Frequentis AG ($FQT)

Good day, ladies and gentlemen, and a warm welcome to today's earnings call of the Frequentis AG following the publication of the financial year figures of 2025. I'm delighted to welcome the management Board of Frequentis with CEO, Norbert Haslacher; CFO, Peter Skerlan, and CTO, Karl Wannenmacher as well as Stefan Marin as Head of Investor Relations, so the gentlemen will guide us through the presentation and the results shortly. Afterwards, we will move over to a Q&A session in which we will be happy to take your questions. And having said this, I hand over to you, Mr. Haslacher.

Yes. Thank you, and also welcome to this conference call. Before we start our presentation on behalf of the Executive Board, I would like to say a big thank you to all 2,600 employees around the globe for their very strong dedication and team spirit and achievements in 2025. This gives us every reason to look back with joy and pride, and the culture of appreciation and personal development is of central importance to us. It creates the conditions for us to grow sustainably and also remain successful from a long-term perspective. So I would like to start the presentation with Slide #2, the highlights of 2025. We are again able to report a double-digit increase in order intake, orders on hand and revenues. So the order intake rose by almost EUR 100 million to EUR 680 million. The orders on hand are now close to EUR 800 million, EUR 795 million to be exact, and revenues grew by 21%. In other words, EUR 100 million to EUR 580 million. EBIT climbed to EUR 47 million resulting in an EBIT margin of 8.1%. But please note that the EBIT margin includes a positive effect from an EUR 8 million claim settlement. And without this, the EBIT margin would have been 6.7%. At the end of December '25, we had a net cash position of EUR 105 million, including EUR 87 million from advanced payments from customers. On the order intake highlights. When we look back, we had really great highlights to report last year. So additional orders in the U.S. contributed to the high revenue growth in the Americas region that we will show you later, the rollout of the air-to-ground protocol converter system in the United States is on track. And in addition, first orders for a modernized voice communication system were generated. The countrywide drone traffic management system ordered by Sweden shows that we have the right applications and solutions to participate in this growing drone market. even if it is only a small. The multi-domain communication systems for the Australian Defense Force's next-generation AIR6500 Joint Air Battle Management System are the type of order that moves us forward, especially in terms of both our product portfolio and our knowledge base. Following completion of an order from the U.S. Defense Department, there is now a digital air traffic control system in place at the U.S. Army Garrison in Germany. And another drone management system this time for testing the integration of drones into military air space for the German Army. In Public Safety, we won sizable contracts from the German Federal State of Thuringia. Here, Frequentis is the general contractor. Multimedia functionality, in other words, managing multiple types of inquiries via phone, radio, video or digital channels is an important element in the LifeX rollout for all fire emergency call centers in Norway. Slide 4 shows that the order intake grew again this time by 17% to a new record of EUR 680 million. Both segments contributed to that success. And Air Traffic Management posted a rise of 11.8% and Public Safety & Transport a rise of 26.6%. So the order intake shows confirmed orders only, as you know, not the budgeted value of the contract. Looking back the past 4 years, we averaged an increase of 20% per year. And looking forward, we have a well-filled order pipeline and also very good opportunities for 2026 and beyond. The orders on hand increased by 9.8% to EUR 795 million compared with December 2024. And the composition of order intake on orders on hand is quite similar, roughly 2/3 for Air Traffic Management and 1/3 for Public Safety & Transport. We now go more into details about financials, and I would like to hand over to Peter Skerlan, our CFO.

