Good afternoon. A very warm welcome to our yerly analyst and investors conference of 1&1 AG in Frankfurt. It's my pleasure to welcome you here today. Making a long story short, Mr. Dommermuth, CEO of 1&1 AG, and Sascha D'Avis, CFO of 1&1 AG, will guide you through our business year 2024 and the outlook 2025. Afterwards, we'll have time for questions and answers. Thank you very much, Mr. Dommermuth. You have the floor. Thank you very much, Mr. Karl, ladies and gentlemen. As it was said, I have the pleasure to present to you the financial year of 2024 and the current situation of our network. Mr. D'Avis will talk about the outlook. You are familiar with our business: broadband business and mobile telecommunications, 3.95 million broadband connections, VDSL and FTTH, complete packages including voice and IPTV.
We purchase via 1&1 Versatel, our sister company with a federal network, and transfer points to carriers and municipal networks for the last mile, not via Versatel, but via Deutsche Telekom and city carriers. For mobile telecommunications, 12.44 million mobile contracts in the past year. We are first Open RAN in Europe, fully virtualized. We have a broad market coverage with our main brand, 1&1. We have our co-brands, GMX and WEB.DE, and the discount brands that you can see here: yourfone and others. Our customer contracts of the past year: 130,000 more to 16.39 million, 12.44 million coming from mobile internet, 3.95 million broadband lines. For mobile internet, we would have liked to see more growth. At the beginning of the year, we were at 250,000 new contracts. However, last year, with the migration of our customers to a new network, we saw some changes.
We had some significant outages, and that cost us performance. For our development, as in the previous years, a slight decrease, and we would have liked to have a different situation. However, the earnings are on the right track, as you will see later. That is why we decided not to be too aggressive and rather make sure that the business works well. Our revenue of this past year is slightly down, EUR 4.06 billion. Our high-margin service revenue is + 1.8% to EUR 3.303 billion. Our low-margin other revenue with smartphones, notebooks, tablets, also a decrease. Now, hardware business is a low-margin business for us. We calculate the price with the contract term. If the hardware cost is maybe EUR 1,200, then the monthly payment has a plus of EUR 50 for 24 months. We get the money back.
The revenue always depends on how attractive a device is or what kind of incentives we offer. The better the deal, the more hardware we sell, of course. Our operative performance for the segment Access is, as I said, it's working well. We have an EBITDA of EUR 856.1 million. This is our traditional business. We have sales, we have customer service, operations, and so on. We have our upfront costs, and we purchase the services the same way we would in case we were to get them at Telefónica or others. Our second segment is the 1&1 Mobile Network with EUR 265.3 million EBITDA increased startup costs. You will see that it's going to be similar for this year. In the following years, you will see that the losses are not as high.
The goal, of course, is to, in the future, be able to produce at a lower price than having to purchase at a higher price for each gigabyte and to then make sure that this segment can also contribute to our profit. This will take some time. Our CapEx we have EUR 290.6 million in this past year, then EUR 21.7 million for the segment Access, EUR 268.9 million in the segment for the 1&1 Mobile Network, hardware and software. Here you have an overview of the numbers: 16.3 million new contracts. We have service revenue of EUR 3.2 billion, then the EBITDA of the Access segment, EUR 786.2 million, 1&1 Mobile Network EUR -132.4 million, and the EBIT, EUR 455.8 million. You can see here the effects of depreciation of investments in hardware and software, which will change over time, EUR 0.05 per share, the minimum dividend.
Now, let's quickly talk about the current situation of our Mobile Network and our telecommunications architecture. We have a completely virtual network that we are building. Today, we have four core data centers. They are redundant. We have 24 Edge data centres around the core data centers. Around those, we have the Far Edge data centers. The Far Edge data centres, the antennas, they have the antennas do not have their own intelligence, so there is no storage in some basement. The Far Edge data centres produce the IP traffic, transform it to radio signal, and from there, it continues. More than 1,000 active masts are connected to the far Edge data centers, so we have more than 1,000 active hubs, and more than 5,000 more are in development. Now, what does that mean? We have a contract for the use of a specific location.
At this location right now, the antenna is being built, or the mast is being reinforced, or we have an in-house cabling process going on at the moment. If we do not have a location provided by a tower crew, then we may use rooftops and look into statics first, of course. We will then build the masts, we will build the antennas, and include the cabling, fiber optic, and so on. 5,000 locations, as I said, right now are in development. We are seeing progress here with the goal of the 1,000 active masts. That was our goal for 2022. It took a bit longer. However, from quarter to quarter, we are having better numbers, and now we have a critical mass compared to Telefónica and others in monthly development.
