long-term growth. It would be a full mistake. Some companies have done that in the past in chemicals, and now they regret it. We try to stay consistent over the years. On the other side, we were in 2024 at EUR 750 million. We are at EUR 650 million this year. Next year, EUR 600 million.
I think it's a good adjustment of the CapEx in order to take into account the evolution of the macro. With regard to which part is maintenance, modernization, legislation, CapEx, I would say that we have around, yes, EUR 400 million of when we were at EUR 650 million, we were on about 60, 60/ 40. This means that it's EUR 400 million, yes, around EUR 400 million. The rest are really productivity and development CapEx, but which are more of smaller scale compared to what we have been doing in the past three years with the major projects.
Thank you.
The next question is from Jean-Luc Romain, CIC Market Solutions. Please go ahead.
Hello. [audio distortion]. Can I tell you? Is that better now?
No, we hear nothing.
Romain, could you please get closer to the microphone, please?
I am in the microphone.
We hear a second voice behind you.
Yes, okay.
Okay. Maybe we can take another one and then come back to you when we fix your topic.
Okay. The next question is Chetan Udeshi, JP Morgan. Please go ahead.
Yeah, hi. Thanks for taking my question. I just had a bit more philosophical question. It seems from all your comments that your view is this is much more a cyclical phenomenon in the sector, and you do not want to necessarily take any radical steps for that reason. Is that understanding correct? Because when I look at your numbers, and it is not just Arkema, right? I mean, to be fair, it is across the whole sector. Everybody is suffering from the same issue. A lot of time, at least my personal opinion, it is blamed on demand weakness, which may or may not be the only reason. I am just curious, from your perspective, you do not see any structural changes in the industry that would warrant more structural changes in how Arkema is set up?
I mean, in terms of whether you need to have all of the businesses that you have within Arkema, or maybe it's better to monetize some of them given the multiple. Thank you.
Thank you for the question. It's a good question, at least on the paper. Is that from what we see today, which part is, let's say, short-term cyclical, and which part could be more structural? I would say we have been in the business, as you know, since 20 years, and frankly speaking, and you may have forgotten, what we have seen in 20 years are significant shifts of the world. You have at the same time, and it can be positive and negative. When you see structural change, the negative is always one-to-one with the positive. You have opportunities and challenges at the same time. It's true today. It was true in the past 20 years.
This means that what we see today is a combination of, and this is the majority of what we see by far, some macro-related cyclical issue, and it will come back. We are absolutely convinced of that. Then you have some more structural change. You have more, let's say, protectionism. You have more regionalization. You have more aggressiveness from Chinese competitors. By the way, we also take advantage of it because we are strong in China, and we have enjoyed quite a good year in China. Yes, this is the world of today. You need to be agile to adapt permanently. Where I do not agree with you is this question of radical steps or whatever. I think if you look at the, for example, portfolio of Arkema, 60% of the business, which was at the origin, has been sold. We changed completely. We made acquisition.
I think it's part of our business life, and we will continue to take the steps that make sense to do. We are thinking all the time, but not necessarily sharing what we are thinking with all of you, obviously. No, I think we are at a good level of reaction. We try really to manage the two horizons at the same time. One is really short-term, working on the cost, working on the cash, and we think we are doing quite a great job thanks to the team on that. At the same time, having in mind that the frame in which we are operating is shifting a little bit, we are moving in order to adapt to that. This is what we have sold in the recent years, the PMMA, and we bought Ashland and PIAM.
Even if these two are not in terms of ramp up exactly where we would have thought they would be, I can tell you they are far more resilient than many businesses in chemicals. It goes in the right direction. It takes time. I think we are really on the right topics, and we are doing what we need to do. It does not prevent us from thinking all the time if another evolution could be meaningful or not. No, no, we recognize that. It is good for us. I have always said that Arkema would not be the Arkema of today. The world has been easy and lighter than you would have had the same players as they were 20 years ago.
I think the fact that there is disruption, maybe we suffer short-term, but we think that in the long run, it gives opportunity to companies like us, to companies that are reactive, agile, ready to question themselves, to take new opportunities, and to make a difference with the other guys.
Okay. Thank you very much.
The next question is from Emmanuel Matot, Oddo BHF. Please go ahead.
Hello, Thierry. Hello, Marie-Josée. I have three questions. First, do you think we are close to the bottom of the cycle now that the issue of customs tariff has been settled in the U.S.? What your main customers are telling you in that country? It seems you are very cautious about the scenario of recovery next year, considering your decision to reinforce your efforts on cost. Second, can we have an update on PIAM? How is it delivering in the current context? And my last question, your inventory levels at the end of September are stable in value compared to the end of June at EUR 1.3 billion. Does this mean that it will be very difficult to reduce stock in the future if demand does not recover? Thank you.
Okay. Clearly, as we mentioned, we are in a low cycle. Where are we exactly? I think everybody has to be modest on that. Only analysts and all the players in the industry, we have all been wrong. My feeling is that we are really at a low point. It has been also a long period of decline quarter after quarter. I do not know if we are at the bottom, but we get close to the bottom. What we cannot say is when the recovery will start. This is why I think we get the message. We want to be ready for whatever scenario. This is why, at the same time, we really optimize what is linked to cost, cash, etc., but not jeopardizing the ability to rebound when the cycle is coming back.