Thank you, Norbert. Hello, here is Peter. Now to Slide #5. The revenues rose 20.8% to a new record of EUR 580 million in 2025 and the book-to-bill ratio was 1.17, underlining our growth. Both segments contributed to the growth. ATM grew by 18.5%, while Public Safety & Transport grew by 26.4%. In the past 4 years, we have managed to increase revenues by 15% per year. And this increase fills the need for additional employees on average 8% during past 4 years, bringing our head count to about 2,600. We are able to find enough qualified employees around the world to handle this growth as we are an attractive employer and some industries, notably the automotive industry, are having to downsize the workforce. The revenue split by segment was almost the same. 69% ATM and 31% PST. I'm now on Slide 6. revenues grew in all regions, except for Asia. We saw postponements of tenders for projects and therefore, revenue in Asia. Some of these projects were then awarded to us in first quarter 2026. Europe is still our biggest market. We are happy with the new growth in North and South America. The second highest relative growth rate was in Europe with 50%. The growth rate in the Americas region give this region a 27% share of total revenues compared with 18% in 2024, while Europe's share fell slightly to under 60%, but we are totally happy with the increase of EUR 40 million. Slide #7 shows the figures for revenues, EBITDA and EBIT for past 4 years. The EBIT margin increased to 8.1% and without the positive effect from the EUR 8 million claim settlement Norbert mentioned before, the EBIT margin would have been 6.7%. The settlement occurred in the Public Safety & Transport segment. Even without the settlement in the EBIT margin in PST is above 10%. The slight decline in the EBIT margin at ATM was due to R&D costs and to set up costs for larger projects. As a reminder, we don't capitalize research and development costs, they are recognized in the profit and loss statement. Now let us return to the numbers for the group in 2025. The impact on EBIT mainly came from the cost of materials and purchased services, which rose by 38%, outpacing revenue growth of 21%. This was largely attributable to a more material intensive projects in the Americas region. By contrast, personnel expenses increased only by 12%, a relatively lower increase than the rise in revenues. For the comments on our strategy and outlook, I would like to hand you back to Norbert.

Thanks, Peter. So let's go on Slide #8. It shows the strategy of becoming #1 in control center solutions. So on the right-hand side, you can see that our current product portfolio can address a market of around EUR 3.8 billion. According to our research, the global market for control center solutions, including equipment, is around EUR 14 billion a year. So our strategy is based on 3 main aspects R&D, new deployment models and M&A, to make more and more out of the EUR 14 billion addressable for Frequentis. So R&D results in new solutions and products that cater for the changing technology and user landscape our customers face. So for example, critical communications, such as mission-critical services known as MCX, drone management, that is the integration of and defense against drones and the future communications and tower infrastructure that can run in virtual and cloud environments systems also supported by artificial intelligence. A word about R&D. So expenses for internally funded R&D amounted to EUR 19.2 million in 2025. It should be noted that the proportion of funding provided by customers was higher in 2025 than in previous years. Expenses for internally funded R&D are expected to amount to a normal level of around 6% of revenues in 2026. When we look to the next box on the left, new deployment models, so like cloud solutions and Software-as-a-Service meet changing customer demands and are part of our strategy package. Nevertheless, we have to say that our customers are governmental entities usually have having funding available for CapEx expenditures. Nevertheless, we think that over the time, maybe a more OpEx-based model could be of interest for our customers. Our decision to acquire companies and interests in businesses that enhance our product portfolio has proven right. So take the emergency services program in the U.K. The contribution made by our colleagues at Nemergent where we acquired a stake in 2020 has been vital in enabling us to partner with IBM. So the proactive search for interesting M&A opportunities remains part of our strategy. When making acquisitions, we focus on parameters such as expansion of the product portfolio, a profitable business model, cultural fit and the acquisition price, of course. So we will continue to look at which technologies and products we develop ourselves and which we buy in. In an increasingly unstable world, systems that function reliably even in challenging conditions are becoming ever more important. So Frequentis is a global leader in this field. We are not passengers. Our place is in the cockpit. So our course is very clear, open standards and true interoperability, so that emergency services can work together across systems and borders. And there are currently 3 major focus areas with potential for growth in addition to our current business, which is mission-critical services called MCX, drones and remote digital tower. And I would like to go a little bit more detail into the area of MCX. So let us start with MCX, which represents a quantum leap for security critical communications used by merchant services, such as the police, fire brigades, ambulance services, but also railways and other customers. Technologies, you are probably familiar with, such as TETRA or GSM-R in the rail industry do not longer meet current requirements and become obsolete. So MCX operates on existing 5G-based networks enabling the real-time transmission of voice, mass data and video across national and organizational boundaries using a single standard. Frequentis is already involved in large-scale implementation such as the MCX rollout in the U.K. for around 300,000 emergency service personnel. On Slide #10, the growing volume of drone traffic is also changing air space and placing new demands on safety and on coordination. And it is estimated that the drone ecosystem in Europe only will be worth more than EUR 10 billion by 2030. For civil and military aviation customers, emergency services and railways, we develop and supply solutions for the safe monitoring and management of drones, combining traditional radar systems with intelligent sensor and detection technologies. And by integrating safety critical data and linking drone traffic management, air traffic control and operations management systems, Frequentis is strengthening its position in the management of lower airspace and the protection of critical infrastructure. I think you're aware also of our remote digital towers. So our remote digital tower enables airports to remotely monitor takeoffs, landings and aircraft movements using cameras and sensors for both civil and military users. So airport operators benefit from the centralized management of multiple airports and from the relief this provides in the event of staff shortages among air traffic controllers in remote areas. This technology has already been in use for several years in Germany, England and in Brazil, and we see good growth prospects for Frequentis because there are only a few other companies that offer remote digital towers. And our top regions for remote digital towers are Europe, the United States of Americas, Asia and Australia. So our local subsidiaries in these regions are staffed to manage tenders and the rollout of remote towers. Regarding the United States, our remote digital tower system is the only one currently being tested for U.S.-wide certification at the tech center of the U.S. Federal Aviation Administration. And we still expect to obtain this certification in Q2 '26. Nothing has changed there. Given our annual volume of over 1,000 projects, questions about our capacity to handle multiple requests at the same time are common. So in practice, however demand is well distributed, customers initiate tenders when their current solutions reach the end of their service life. And these time lines vary across customers. As a result, there is no multiyear surge in demand for individual products such as remote digital towers. So let me conclude with the outlook for 2026 on Slide #12. Based on orders on hand of EUR 795 million, we are working at a good level of capacity utilization. We aim to increase order intake further and the sales pipeline for '26 and beyond remains well filled. Revenues are set to increase by about 10%. And profitability, inflation and now above all, shortages, long delivery times and limited commitments from the suppliers on the IT hardware market will weigh on EBIT in 2026. Temporary shifts in milestones, revenues and potential start-up costs at the beginning of projects are common challenges for a project-driven companies like us. Overall, we expect an EBIT margin of about 7% in '26, which is above the 6.7% margin we achieved in '25 if we exclude the claim settlement of about EUR 8 million. Capital expenditure, mainly for notebooks, office equipment and production machinery, will be around EUR 15 million and company funded that is self-financed R&D expenses will amount as mentioned already, to around 6% of revenues in 2026. So to sum up, we are proud of what we have achieved as a team in '25, and we are really energized for another year of growth, 2026. We are now ready for your questions.