You can say that the business is running, so to speak, and we're finding our place. It's easier to plan. The next 5,000 are in development right now. As I said before, we are building the first Open RAN in Europe. What is the main difference? We have open interfaces. With that, we are independent. We can decide ourselves what partners should provide us with hardware or software. We have our own network that's built in a way that we always have fiber optic antennas, and we are prepared for real-time applications because we have the necessary capacities in the data centers so that we can have a quick and smooth transfer. We have a low electricity consumption compared to traditional networks. 3,200 employees at the different sites.
We have a very broad ecosystem with more than 100 partners who help us build and operate the network, 50% coming from Germany, 40% coming from the rest of Europe, and 10% from the rest of the world. With the exception of China, we do not have partners coming from China. So far, so good. That is 2024 from the first perspective. I will hand over to my colleague, Mr. D'Avis.
All right, thank you very much, Mr. Dommermuth. I would like to wish you a good afternoon as well. I would now like to proceed to sharing the financials with you for the current financial year, which is 2024, and also give you an outlook to our expectations for 2025. Let me start with earnings. Revenue in 2024 was EUR 4.0643 billion compared with EUR 4.0967 billion in 2023.
Therefore, the high-margin service sales dropped to EUR 3.303 billion in 2024, while low-margin hardware revenue, particularly from smartphones, fell by -10.8% to EUR 0.76 billion. The cost of sales rose from EUR 2.938 billion in 2023 to EUR 3.22 billion in 2024. The gross profit fell accordingly from EUR 1.159 billion in 2023 to EUR 1.042 billion in 2024, which is a decrease of 10.1%. The gross profit in the Access operating segment increased from EUR 1.326 billion in 2023 to EUR 1.401 billion in 2024, which corresponds to a strong increase of 5.7%. The gross profit in the 1&1 Mobile Network segment, 1&1, amounted to EUR -359 million in 2024 compared to EUR -166.8 million in 2023. This included depreciation and rising expenses for the construction operation and the 5G frequencies in the year 2024, amounting to EUR 129.1 million after EUR 47.3 million in 2023.
Distribution costs rose from EUR 513.2 million in 2023 to EUR 535.7 million in 2024. This corresponds with an increase of 4.4%. This is mainly due to higher commission costs in previous years, which are deferred over several years. Administrative costs totaled at EUR 112.2 million in 2024 and were slightly below the previous year's figures of EUR 115.6 million. Other operating income and expenses was at EUR 37.0 million, and the balance was above the 2023 level of EUR 31 million. The change is due, in particular, to improved income from the recovery and collection processes. Impairment losses on receivables and contract assets increased from EUR -105.4 million in 2023 to EUR -121.9 million in 2024, mainly due to higher payment defaults. We will recover part of this through the higher income, as I said, from the recovery and collection process.
The EBIT was at EUR 309.4 million in 2024, which was below the 2023 result of EUR 455.8 million, which is attributable to the higher expenses for the 1&1 Mobile Network. At EUR -4.2 million, the financial result in 2024 was below the previous year's figures of EUR +9.1 million. The increase in financing expenses was primarily due to the interest expense from the recognition of leases in connection with our network expansion. The decrease in interest income was mainly due to a lower average cash investment at United Internet. Earnings before taxes thus amounted to EUR 305.2 million in 2024 compared to EUR 464.9 million in 2023. As a result, tax expenses fell to EUR 92.4 million in 2024. In 2023, tax expenses still amounted to EUR 149.9 million. We will therefore achieve a consolidated group result of EUR 212.8 million in 2024 compared to EUR 315 million in the previous year.
I would like to move on to the balance sheet. Balance sheet total rose from EUR 7.74 billion at the end of 2023 to EUR 8.13 billion at the end of 2024. This increase of EUR 389.3 million is mainly due to the following effects. Current assets amounted to EUR 1.844 billion and were therefore EUR 83.7 million lower than the previous year's figures, which was totaling at EUR 1.928 billion. On the one hand, this is due to the EUR 107 million decrease in receivables from affiliated companies. The decrease is primarily due to the reduction in cash investments at United Internet by EUR 91 million versus 2023. On the other hand, inventories fell from EUR 176 million in the previous year to EUR 119.6 million in 2024 at the reporting date as a result of the scheduled reduction in inventories.
In contrast, deferred taxes increased by EUR 69.5 million, which is primarily due to the advance payments to Deutsche Telekom as part of the contingent contract. The non-current assets increased by EUR 473 million from EUR 5.13 billion in 2023 to EUR 6.286 billion in 2024. This is mainly due to the increase in property, plants, and equipment as a result of investments in the 1&1 Mobile Network. Current liabilities rose from EUR 716.6 million in 2023 to EUR 730.6 million in 2024. This is due to an increase in trade payables as a result of delayed invoices from upstream service providers in the network expansion. Non-current liabilities also increased from EUR 1.137 billion to EUR 1.306 billion in 2024. This increase is due to higher lease liabilities of EUR 222.2 million as a result of the growing number of antenna sites in the 1&1 Mobile Network segment.