Your experience has been that each time the light has come back, we are one of the first ones to take advantage of the recovery. We want to stay with this mindset while recognizing the challenge of the current macro and adapting what we have to do, but without losing, I would say, the focus on the long-term development of Arkema. It is a fine balance, but I think this is what we try to do so far. On PIAM, as you know, it depends freely on the end market, particularly advanced electronics of PIAM. The vision per quarter can change because it depends on the stock policy of the customer, etc. Globally, we have done a good year, a good progression for PIAM with a margin around, again, 30% in Q3, which means the resilience of PIAM far above any decline in specialty chemicals.
They were rather stable in Q3 after a very strong growth in Q2. I do not want to talk 2026 for the time being, but with regard to PIAM, I can make just a short comment. This seems to be quite positive on 2026 and from their discussion with the customers. I would say PIAM certainly lagged from their initial business plan, but quite resilient, growing this year, rather positive for next year. On your last question, I think the difficulty first, we are doing a good job on stock. You can see that in the cash generation. The difficulty when you are in the middle of the year, especially at the end of Q3, is that you are still at the point where the sales seasonality is rather stable. You cannot also take the risk of losing sales.
We know that we have opportunities to reduce further our stock until the end of the year, and we will do it, and we know how to do it. Maybe Marie-José, you want to complete?
In fact, Emmanuel, when you look at the ratio of inventory on sales, frankly, we are very much aligned. If you compare with last year, last year, end of the year, we finished with EUR 1,350,000, which represented 14.7% of our sales in terms of level of stock. This year, we are at EUR 1,309,000, as you mentioned, which represents 15% of our 12-month sales. Ultimately, because it is still not year-end, there is a very limited, let's say, excess of stock in the chain compared to last year's year-end level. Definitely, we are adjusting permanently to the forecast, and there is an additional expected reduction for year-end on this metric in particular.
Okay. Thank you very much.
The next question is from James Hooper at Bernstein. Please go ahead.
Good morning, and thank you very much for taking my questions. I have two, please. The first is about the cost savings and the delivery. How does the increase fit into the ambition that you had at the 2023 CMD to deliver EUR 250 million savings over the five years? You have added the incremental EUR 50 million this year. Is it the opportunity has become EUR 50 million bigger, or is this more just pulling forward savings to kind of take advantage of the current situation? The second question is about the kind of special necessity material projects. Initially, when you guided 2025, you had EUR 100 million. They have been downgraded to EUR 60 million. What kind of growth potential contribution are you expecting if we assume a similar macro environment in the coming years?
Kind of put another way, if these projects have contributed EUR 60 million more and EBITDA is down EUR 250 million year-on-year, what actions are you taking for the rest of the business? Can these businesses grow fast enough to offset weakness elsewhere despite your cost savings? Thank you.
Okay. With regard to additional cost savings, yes, clearly, what we are doing is not just to get quicker on the cost savings. It is to add cost savings. This means that if we say, for example, this year, we increase by EUR 50 million, this means EUR 50 million you will find then at the end of the five-year period. This means that we are at EUR 300 million. If next year, to deliver the offset of inflation, we will be again significantly above the EUR 50 million. This above will be also on top of this EUR 300 million we have just talked about, etc. There is no way we will say up to EUR 50 million. It will be at the end far more than EUR 250 million. I think it answers your question. On the project contribution, the growth potential for us is intact. The question is more it depends on the scenario of the macro.
It depends if the macro is coming back already in the course of next year or later, whatever. What is clear is that we believe that the numbers we have shared with you are still the same. The question is more the timeline. This means that if the macro is coming back sooner than later, I think we can certainly still target EUR 400 million in 2028. If we take more time, it will be difficult to deliver the EUR 400 million. I think the best would be to have an update in the course of next year on all these projects. What is very important for me is that this project is from a strategic standpoint, and it comes to a certain extent, it is consistent with the answer I made to the question before. The strategy that is supporting this project is still completely valid.
Even if the world is changing all the time, I think these projects are still completely relevant from what we see of the evolution of the world, and this is the most important thing. Now we can debate on the timing if we can still maintain what we said for the achievement of the EUR 400 million for 2028, or if the macro remains similar, as you mentioned, then by nature, there will be some delay. At the end, the contribution will still remain very significant. The endpoint will be the same, clearly. Now, as we said, I do not know if it is really your question on additional action on the project. You got the answer? No. The work of Arkema is not limited to this project. There are plenty of new businesses which are absolutely not linked to this project.
Clearly, looking at, for example, when we discussed data center, which was not so much in our vision up until recently, is something that we will develop. It will be not linked to this major project. It is something else. We permanently continue to complete our new business development prospect based on the evolution of the world.
The next question is from Jean-Luc Romain, CIC Market Solutions. Please go ahead.
Good morning. Hopefully, you hear me better now.
You have to talk a little bit louder. I do not know because it is very low.
My question relates to five outlets which have 20% growth. When you first talked about those five sectors, it was representing 50% of your sales, and your target is 25%. Where are you now in terms of the weight of those five outlets which are growing faster than the rest of your business?
Okay. If I understood well the question, this 20% growth belongs to the five markets which represent 15% of Arkema, but we do not make 20% on the full 15%. It is an extract of the market we are mentioning at the beginning, batteries, sport, etc. It is a part of the five markets, of the five platforms. This means this 20% applies more to around 7% of the sales. On the rest, the other 8%, we are far more stable. Does it answer your question?
Thank you very much.
We take the last question.
Gentlemen, there are no more questions registered at this time.
Okay. So if no other question, I would like to thank you very much for taking the time to. The rest, and I wish you a good day. Do not hesitate to come back to Béatrice and to James if you have any further questions. Thank you very much again.
Ladies and gentlemen, this concludes this conference call. Arkema thanks you for your participation. You may now disconnect.