[Operator Instructions] And then we will move on with Elias New.

Yes. I hope you can hear me well. I have 2 questions. I'd like to take them one at a time. So firstly, starting with the '26 revenue outlook. You specified your target for around 10% of growth. So could you perhaps just give us some additional color on the underlying assumptions here? And why we shouldn't expect a similar growth rate as we saw in 2025? And perhaps if you could also just shed some light on why 2025 was so strong in terms of revenues, that would be great.

Yes. Maybe I will answer your question about the '26 outlook, and Peter will then give you feedback around the 2025 revenue explosion. So '26, we have a very conservative view on '26 as we see a lot of tensions currently coming up in the Middle East or already ongoing in the Middle East and also on the delivery time of hardware, especially standard IT hardware equipment, we currently do not get any commitments on prices neither on delivery time from HP and Dell. That's why we fear that maybe our execution path in transferring orders into revenues will probably be challenged by the delivery time of IT standard equipment. Karl can give you -- maybe give you a little bit more insight about the problems we have there due to the key AI boom, artificial intelligence boom. So in orders, we see very good IDIQ contracts in the U.S. Nevertheless, I think the U.S. is currently not the most stable country in the world when it comes to decisions. Therefore, we are more on the conservative side. We are -- have a good run under our IDIQ program execution, and we get orders in from the U.S. continuously. Nevertheless, we have midterm elections in -- very soon. And the question is what's the outcome there? And is that influencing the continuation of our programs, which have been set up during the new administrative time of the U.S. administration? That's why we are a little bit conservative about 2026. When it comes to revenue jump in '25, Peter, do you want to add something?