This was offset by the annual repayment of the liability from the frequency increase. Equity rose from EUR 5.887 billion to EUR 6.094 billion due to the positive group result we have shown. Moving swiftly on to cash flow. Net cash flows from operating activities amounted to EUR 311.4 million in 2024 compared with EUR 225.6 million in the previous year. These were made up of EUR +494.3 million cash flow from operating activities, increased tax payments of EUR -99.9 million, which mainly relate to back payments of income taxes for previous years and excessive prepayments for the current income taxes in 2024. These excessively high advance payments will be refunded as part of the general assessment and result in cash inflows.
Reduction in inventories of EUR 58.4 million, the change in contract assets of EUR -64.7 million as a result of the reduction in hardware sales, and from the change in deferred expenses in the amount of EUR -223.5 million due to the advance payments under the quota agreement with Deutsche Telekom. Last but not least, obviously, from the change in working capital amounting to EUR 17.4+ million . Cash flow from investment activities amounted to EUR -180.8 million in 2024 compared to EUR -125.2 million in 2023. This can be broken down as follows. A cash CapEx in 2024 amounting to EUR -290.6 million, mainly from investments in the 1&1 Mobile Network, EUR 91 million from the investments of free cash and cash equivalents at United Internet, and EUR +19.1 million in interest payments received from the investment with United Internet.
Cash flow from financing activities amounted to EUR 129.7 million in 2024 compared with EUR 101.8 million in the previous year. This is broken down as follows. EUR 14.1 million is repayment of lease liabilities as part of the accounting of the rental agreements for our antenna sites, EUR -8.8 million from dividend payments, EUR -61.3 million repayment of liabilities for the 5G radio spectrum, EUR -31.3 million other payments with interest character as in previous years. These relate to payments in connection with the [1 Sports] project and EUR -14.1 million interest payments from leases. As a result, in 2024, we achieved a free cash flow of EUR +20.8 million, a significant increase compared to the EUR -70.1 million in 2023. I would now like to discuss the bridge from EBITDA to free cash flow.
We come from an EBITDA in the amount of EUR 590.8 million, trade receivables in the amount of EUR -6.8 million, inventories amounting to EUR 64.6 million, contract assets amounting to EUR 64.7 million, deferred expenses of EUR -223.5 million, trade payables in the amount of EUR 9.7 million, other working capital to the tune of EUR 11.7 million, tax expenses in the amount of EUR -199.3 million, and then the cash CapEx amounting to a total of EUR 290.6 million, which brings us to a free cash flow value of EUR 20.8 million. Let us turn to the lookout for 2025. The service revenue in 2025 is expected to be at the previous year's level of around EUR 3.3 billion. Regarding the EBITDA, we are expecting it to fall by about 3.4% to roughly EUR 571 million after EUR 590.8 million in 2024.
For the operational segment Access, we expect it to fall to around EUR 836 million compared with EUR 856.1 million in 2023. The decline in EBITDA is due to the expiry of the national roaming agreement with Telefónica, which provides for one-off payments every five years that are capitalized and then amortized. The national roaming agreement with Vodafone, which is commercially equivalent for 1&1, does not provide for such one-off payments. The use of the Vodafone network is recognized as EBITDA in the wholesale costs. In this respect, there is no change in EBIT due to the switch to Vodafone. The burden on the EBITDA is offset by the relief in depreciation and amortization in the same amount. The EBITDA in the 1&1 Mobile Network segment is expected to remain unchanged year on year at around EUR -265 million.
This includes roughly EUR -100 million in expenses for customer migration and for network upstream services, which will no longer apply after the complete migration of all customers in 2026. The investment volume, cash CapEx, is expected to amount to EUR 450 million. It was EUR 290.6 million in 2024, the previous year, mainly driven by the network expansion. Thank you very much for your attention, and we will now move on to answering your questions. Thank you very much indeed. I have a first question. That first row, from my point of view, to left-hand side, Mr. Polo Tang from UBS, and then we go to the right-hand side to the from Citi. Polo, please.
Thanks for taking the question. It's Polo Tang from UBS. I've got three questions. First one is in terms of guidance. You're expecting stable subscribers over 2025 and stable service revenues. Can you talk through what you're already seeing in Q1, and is there any reason why Q1 trends would be different from the full-year trends? Second question is just on the CapEx guidance of EUR 450 million in 2025. Can you maybe break down some of the key elements in terms of what is in this EUR 450 million? Specifically, how much is a spillover from the CapEx that you did not spend last year in 2024? Also, how should we think about the RAN CapEx, and how many cell sites or antennas are you expecting to activate this year in 2025? Can you maybe also talk about data center CapEx?
My final question is really just about the network build. Can you clarify if you intend to build to 25% population coverage or 50% coverage? Would you consider RAN sharing or some form of network sharing as an alternative to doing everything by yourself? Thanks.
