Good morning, ladies and gentlemen. I'm happy to welcome you on behalf of the board of Dr. Dommermuth and Dr. Raß to the board meeting of our presentation of the fiscal year 2022, and we will give you an overview of the year and the outlook for 2023. After the presentation, you will have opportunity to ask questions, and I hand the stage to Mr. Dommermuth.
Thank you, Mr. Keil. Good morning, ladies and gentlemen. I want to start with the company development 2022. Many of you know our business fields. We have the broadband connections, where we provide DSL and fiber. We have good network quality and about 4 million access points. Beside that, we offer mobile contracts. Here we have a lower value contribution than the broadband.
We produce some of the products ourselves, either we ourselves or sister company, Versatel. On the landline, we get the primary provide by Vodafone and Versatel. We have 11.68 million mobile service contracts. The customer contracts in general in 2022 have increased to 15.78 million. That's a growth of 350,000. 490,000 of these are from the mobile network, and in the broadband, we had a drop of 140,000. These contract figures include shifting effects from the TKG amendment 2022, which cost about 250,000 contracts by earlier cancellation dates. Operational, however, we had a growth of 600,000 contracts. This year we will see a bigger growth.
The revenue was increased by 1.4%, the service, in this by 1.7%, and the other revenue, which is rather weak in margins, by providing smartphones, for example, increased by 0.3%, all together EUR 3,963 million in revenue. Our EBITDA by segments, we operate our access business, that's DSL, fiber, mobile. In the access field of business, the operative operating EBITDA increased by 5.1%, and that means if we look at the year 2021 comparing, we have to have off-period revenues out, and the actual EBITDA in 2021, as we had said, was EUR 709.7, and it has increased in 2022 by 5.1% to EUR 745.7. That is a very good development, fully meeting our expectations.
I just said we are building up our own mobile network. We have start-up costs for this and they amounted to EUR 52.4 million last year, slightly higher than the year before, where we had EUR 37.9 million. The CapEx also increased by our own mobile network, where we invested in antenna, computers, and software. This was EUR 245 million in 2022, EUR 232.7 out of these for the mobile network and EUR 12.3 for our ongoing business. Here you see an overview of the year 2022. Customer contracts increased by 2.3% to 15.78 million. Revenue by 1.4%. Service by 1.7% to EUR 3 million 150... 75.
EBITDA has slightly dropped, if we take out the off-period revenue, we have a growth of 3.4% in 2022, increasing to EUR 693.3 million. The EBIT, similar if we take that undiluted, it increased by from EUR 507 by 5.4% up to EUR 534. The EPS without the off-period accruals increased by 7.2 from EUR 1.49 to EUR 2.08. As we're gonna present a proposal for the dividends on the shareholders meeting at EUR 0.5 as the minimum dividend that we have to pay according to the legal stipulations. You know that we only pay this minimum dividend in order to fund our network building.
That takes us to the most important topic, which is the operative business, which is going well. We have the mobile network, and I'd like to update you on where we are in that respect. On 28th of December, we started the network. We are approaching two core computer center out of four, which we're gonna need. We operate seven decentralized edge data centers out of 24, which we'll need. We have started the first 61 far edge data centers. We're gonna need 105 and 150. They are connected by 94 antennas, which we have taken over. 14 of these are in operation. There is a certain lag in this because the antenna site is passive infrastructure, so the mast on the roof or on a tower. This passive infrastructure has to be fitted with optical fibers.
It has to be wired up in-house, and at the end of the in-house wire, we have the mast with the antenna, one or several antenna. That takes a while, usually three months that we calculate to do that. We are currently testing our network and preparing it for mobile application. At the moment, we only offer wireless access because we don't have national roaming. That's gonna be provided by the end of the year by Telefónica, and then we can offer so-called eMBB services, which we would need for smartphone operation. We are currently testing the national roaming with Telefónica. We are testing international roaming as well. Here we decided to go with Orange as a wholesale dealer who provides this worldwide.