Yes. Thank you for the question. So as I tried to show you on Slide #6, you can see that large portion of the revenue increase comes from the Americas as well as a substantially increase from Europe, which could more than compensate the decrease in Asia. So I think revenue comes from our projects from larger and midsized programs. where we have the possibility to do a little bit more in the Americas.

And Karl, maybe you can say something about our current challenges on the standardized IT hardware equipment side?

Yes. So the challenge here is that the big memory manufacturers are currently focusing very much on providing high bandwidth memory chips where they can make higher margins and sell them to the big AI data center providers. And this leads to a sharp increase in prices for the traditional random access memory chips that are used in our COGS IT hardware servers. So we see sharp increase in prices of our standard manufacturers and we decided to buy on stock those servers so that we are able to fulfill our contract obligations. But this will remain a challenge throughout the entire year and probably also into 2027.

I understand it's not only a price increase. It's also a delivery time increase.

Right.

Daniel, did we answer your questions? Okay.

Yes, that's very helpful. And second question would be on the margin outlook. And I guess, what you mentioned in terms of memory pricing sort of feeding to that. But I was just sort of wondering around the sort of moving part within the 7% guidance for next year. So if we look at the clean EBIT margin for this year of around 6.7%. I was just wondering where this progression comes from? Is it sort of more the ATM segment with software transition or higher defense share? I was just sort of wondering how you think about that.

May I ask you mean the -- you mean the increase from the 6.7% to around 7%. That's the equation where does that come from? Is it right?

Yes, exactly. Just I guess, parting to that, the underlying margin dynamics by segment ATM versus PST. Just sort of how we should think about that trajectory as well in terms of not just '26 margins, but in terms of your midterm targets of 10%?

Shall be come from the ATM section. .

Yes. ATM and PST margin dynamics. I guess most of the uplift...

Yes, uplift don't come from ATM. I think we are quite satisfied with the PST margin.

Okay. So that entire uplift also toward the 7% will come from ATM and the software rollout? Or what are the dynamics there? Because you also mentioned higher R&D costs in '25. So just wondering if you can shed some light on that. And should we then expect a kind of linear improvement each year as the software rollout gains traction in '27 and beyond?

Yes, yes. When you look on the history, then you see that in the past, we had also the possibility to pay up a little bit more margin in ATM and due to starting costs from programs. That's the reason why our new products, the 2 reasons why the margin declined. We want to gain -- come back to better margins and improve them as well as when the transition to the new product is then fulfilled, which shall be done within the next 2, 3 years. the transition shall be done then to more software-based business. .

Okay. Great. That's very clear. And just on the PST margin as well in terms of '25, I mean even if we strip out the one-off gain, it looks like it's quite a step up as well compared to the '24 margin. Just wondering is there anything to mention there? Or is it just a product specific -- sorry, project specific, and that's the main driver.

Of course, it's fluctuating from 1 year to the next year, but I think we have reached here a quite acceptable level.

The next one would be Daniel Lion.

I would like to follow up a little bit on the margin discussion. Can you confirm that you see basically everything on track towards your midterm margin target level of around 10% at the end of the decade, is this still valid?

Here's Peter again. There are a lot of dependencies here, but still, we try to follow this plan.

What do you see as a major risk in meeting this midterm projection?

I think on the last slide, Norbert mentioned some major threats for our profitability. So we book good phase without the crisis, a worldwide crisis would be quite nice for the margin. I think that if you make up business worldwide, portfolio problem worldwide. Next thing is to succeed in finishing the transition, concerned finishing our new products, especially in the ATM business. And the next thing is that large projects when we fulfill the larger projects that there are no postponements but also the customer is able to finish his tasks so that we can deliver in time. And these are the main things. And I think all the restated risk, I think we have declared them in our financial statements over several pages, political risks and inflation. Inflation in Europe is a tough thing for us. As you know, we had inflation rates of at 7%, 8%, 9% here in Austria as well as in the other subsidiary countries, Romania. That's tough for us in comparison to the international competition. These are the things that could hurt us in the future.

Okay. Understood. Then one more on the claims settlement. You mentioned the EUR 8.5 million one-off. Could you provide some more insight on the settlement, just to get a feeling of to what extent we should really treat this as a one-off?