Yes, thank you. Let me begin with the subscribers. This year, we are involved in a huge customer migration project. At the moment, we are actually migrating 50,000 customers per day from the Telefónica network that we have used on a wholesale basis, and we're now pulling these customers into our own network. In the framework of this migration, we are actually losing more customers than would otherwise be the case. That meant there were obviously various reasons for that.
It begins with us writing to customers who have an inactive contract or who always thought about canceling and have forgotten, and we're now actually reminding them that they still have this contract and that we want to change it over, and they now take that as a trigger point to cancel their subscription with us. It is also got something to do with the fact that the process is cumbersome. About 96% of our customers are migrated more or less automatically, where we basically have a profile SIM card, and we basically take a second profile, and we can do that over the air, basically.
With the change of the telephone number, we activate the second profile, but not every customer has a dual-profile SIM card, and only 4% of people have a dual SIM card, but the process still does not work because they have a special type of mobile phone that they are using, and we are supporting thousands of mobile phones, but not the finite last model that they might have bought, Lord knows where, or they have very special combinations. They have several dual SIM cards that they use combined and what have you. These processes are laborious, and we actually do lose customers by that. If we have to send a new SIM card and the customer calls the call center and possibly has to wait, there can be frictional energy losses. That is why we think that this year we are not going to show any growth.
I think we would that the normal growth that we would normally be making cannot be the same as before because we are migrating many more customers than last year. At the moment, 50,000 a day, that is a gigantic migration project, and that is why we are working towards keeping our customers stable as much as we can. Now, in the first quarter, I see a slight lapse with mobile phone contracts, possibly 10,000-20,000 customers less, particularly as we are migrating so strongly and so many. Those migration efforts are going to be huge in the first and second quarter. In the third quarter, it will tail off, and in the fourth quarter, there will be nothing, that small remainder left then. Towards the end of the year, you should be able to see more marked and more healthy growth again.
At least that is the plan and our perspective. My colleague, Mr. D'Avis, is going to say something on the EUR 450 million and the CapEx. Yes, we do plan to reach 25% of our customers and 50% of our customers. Nothing has changed with these targets, and we are building our network. If there is an opportunity of usefully and intelligently doing RAN sharing, we would do that as well. So far, we do not use any RAN sharing, and we do not have any offers on the table that would suggest RAN sharing to us, where we would say that this is stable enough that we could or should be doing this in the next few months. Why not? I mean, this exists in other countries, and we basically have it in Germany, where you have antennae used by several providers.
If there were to be the opportunity, we'd obviously use this. Obviously, the conditions would have to be right.
In regards to your CapEx question, as you know, our CapEx in 2024 was significantly below our guidance due to phasing effects. For our planned 2025, we see the phasing effect of 2024 significantly. That is getting to this year. Also, investments for our capacity expansion until the end of the year. We have the network expansion that also has an impact in our CapEx. Our CapEx budget includes also investments for our own buildup of new sites.
By how many sites you plan to deploy or how many antennas you plan to deploy in 2025?
I cannot publish the number now because if I then cannot fulfill this target, then we'll read that in the newspapers. We have now fulfilled our goal of 1,000 that we wanted to conclude in 2022, and we have been criticized for that, and we've had negative coverage due to that. Now, as I said, we are improving from quarter to quarter, but I don't want to say an exact number because then I'll be compared to that number always. We are on the right track, and we think that we can meet our goals. The second question, CEE of Citi.
Hello. Thank you for taking my questions. It's CEE from Citi Group. I have one question and one clarification, please. And my question is, your guidance on access EBITDA, you expect access EBITDA to be flat year on year, but in theory, you should have more benefits coming from Voice over LTE and also international roaming. I'm just wondering, when you're thinking on the EBITDA development, what other costs you factored in your guidance for 2025? The clarification question is on the migration. You talk about that your migration to your platform is in full speed. Is that the same case when it comes to Vodafone's national roaming and whether you still think that the migration to Vodafone's national roaming can also be completed by the end of 2025? Thank you.
[Foreign language].
Thank you very much. The access business this year is flat because we are not expecting any customer growth, no sales growth, or revenue growth. You're right, there are savings in the upfront input because of Voice over IP, because of national roaming. That is coming from mobile communications. If you look at the access business, you can see what we usually would have to buy.
In the mobile telecommunications business, we have our own production with Voice over IP. Similar to the MVNO contract, you have the effects due to lower purchasing costs and that compared to the costs we have in the mobile telecommunications business. Vodafone national roaming. We are migrating customers from our wholesale networks to the 1&1 network. They use Telefónica roaming, and later we will switch the roaming. The customers will not even notice. They do not need a new SIM card or a new profile, nothing like that. They are in the 1&1 network, and how we configure the network, our customers will not even notice.