We are testing the mobile operation. We have a number of tests and certifications ongoing with device manufacturers because we need to have the certifications for that as well. It all looks very well for the time being. There is ideas whether everything is working, and I can assure you it is working well so far. We have absorbed the first tests with subsidiary of Accenture, and they show that the quality and the data capacity that we have, the speed that we provide is quite competitive with the networks that are already available in Germany. This is why I can say the network is up and working, and I'm confident that we'll be able to take the next steps over the couple of months to come. We also have question concerning the costs of that network operation.
Doesn't Open RAN need too much energy? Is it all calculable? I can say tell you by now that after we have been up and running for three months, the measurements that we have carried out over this point in time, our assessments and assumptions are confirmed, and that means that we have three rough working hypotheses. One being that if the customers have migrated 2025, we want to migrate the Telefónica and Vodafone network customers to our own network. The costs that we have for the computer center, the core computer centers, for the edge computer centers, and the far edge computer centers, and our team are gonna be, in the sum, as high as the savings that we generate by voicing and international roaming traffic. How is that?
We buy voice by a minute price at the moment. As soon as we operate our own call, we go buy voice as data. That is much cheaper because voice is not very data intensive. We have international roaming, which will get much cheaper from Orange than we do today. That means once the customers have migrated, we have no single antenna on a calculation basis where we have minus EUR 127 million now, we will have a small black figure. Even broken down to the EBIT, not EBITDA, where we have to calculate depreciations, even there, we would profit from it. We don't only want to keep the decentralized and core data centers. We also want to build a complete network. For that, we need antennas.
As of today, we can say that if we have an antenna in full operation, including the low-band frequencies, which we still don't have, we want to have them from 2026 on, and the home coverage that we want to have, first of all, that's gonna be 15% of the German households. The operation costs of a single antenna site are gonna be high enough to parallel the costs that we have for data as if we weren't operating that antenna ourselves. For example, we have a site that in all with rent and optics and power costs a sum X. Then for the data that we send from that antenna, we will pay exactly that amount today to our suppliers.
That's the assumptions of our business case, and these are confirming. Besides the data center and the antenna sites, we need a radio range. Here we assume that the frequency costs are going to be generated by the new business fields, which will be only possible by our own network. That's fixed wireless access, branded reselling or B2B tariffs. That means from the network operation in the first three months, we can see that we have the assumptions confirmed. Looking at the antenna locations, we have to be honest and say it doesn't look as good as with the network operation itself. You surely noted before that in 2022, we could only get five antennas into operation out of 1,000 sites that we have aimed for and that we could have and should have had due to the frequencies.
We have a number of companies that we cooperate with. Our frame contractor, the biggest partner is Vantage Towers. As you may have heard, in February, we filed a complaint to the Cartel Office because the delivery duties by Vantage Tower weren't fulfilled. How are we gonna move on from here? If we look at the forecast of our partners, they are now planning to provide a total of 1,207 sites by the end of 2023. That's the path of infrastructure. We have heard about delays, which means that during the year there's gonna be delays and we are gonna be mainly supplied in Q3 and Q4. The later we get a site, the more difficult it'll be, of course, to get the operation or the antenna into operation in 2023.
If we get the site in November and December, we'll probably not be ready to put it into operation this year. That will be delayed, but we'll have to see how we'll play along with that. The important point is that we do get the site, that we do get the tower. We can, we put in the fiber optic with our sister company, 1&1 Versatel. Building up the antenna is where we have our partners for which doesn't issue any problems. The forecast is at 1,207 sites. We know these sites, and we have clearly analyzed the individual building blocks and what time slots we have for the building blocks, and this planning is realistic.