Yes. It would be nice to have this each year, to be honest, yes. So that would be very nice, but it's a situation that I don't want to have each year. It's a large program where the customer decided to stop it because the prime contractor did not perform and the customer decided to stop the program. Frequentis performed, but we had a contract that more or less if the prime does not perform, we have a problem. And so we had to fight it and it took several years to fight it through. So we have done all our work, but we were not paid by the prime contractor. And that's the reason of this effect in 2025 because all the costs were in past periods but we had no possibility to capitalize it or do something with it because we had this fight there. And the question was, will we win? Or will it last for years? Because, as you know, some current decisions take a lot of years.

Okay. I understand. Yes. And then maybe a last one. Working capital sales, it's flattish for the first time since years on a year-on-year basis. Do you see this level now as sustainable going forward, especially now reflecting on the shortages that we see on the memory side? But there might be other shortages coming up as well? Or should we prepare for this ratio maybe to even increase higher?

Thanks for the question. It's a very good question. So we have to make decisions each day. So shall we put something on stock so that we are able to deliver or shall we risk that we cannot deliver, shall we take the contract because the payment milestones are ugly, but it's a tender, and we have no possibility to improve the payment milestones. So we will receive the payment at the end of the contract. So it's decisions that we have to do each day. But our overall target, especially my target is to keep it stable and even to improve it because we had lower rates in the past. And due to the growth of the company, we have to keep the working capital under control. Otherwise, it will cost a lot of funds to prefinance the working capital.

And then we move on with the question from Ms. Becker. So Mr. Becker poses question in the chat and he would like to know, given the continued search for acquisition targets, have potential targets already been identified in 2026 and beyond.

Yes. So I will take that question. We are now a stock listed company since May 2019. So that's around 7 years. And within these 7 years, we have acquired 10 companies and I think have learned what it means to acquire a company and then to focus on a post-merger integration phase. We are in some larger acquisition areas still working on that post-merger integration, which is, of course, a good learning for us. So yes, we are hungry to and keen to do further acquisitions. Are there targets out we are interested in? Yes, there are. So we have around 30 to 35 teasers a year. where we -- not in all of them, we go into details. But in some of them, we are interested to learn more about their business model, about their profitability level, about their pricing level. about their culture and so on. So yes, M&A stays a major cornerstone of our strategy. Will we acquire something in '26? I cannot tell you now. But M&A is one of our prime activities when it comes to growth of the Frequentis Group.

And then we move on with the questions from Mr. Zuzak.

Yes, hello, can you hear me?

Yes, Miro, we can hear you.

I have a couple of them. And if it's okay for you, I'll take them one by one. The first one is for Peter. The EUR 8 million settlement one-off payment, in which P&L line was it booked?

Revenues.

In the revenues.

Yes. We have the possibility to bill that is one sum. And part of it was already paid 2025 and the other half in 2026, but it's in revenues. So if you can deduct it, if you want to know, without this onetime effect, you can deduct the amount from revenues, and then it's also deducted from the EBIT.

Okay. Very clear. And it's in PST, as you said that. Okay.

Yes.

Then next question. I've seen that you released quite a large amount of provisions in H2 -- sorry, you didn't release, you built provisions, EUR 17 million in bonuses, Congratulations, gentlemen. And -- but the question is, are these EUR 17 million of bonus provision already included in the personnel cost in H2, the EUR 148 million? It's also a question for Peter probably.

Yes.

They're already in there?

Yes.

So then I have a question because if I deduct EUR 17 million from EUR 148 million, which is the H2 personnel cost, then I get like EUR 131 million of pre-bonus personnel cost, which is in line with the number you had in 2024 already, roughly EUR 130 million per semester. And now the question is why was H1 already basically at EUR 144 million, which is quite high? So despite the EUR 17 million personnel cost bonus provision booked in EUR 148 million, we have EUR 144 million for H2.

Yes. I'm not sure if I understood your question, but what we have to do is we -- for the half year, for the first half year, we do also a provision for bonus if we think that by the end of the year, the people will meet the targets, and if we have a good idea how to convince our auditors that I don't have to book these provisions when the half year result shows a very low result, then I would appreciate that. But the opinion is the years if we think by the half year that we will achieve the target, I have to make a provision of half of the bonus. The question is if the result is then better, the provision at the end of the year has to be then a little bit more here. But I'm willing to pay more bonus if the results are better.