We're migrating about 50,000 customers for Vodafone national roaming, now going up to 80,000 because we have many customers with Telefónica roaming already. In parallel to the migration, we're also switching the national roaming and want to conclude that process by the end of the year. This is within our timeline so far. Adam Fox, HSBC.
Thank you very much for the presentations. I've got three questions, please. The first one is on towers. In the past, you had spoken about potentially spending EUR 100 million-EUR 150 million of incremental CapEx on your own towers. There was a bit of a mention of that in the presentation. I wondered if you could just tell us where we are on that. Is there a moment during the year where we'll get an update on that, and what, if anything, are you waiting for there? The second question is on spectrum. The regulator has taken an important step in recent days, but the way that I look at it, at least, it could be quite a long time before you actually get access to the sub-gigahertz spectrum. Obviously, you've got negotiations with the other operators to go.
I wondered if you could talk a little bit about your perspective on that process, whether or not it changes anything in the way you're thinking about the build this year. I don't think it sounds like it does. Do you have enough visibility at this stage to begin to think about the capital markets day that you might want to talk about? Then my final question. My final question, which I appreciate, is a sensitive one, so I'll leave it to you as to how you want to handle it, but I'd love to hear how you characterize your operating relationships with Vantage and Rakuten at the minute. I'll leave it there. Thank you very much.
Now, could you please repeat the question about the capital market? My colleagues and I didn't quite get it. Please repeat that one, then I have all the answers to your questions.
I suppose I was saying we've been waiting for a capital markets day for many factors. Spectrum, I think, was one of the last ones that we were waiting for. There is some clarity from the regulator now, not perhaps perfect clarity. Do you think we're at the point where you're willing to host a capital markets day for the market then?
[Foreign language] .
Now, EUR 100 million-EUR 150 million for own towers you mentioned. I can't remember to have said something like that during the presentation. What we are spending per year will be much higher, about EUR 500 million , I'd say, in total for our own towers that we are going to build. The question is whether we will keep them or sell them to a tower group. We have not decided on that yet. In regards to the spectrum, the regulator has decided that the frequencies for Telefónica, Vodafone, and Deutsche Telekom will be prolonged for another five years. 2028 will then maybe bring another option with more diversity of what we have today for us. This is very positive. We have a large spectrum that's being offered, and that is better than having an auction with a smaller volume.
Of course, if I only need one-tenth, then of course, the impact is different from if I needed 1/3. Let's say if I need only 30%, then I don't have to pay as much. Obviously, if you calculate that, it's always impacting on how difficult it is for us as a market participant. Now, we would have liked to see a spectrum auction to then know what volume we can use. We will only see that in a few years. For now, we have a lease for the Telefónica spectrum. 2x 2 MHz is the current spectrum that we are leasing from Telefónica as part of a remedy that was part of the merger release. This lease ends this year, and the regulator has said that this spectrum will be continued to be leased for another five years.
Since we have the current lease agreement, we only need to prolong it. This should be a fast process. The national roaming negotiations were supposed to go quickly as well, and then it took two and a half years, so you never know. In general, I'm optimistic that this will be quite fast. The low-band frequencies, 700 MHz, 800 MHz, 900 MHz , this will be more tricky because the regulator said that they will need to negotiate with us again. We contacted Vodafone and Telefónica and asked for their offer so that we can use their frequencies in a cooperative manner where we have active masts. It could be that in two weeks' time we get a good offer from them and that's it. However, my experience tells me that this is not what is going to happen. I'm already in touch with our lawyers.
I told them to buy a new cabinet for all the files we'll need because, as you can imagine, that's three gigantic machines. There will be a lot of expert statements coming for the coming months. I hope that the regulator will keep a close look at it and will make sure that it's a fair process. The cost for the frequencies are clear, and that's what we pay in order to prolong the agreement with the application we have. Let's say we need 20% of the surface, the frequency makes us EUR 10 million per year, just to say a number. We should pay EUR 2 million. That would be fair. Who knows what the process is going to be like? We have competitors who will come up with a thousand reasons why this is not possible.
Maybe there is a hospital right where these frequencies are or a senior citizen's home, and the senior citizens will need the frequencies. If they can provide them to them, then they'll have a problem with the network and so on. In some negotiations in the past, they said, you know, we only have 500,000 capacities left for our own customers, so there is nothing to share with you. We will continue to build our network. As I said before, 5,000 locations right now. We are in development, and the low-band frequencies are part of that, of course. It's not just mid-band and high-band. We also have a lot of locations, not 100% of the locations, but many locations where we are developing and expanding right now. That is what we are going to keep doing.
We will also be able to use low-band. The earlier, the better, but I can't make any promises, but it will not hold us back.
About that capital market day, and I apologize once more for not having understood that. I mean, you're absolutely right. We do not have any real clarity about what the frequencies will look like. Besides that, the image is, in fact, getting clearer. We know what the speed of expansion is. We see what the customer development looks like. We do know the cost factors incurred. The image is getting ever clearer, and we know where we're going. I think that frequency topic is one that we're keeping an eye on because that could obviously make a difference depending upon from when we can use these frequencies and what the conditions will look like most precisely.