Whether it really, in the end, will actually turn out a supply in accordance with this planning is something I cannot promise as of today. Due to the experience that we've had in the past, we need to proceed one or two quarters. In the Q1, the forecast was fully met, but at a low level. We'll have to see whether the ramp-up will be kept as promised. If the ramp-up works as planned, next year we are going to talk about another 3,000 antennas, which is because the TowerCos will supply more Vantage and merchant tower. We will also start building ourselves. We know about the Eubanet contract. They are building up a team for us which will acquire sites across Germany.
If we add it up, built and acquired and the tower company's sites, then we have a contractual scope of about 3,000 antennas in 2024. I am expressing myself as carefully as I do because we had our lump backs in the past year. We'll have to see whether the contracts are fulfilled in 2023 and 2024, and we need a little more confidence than we have gained in the Q1 this year. Of course, you can imagine that I would have liked to proceed much quicker and faster. It didn't work last year. This year looks better. It is not so that we can let our guards down and we can 100% rely on the machinery running. We'll have to wait some more time down the line. What's planned next?
The planning steps have not been changed according to what we have presented to you before. In the next quarter, we're gonna connect the international telecommunications networks. We're gonna test the international roaming. In the Q3, we're going to have the first and fourth core data center, the third being set up already. We're building the fourth one. National roaming will be handed over by Telefonica, that will be the starting gun for marketing the national tariffs. We're going to stop the MVNO sales and market our own network. In order to utilize the network in the Q4, we're gonna start migrating the existing customers until the end of 2025. Well, that takes me to the end, I'd like to hand over to Mr. Huhn, who will provide you our financial figures.
Well, thank you very much, Mr. Dommermuth. Good morning again. Let's start with the earnings. On page 17, we can see the revenue at EUR 3.963 billion, a plus of 1.4%, as has been indicated by Mr. Dommermuth already. I'll continue with the profits, the cost of sales, which increased from EUR 2.709 billion to EUR 2.734 billion, which leads to a gross profit from turnover of EUR 1.229 billion, a plus of 2.5%. That, of course, includes the EUR 39 million of out of period revenue that Mr. Dommermuth had mentioned earlier.
If you consider that, we would have a result of EUR 1 billion of EUR 60 million. That would be an operative growth of 6% if we consider that. If you see that our revenue increased by 1.4% and the service revenue of 1.7%, we can see that we have a nice disproportionate development in the overall profits. The sales costs increased to EUR 509 million. This is largely due to marketing efforts and other efforts for distribution and marketing. Administrative costs increased from EUR 126.1 million to EUR 110.9 million. Sorry, decreased to EUR 110 million.
That is largely due to the fact that in 2021, we had the expense for the 1&1 mobile phone network shown in the turnover figures. In the past, that had been included in the administrative cost because the network hadn't been operational yet. If we account for this, we had a proportionate increase of the administrative costs. The other operating income and expenses increased by 20%. The impairment losses from receivables and contracts increased from EUR 78.4 million to EUR 107.3 million. That is largely due to a changed payment behavior that we find in our customer base after the corona pandemic. During the pandemic, customers paid punctually, which now changed in the business year 2021. That's an important effect.
Another effect is the change of the blocking limit after the legal changes, which means that the receivables are higher than they were in the past. That means higher impairment losses from receivables as well. The profit from operative activities decreased slightly from EUR 546 million to EUR 534.9 million. If we eliminate the EUR 34 million in 2021, the operative development would be EUR 503 million in 2021 to EUR 534 million in 2022, a plus of 5.4%. The financial result in 2021 was negative at -EUR 11.6 million. In 2022, it was a better financial result at only -EUR 3.4 million. Two effects are accountable for that.
The investment for white spots that are shown here, which in 2022 were EUR 5.4 million less than in the prior year. Income from interest increased by EUR 2.8 million compared to the last year. The profit before taxes decreased from EUR 535.1 million to EUR 531.5 million, a very slight decrease. After accounting for the EUR 39.4 million in 2021, the profit development of EUR 495.7 million to EUR 531.5 million would have resulted. Tax expense was at EUR 164.2 million in the last year.