Yes. Okay. Very clear. So you already booked some of the provisions in H2.

Unfortunately, I have to.

No, it's fine. Next question, what are your hiring needs in the next 2 years without acquisitions? So if you just want to grow the 10% around, which is in fact more because you had the EUR 8 million one-off, right? So it's maybe 11% or 12%. Do you also need to scale employees and what by what extent?

I had a slight noise. So did you ask for how many people do we plan to hire in the future? Was that your question?

Yes.

Okay. Okay. There is a special effect. When you look upon the ratio, personnel cost, in relation to turnover, you will see the personnel cost is declining from where we are almost up to 54% of revenues was personnel costs and now we declined to 50-something percent, yes. The decrease was because we have a little bit more material and more procured services. And I think if we need more people and how many of new employees we need depends on the structure of the tenders that we win. I don't want to sell IT hardware. I'm not interested in selling IT hardware. But if the customer attends these things, we have to offer it to procure it and then a portion of the turnover is created by delivering this IT hardware, so you will see more material costs. What will I expect in the future? I would expect in the future that probably we really transform them in a software company by not having -- not being forced by the customer to deliver hardware. So that means that we provide more labor to the customer, more software. That means the relation personnel cost to turnover will increase in the future.

So you can't give a number, like 200, it was always like 200 last couple of years.

If you want to have a percentage, yes, because when you think that will be the turnover, then from the 50%, we probably will go back to 52%, 53% of revenues. Yes, I think it's only better than people because people, the question is where do we hire the people and how much personal payroll do you then calculate. So if you take this ratio, it's probably easier for you.

Okay. Next question is basically related to what you just said. So I mean, you've seen very low personnel costs in percentage of sales. You also have seen quite a low gross profit or high cost of material related to sales, which led to a lower gross profit margin, especially in H2, but also for the full year. And we also see a lower EBIT margin in ATM. Can you explain better how much of this low-margin hardware straight through selling was included in your top line? There must certainly be such an element in your P&L from one of the numbers I see.

That's hard to say because the project has hardware in it or not. So I think in this case, I can't give you a figure that could be useful for you. And as I mentioned, concerning the material costs, we hope that in the future that's declining because especially in IT hardware, the margin is lower.

Okay. And different question regarding margin in ATM. I think the ambition in the beginning was to have clearly an increasing margin there. Right now, it was a decline in ATM, and I guess this is related to a specific like project from the U.S. That's my guess. You don't have to comment on that. But will these profits, have they disappeared or have they just been moved into 2026?

A very good question. Unfortunately, a large portion of it was used up in 2025. And to regain it in the future, the plan then on change requests, if we can sell additional stuff with probably -- additional things with probably higher margin here. So as I mentioned before, the increase for 2026 that we have shown already when we went ad hoc in -- was it in January, February? In February, when we went in ad hoc in February will mainly come from ATM.

I'm not sure but did you -- when you answered it did you talk about cash flow, the usage of the like the prepayments or were you...

Not much in EBIT margin. We're talking about EBIT margin because what we elaborated, I don't know, 10 minutes before was the increase from the 6.7% margin when we take off this onetime effect to around 7% EBIT.

But can you better explain -- because I was expecting positive operating leverage in the ATM division. Also given the fact you note that was the not capitalized R&D cost should increase to an increase in the margin in the ATM division. I expected that the ATM division would actually increase the margin year-over-year. Now it was lower. Can you better explain what was happening there? And how this is going to develop into 2026? Do we have to expect that with continued growth in the ATM division that the margin is going to go down again? Or how does that work? Maybe can you explain better what's happening there in this division, like a problem, a specific problem happening, something which you didn't expect or so or...

As I mentioned before, it's the transition takes further 2 to probably 3 years, the transition to these new products. That's the point.

And during the this transition, the margin will decline or just no decrease...

No. As I mentioned, overall in ATM, we think the improvement for 2026 from 6.7% to 7% or around 7% will come from ATM. So in other words, a slightly recovery shall be seen in 2026.