Then we will be in a position to make a proper statement and make a forecast that we will actually match in the years to come because otherwise it does not make sense because if you have so many unknowns still in an image, then it is not worth anything. I think this is something that we should be able to do very soon. Now, on that operating relationship with Vantage , I mean, we actually have good relations on the working level, but also with Rakuten, we have got the same. With Vantage Towers, we have got the antitrust proceedings. I do not know if I am allowed to call it antitrust proceedings, but the antitrust authorities are looking into that.
I don't know what the right technical term is for that, but we've been dealing with that for two years, and I'd very much hope that this will come to an end so that either the antitrust authorities say they find that there has been an infringement or not, because I have no influence on that. The operational relationship is working fine. That's okay.
If I could just clarify the towers question, because it may be my misunderstanding. Your CapEx budget is EUR 450 million this year. I thought there was a chance that because of some of the challenges you were facing in acquiring sites, that you might, on top of that, build out some own sites with a view to a sale and lease back subsequently. Is that the right understanding or not? Is there an envelope of CapEx that you might end up spending or money for cell sites that you might end up spending during the course of the year if you have to do more builds yourself? Have I misunderstood?
As I said, in 2025, CapEx includes investment in passive infrastructure and our own construction. I mean, if we could speed that up, we would obviously do so. Obviously, yes, there would be a higher CapEx need. At the moment, we feel quite happy with the approach that we're pursuing there and with the numbers that we've mentioned. I think this is sufficient. There is enough room in this coffer of EUR 450 million for our own construction too. Thank you very much. The next question, Mr. Wolf Martin from Shareholder Value, and then we go to the third row on the left-hand side.
Hello. Wolf Martin from Shareholder Value, could you give us an update on your negotiations with Rakuten? I have another question. You have adjusted your dividend for United Internet, but the savings that you've made with the frequencies goes to 1&1. Why haven't you adjusted the 1&1 dividend as well? I don't know what you mean about an update of the negotiations on Rakuten. We obviously negotiate with them basically daily about change orders, about new things, whatever there is. I don't know what you're trying to get me to. My question is that you say in the guidance for 2025, there aren't any payments for delays in your dealings with Rakuten because weren't you going to negotiate that with them again? Take out Rakuten. It is absolutely right. We're negotiating with a partner about compensation and damages.
As we have liabilities of not telling about, I mean, keeping these things secret, this is not included in any guidance if there were to be a payment. Now, as to the dividends you were talking about, you know, in United Internet, we have a clear dividend policy, and that dividend policy has been undercut repeatedly in the past few years. I mean, we basically had very small amounts, and we did not do that because United Internet was not earning enough to pay the dividends. We actually did not know how much money 1&1 would have needed if there had been the opportunity of buying additional frequencies. Until this weekend, I mean, that could have still happened.
If there were to be an auction for a frequency, then obviously we want to have our coffers appropriately full and then have the cash pool with United Internet. By now, we have used the cash pool that we still have with United Internet because we have huge CapEx for the next few years, and our share is not trading particularly high at the moment. If there were to be an opportunity and new money was needed, then we would have called upon United Internet to take out a loan for us, some financing vehicle for us. That is why we have kept some powder dry so we could have helped if the opportunity had arisen. This has now not happened, and the opportunity is postponed.
We assume that in a few years to come where the wide range on offer is going to be much broader, and then hopefully the auction and the tender with the bids is going to be hopefully cheaper because there will be more available so that we can basically pay for this auction and any auction that might come for a new frequency. Obviously, we could pay a dividend now, and then we would have to increase our capital in three years' time. That actually, I think, does not make much sense. Therefore, we have said, let's stay prudent and straightforward. Let's keep our hands on this money at the moment. 1&1 is debt-free, but we have to continue building the network. We have also seen last year the risks in the network. I mean, we had an outage for three days last year, which was huge.
All of that is not without risk. I mean, building the network is not risk-free. We do not know what the frequencies are going to cost. Will we be able to maintain and keep our towers? Will we have to do a sale and lease back? How expensive will the next auction be? I mean, we really have to make sure that to be able to pay this money, we would have to take out a loan as 1&1 to then pay a dividend and build the network and finance the next bid for frequencies. We thought that that was not the right way to be going. Now, let's move over to the left-hand side. [Frank Ruter] from AHB Capital, please. I have a CFO question, if I may.
In the last meeting last year, when we talked about that you had a decline in the gross profits with stable income at the time, could you possibly show the effects again that led to this gross profit decline? I would actually like to understand whether there are effects included as a consequence of building the network. Thank you. Yes, the decline with the gross profit is completely owed to the network building activities because in the operational segment of Access, we have done really well. The gross margin was increased very noticeably last year. It was a strong year. I think the burden more or less exclusively comes for financial effort for the 1&1 Mobile Network. If you were to roughly quantify that, I mean, those effects, network gains to improvements in the Access segment.
If my memory serves me right, I think we have an increase in the Access segment of 5.7% and correspondingly a decline in the 1&1 Mobile Network. Up here, Camillo Azzouz from Amber Capital in the first row, please.
Publication. Three questions on my end. Last year, you were communicating on churn from migration to be around 250,000 [netads]. That would be largely low-value [netads] for the reasons described by Mr. Dommermuth. If we take that into account and the guidance of stable [netads], should we assume that your commercial additions should be around 250,000 for the year, including mobile and fixed? Keeping in mind that fixed could be slightly negative, mobile could be closer to 300,000, highlighting a strong pickup in commercial momentum. This is my first question. My second question is on the 5G costs.
I see that in 2026, you're going to start reducing that number for EUR 100 million, which is better than expectations. Can you help us understand by when the 5G segment will turn into a positive contribution into your numbers? My third question is around the optionality around RAN sharing. I welcome the comments made by Mr. Dommermuth that you would explore potential optionality around RAN sharing. How should we think of the CapEx profile of the company if you were to use RAN sharing as a tool to get from 25% to 50%? I imagine this would be a lot more asset-light and therefore improve the cash flow generation. Thank you.
Thank you very much for that question. On that first question, you're absolutely right. It's true that in 2025, if we weren't doing this huge migration, we'd be talking about 200,000-300,000 of net growth. I think that is the right figure to bear in mind. That is also what we're expecting duly. I think by 2026, obviously, it depends upon the competition a bit, but that is an absolutely correct assumption. That is the order of magnitude that we're expecting. We think in 2025, with this huge migration process, yeah. It is true that we are expecting a very slight decline in the fixed alliance and a slightly higher growth in the field of mobile telephony. On the mobile phone network for next year, we just said this year we have about EUR 100 million to spend this year that will not have to be spent next year.
We basically have to pay compensation for taking over and transferring mobile numbers from other providers. We have higher call center costs because customers are asking lots of questions. There is uncertainty. Not all customers are on the new networks, and we cannot lift all savings potentials like with international roaming and others. Therefore, we assume that next year things will look EUR 100 million better, but we're then still not in the black with our network. I can't tell you how long precisely because you'd obviously have to know what the frequency costs will end up being, which we don't know and cannot know. I can tell you what I really think. Of course, we all want to have our network up and running in a profitable way. I mean, no doubt about this.
I'm now sitting here in my 38th year, and whether we're getting it to be profitable in year 40 or 41 is not going to make a big difference for the company in the long run. We're obviously trying to do it as quickly as possible, but it all has to be a good quality. We have to understand what we're doing at every point. I've just shown you our map, and we basically are tying each antenna in with a fiber cable. I mean, we have kilometers of fiber optic connections that we're putting into the ground. I mean, this is just going to make us very stable, but it is not cheap. Others do it much more cheaply.
I mean, all of these data centers that we're building, I think most of that could be very slightly cheaper, but we have decided to build something that is future-orientated and does leave room for further development. This is the most modern infrastructure that you can find anywhere. I think what we're doing is the most modern structure and infrastructure in existence, and that is what we are guided by. We want to do this state of the art, and we want to do it best possible. While the RAN sharing question is a good question, I mean, what would happen if we had between 25% and 50% of RAN sharing? I think those are dimensions in RAN sharing that I don't think about, I have to say.
I think about the odd location that we could cover by RAN sharing, but I'm not imagining that we are no longer building any locations or any towers or whatever because we have a RAN sharing partner and use their sites. Yes, it might be, but it might also snow in December. I don't actually see any reason to believe that right now. Mr. Schlote here at the first row, and then we'll move into the third row again.
[Foreign language] .
Klaus Schlote at Solventis. Mr. Dommermuth, you said 5,000 towers in development, and per quarter, you can be compared to the other three large stakeholders. Now, maybe you would want to tell us a range of their numbers so we have an indication of where you're going. A more general question. We'll have a new government in Germany, a new federal government.
We already know that. We still do not know what the government is going to look like exactly, but it seems to be a mixture coalition of the Red and the Black Party, the SPD and CDU. That means investments in defense, in infrastructure without precedence. Is there a business for you in that plan of the current government that is being formed? In regards to the towers, we have about 200-300 locations per quarter, and we want to see an upward per day, and we want to see an upward tendency there. In regards to the new federal government in Germany, so far, I do not see any business we could get from that investment plan. In regards to landlines and mobile telecommunications, for mobile telecommunications, it would be easier if we had a license for new masts, and that would definitely help.