Okay. And then last question, then I go back into the line. Actually 2 last questions, sorry. One is the minorities went up significantly to EUR 5.3 million from EUR 1.5 million and which was EUR 1.6 million the year before. So it was clearly much higher, probably team communication or so I don't know, maybe you can explain what that is and in which -- mainly also in which segment these profits were booked.

Yes. The thing is if you don't owe 100% of a company, there is another owner. And if the other owner has half of the company, half of the profit is then the profit of the other owner. And in this special year, we had the nice possibility to gain profit in these companies where we had not 100%.

No, I know what minorities are. So was it communication technology or was it Elara?

Elara.

And is this in ATM or in Public Safety?

Public Safety & Transport.

That's there in there. And this is a 50% -- 49%.

51%, mainly from this company. But I have to say that also team with team, we are doing a quite good business. So overall, we are happy if all these companies doing good business because even when the other half belongs to another person, half of it belongs to Frequentis.

Okay. Okay. So the entire increase in Public Safety & Transport if you adjust for the EUR 8 million, you went from EUR 14 million to EUR 19 million, the entire increase was basically coming from Elara, I can infer from the increase of the minorities. There was no underlying increase from other than Elara company is quite the opposite.

Yes, you're right.

And Elara, is there a one-off element or Elara is also...

You are in a good track. Congratulations. You are in a very good track and you combine it quite well. Yes, you're right. The onetime effect happens there.

The EUR 8 million was in Elara? No, that's much better.

Congratulations to your combination, Mr. Sherlock, marvelous.

And then the last one, sorry, would be on the EUR 80 million prepayment from the FAA. I wonder why your net working capital change was still negative in 2025. Is it already used up entirely the EUR 80 million of the prepayment of the FAA?

I think now I've missed the point. May I ask you to repeat it once more. I think I've missed a point.

I think there was a large prepayment of around EUR 80 million coming from the FAA in the beginning of the year. And despite this large prepayment, the change in net working capital was negative in sum.

I don't know where the EUR 80 million comes from.

From my notes, I'm not sure where they're coming from, actually, but I have it in the notes. I hope they're not wrong.

I don't know the EUR 80 million. I had appreciated this sum. It wasn't -- I think probably we expected something like that 1 year ago, that's possible because sometimes customer promise a lot. And then the question is how much do they keep. I think there were enormous sums in discussions. But then when it comes to the contract and to the real payment, sometimes numbers are smaller. . But the prepayment was then used to procure materials. And what you see is in working capital. In the working capital, you will see a little bit still concerning that you still increase of materials because the materials were procured, were on stock. So you see a large increase in the inventories. You see also an increase of our trade receivables and also the trade advances. In total, FAA has an effect of the working capital of EUR 20 million, yes? But it affected every position in the working capital. But overall, EUR 20 million.

[Operator Instructions] So Elias, the stage is yours again.

A couple of follow-up questions from my side. So firstly, on your defense business, I was wondering if you could comment on defense revenue share for 2025? And also what the current backlog share of the defense businesses? And perhaps also your expectations for 2026 and beyond, when do you expect orders to pick up more meaningfully? And also in terms of profitability, where do margins currently sit and what you sort of see going forward?

Okay. I mean for defense revenue share in 2025, we have been around 18%. The reason for that is that we had a very strong growing Public Safety & Transport business. In absolute numbers, defense was growing. The order backlog is pretty good for defense. So we expect a higher order intake value of '26 compared to '25 when, hopefully, soon the NATO spending starts in Europe out of this special funding mechanisms in Germany and other NATO countries. We expect that '26, '27, '28, we will see much more tenders following the procurement of fighter aircraft, tanks and other hardware equipment. Margin profile, I think we have reported that last year as well, it's the same this year. Margin profile in defense is much better than the margin profile in civil. And to be honest, in some projects, it's not easy to distinguish between civil and defense and it is a joint infrastructure like in Brazil, where the network is managed by military organization but is also used by civil airspace users.

Okay. Great. So margins still sits somewhere around 10% or slightly double-digit profile.

Yes.