There have been talks about that, but it's not really a new business model. It'll just alleviate the current business. For fiber optics, I heard that they are counting on market-borne solutions. That is about what we have at the moment, whether this is going to be increased or not, and if you can get subsidies or not. I cannot tell you today, as 1&1, we're using the optical fiber of 1&1 Versatel. I don't see how this could be a great new business for us compatible with our broadband business. Thank you. [Simon Stubbeck], Baum Research. I would like to ask the following question. In regards to CapEx, if I look at the delta of your guidance, November 2024, to the numbers published this week, then there is a gap: EUR 130 million phasing effects, EUR 140 million, something like that.
Is that correct? The CapEx expectations, would you say that in 2025 we'll have a peak CapEx from what we've seen? In regards to the EBITDA, 1&1 Mobile Network, you said EUR 100 million additional expenditures. Could you give us the split between the migration and the upfront costs for the network? In regards to the quarterly EBITDA trends, you had a slide with the trend of the EBITDA over the different quarters. It's a negative trend. Is it fair to say that this trend will be turned around next year? You said Q4 will be much stronger than Q1. The last question is this: What would be the worst case, the worst case in regards to the point in time to then have a profitable business? Thank you very much. In regards to your first question regarding CapEx, yes, that's correct.
There is, or there was, quite a large gap. There was one gap between booked CapEx and cash CapEx. There is a phasing effect where you can really say that the payouts in 2025 will come. On the other hand, there will be effects from 2025 going to 2026. You can't really say that you'll have the full EUR 130 million as phasing effects for 2025 because some of that will be moved to 2026. The peak CapEx at the moment, I think we can't say that 100% because we still have those phasing effects. There are delays. With the dimension we are working with, this could easily be between December, January, February. There could be a shift, but we are at a high level already. I think the numbers will improve with the CDCs and final capacities being concluded.
Once we can focus on the rollout, the CapEx will go down. The EBITDA for mobile telecommunications, you asked about the EUR 100 million and what they consist of, and I can't give you the details here. However, we have mentioned the main effects, the migration, and the larger effect is that in the EUR 100 million. In 2026, we won't have that anymore. Additional savings due to national roaming and others will then have a positive effect on the EBITDA next year. Your assumption is correct that the EBITDA in 2025 of mobile telecommunications, of course, from quarter to quarter will improve. Mr. Dommermuth talked about the migration process and the huge number planned for 2025. Most of the migrations will take place in the second quarter. We've already had a lot of migrations in the first quarter.
There's going to be a peak in the second quarter. In Q3, we'll see a slowdown towards the end of the year. We'll then have migrated all our customers to our network. We'll start seeing more and more savings and fewer expenses for the migration towards the end of the year. You asked the question of all questions. Is the network profitable? The own production costs less than what we have to buy. I have this answer that may not be satisfactory to you. It's a fixed infrastructure, and the earlier we use it at full capacity, the earlier it's going to be profitable. It's a question of the costs and sales and volumes. If we're quicker, then it's going to be better for us, but I cannot give you an exact day.
You ask us each quarter, and that's okay. What I really want to focus on more is cash, more than on EBIT or EBITDA. The cash flow now, I think the expenses will soon go down. We will not have to put much more cash in there than we have savings. We have the core data centers. Once we have the Far Edge data centres as well, we will only invest in the rollout of the antenna locations. The more we can lease, the better. The fewer CapEx we need, the more own production we can have. In a few years, we will be able to see that on the cash side. It's really working out. The EBITDA will see some more delay, EBIT as well. For cash, in a few years, this should be done. What we should also look at is the numbers.
You know, we know that these are very large dimensions, and 30% is paid by the government because we do not have to pay taxes on that part. Of course, the other 70% also hurt, but it is only 70% that we cannot pay out as a dividend, not 100% at least. Yes, go ahead. Street Alpha Hans- Dieter Klein, I have a question about the real-time applications. You are building a state-of-the-art network with very large investments, very positive, I must say. Now, what is your feeling or your experience when customers ask you about this? Maybe you can give us two or three examples. Now, I would love to give you many very good examples. We have tested it with a game server, and we had low latencies, you know, below 3 ms. 5G is 2.5 ms, I think, the signal.
We were able to do that in half a millisecond to go to the data center and back. That is quite fast. In principle, it is working. It is physics. If you have fiber optic, then of course it is faster if it is a shorter distance. However, today I cannot give you an application, a real-life application where people are willing to pay for that right now. We know, or we think that this is going to come, and we want to be prepared. We have to do it once there are applications that can actually bring us revenue. People need to be willing to pay for that. For now, we have to be honest and say this is nothing we are going to do in the next few months.
I think we're coming towards the end because our parfent company, United Internet, will also have their press conference in a moment. Maybe two or three more questions. If there are no more questions, we would like to thank you. Investor Relations and our CFO will still be here for you. Probably not today because now it's our colleagues' turn, but in the coming days, we look forward to your reaching out to us. Thank you very much for your attendance and for your questions.