Okay. Great. And in terms of the drone business, I know you mentioned that in your presentation anyway, but could you just shed some light on expectations there in terms of revenue growth contribution going forward and also current revenue share margins? And anything you could share here would be very helpful as well.

Yes. We don't have these figures as we have incorporated drones MCX recording and our C2 development agency into one organizational unit called Acceleration Hub. There, we share resources. And the purpose of that company is to have a much bigger and faster scale than in our standard business model. That's why we have combined them. Overall, we do around EUR 25 million to EUR 30 million order intake already in coming out of that Acceleration Hub. What I can report concerning the drones is that we see currently more activity in the defense space as we have published the research program for one of military airport in Germany where we test the friend foe distinguish based on our technology and based on sensors being available there and the vectors being available there. So we are the data collector and data process in the back us collecting sensor data, processing the data together with the ANSP and then come to a decision if it's a friend or a foe. And if it's a foe, we give data to the vectors. That's our positioning. In the civil area, yes, we have won Sweden. But to be honest, it's a very, very slow developing market in Europe as the regulation is still missing who is responsible for the lower airspace. I think the regulators and the European Commission would love to have competition to the ANSPs. Therefore, they hoped that somebody will stand up and say, "I will be an ANSP for the lower air space, managing drone traffic." Unfortunately, that didn't happen because nobody was willing to invest millions of dollars into infrastructure to manage maybe upcoming drone traffic. Therefore, we hope that within the next hopefully, 24 months, European Commission will come to a conclusion that the only way out here is to give the ANSP responsibility for also managing the lower air space and therefore, the drone traffic.

That's helpful. And final question on R&D. I mean you're guiding for around 6% of revenue for 2026. Is that a sort of reasonable run rate to assume for the outer years beyond '26 as well?

Karl, it's yours.

Yes. No, that the goal clearly is to reduce that rate. As Peter has mentioned before, for the next 2 years, we expect to stay on this 6% rate. And then slowly moving towards the 5% rate. I mean Frequentis will always be a very R&D-heavy company and for a good reason. But the clear ambition is here to move towards the 5%, it will take some years.

So ladies and gentlemen, in view of the time, we only have the opportunity to take questions from Wolfgang Specht.

Okay. So once again, on 1 of your 3 strategic growth areas, drone management. If I look at the military part of the business that's hopefully upcoming here, do we may need some more, let's say, solutions for combating drones before you really see a spike in orders here because detection and competing should be one integrated solution for the military part. So is it maybe a little bit too early to really hope for big orders from the military side.

Our reflection goes more into another direction. So our positioning is very clear. We want to stay agnostic when it comes to sensor equipment and effector equipment, because what we see is that sovereign states try to support their local industry. So you find hundreds of companies developing sensors and you find hundreds of companies developing effectors. So our positioning is we are an expert in managing and controlling air space. That's where we have a lot of domain knowledge and solutions available. Part of them are in our UTM solution, where we can process sensor data, process and align with ANSP and regulation and give it back to effectors, if necessary, to shoot or bring a drone down. That's our positioning. And we see that military organizations are interested in that because you shouldn't forget, we have peace time in Europe. Hopefully, we have peace time for a very long time. And in peace time, you cannot just start shooting around if you see a drone, you have to follow a process. and the procedure. And this has to be aligned with the navigation service provider. Therefore, we think our positioning to be the channel to the ANSP and be the data processor of sensor data and effective data is the right positioning for us. Do we see a major development in the area of tenders? Currently not. It's more a research program. Partially, we see a tender like Sweden coming up and partially then another tender is now coming up in the Middle East. But we think the next couple of years, it will be more a military tender traffic than a civil one as long as the regulation is not clear for Europe.

So ladies and gentlemen, thank you, everyone, for your participation and your interest in Frequentis. So a big thank you also to the gentleman from the management Board for your presentation and the time you took today. So it was a pleasure to be a host today. I wish you all a lovely remaining day and hand over again to Stefan Marin for some final remarks.

Yes, from my side as well. Yes, I'm looking forward to meet you at the upcoming conferences next week with a ODDO BHF in Turin or on the 6th of May with Berenberg in London and our half year results for 2026 will be published on the 11th of August. If you have any further questions, please just drop me an email at investor.frequentis.com. Until then, all the best, and thank you.