For the group result, that the consolidated result then is EUR 367.3 million for the last business year. On the next slide, page, you can see the balance sheet, which increased to EUR 7.063 billion, a plus of EUR 193.4 million. You can see that the short-term liabilities have decreased by about EUR 100 million. That is due to the reduction of the cash receivable from United Internet. This is just opposed to other receivables. Liabilities, so that we have the bottom line that we have. Our long-term assets have increased to EUR 5.4 billion.
That's due to an increase of fixed assets of EUR 220 million, which is due largely to the establishment of a mobile phone network. Increased expenses, which increased EUR 224 million, which are due to the contingency payments to 1&1 Versatel and Deutsche Telekom. On the liability side, we can see that the short-term liabilities have increased by EUR 106 million to EUR 549.7 million. Sorry, they have decreased that far. That's to two short-term liabilities from supplies and services, and also the reduction of other liability of EUR 50 million that have to do with down payments on that.
The long-term liabilities decreased to EUR 1,127.6 million, which has to do with the payments for a frequency range, the EUR 61 million that we paid in 2022. An increase of equity to EUR 5,579.8 million is of course, due to the bottom line of the last business year. On the next slide, we can see the cash flow. The net inflow from operating activities was at EUR 180.6 million, composed of the cash flow from operating activities minus EUR 40 million from a decrease in increase of inventories, change in trade and other receivables from trade and other payables.
Decrease in income taxes minus EUR 13.2 million and the change in trade from receivables and other payables at EUR 110 million. The accrued expenses account for EUR 154 million, and the increase in inventories minus EUR 23.9 million, and that's is the net investments. The cash flow from investment activities was at EUR 97.4 million, composed partially of EUR 245 million CapEx, and the decrease in the investment of free cash with United Internet at EUR 148 million.
Cash flow from financing activities is at minus EUR 83 million, EUR 8.8 million, resulting from dividend disbursements, EUR 12.9 million from repayments of lease liabilities, and EUR 61.3 million from the repayment of the 5G spectrum liabilities, so that we have a free cash flow of minus EUR 64.5 million. On slide 20, we give the bridge from EBITDA to free cash flow. On the left-hand side, you can see the EUR 693.3 million EBITDA for the last fiscal year.
The receivables and other assets at minus EUR 20.2 million, the deferred expenses to the tune of EUR 154.9 million, liabilities and deliveries and services and other account for minus EUR 81 million, contract assets minus EUR 39.7 million, the total of other working capital at minus EUR 39.5 million, taxes minus EUR 177.5 million, CapEx minus EUR 245 million. We wind up at the minus EUR 64.5 million free cash flow mentioned earlier. Let's take a look at the forecast for the current fiscal year, 2023.
We expect an increase of service revenue of about 2% to about EUR 3.23 billion, after EUR 3.175 billion in 2022. The EBITDA is expected to be about EUR 655 million, after EUR 693.3 million in 2022. This is composed on one hand by the access segment, where we expect an increase of about 4% to approximately EUR 775 million, after EUR 745.7 million in 2022. The segment 1&1 mobile network, we expect to have an EBITDA of minus EUR 120 million roundabout, largely due to the start-up costs for our mobile phone network.
Last year, the EBITDA in this segment was at EUR -52.4 million. Concerning new subscribers, we expect about 500,000 new customers in this year after an additional 350,000 last year. We expect the CapEx to be EUR 320 million after EUR 250 million in 2022. Again, just like last year, these are largely due to the establishment of the mobile network, so investments into antennas, computers, and software. Well, so much for this. Thank you very much. I'd like to give the floor back to the operator for the questions and answers. We start with the questions and answers. If you wish to ask a question, please press 1 and asterisk on your phone.
First question is from Mr. Hammerschmidt.
Thank you very much. The first question which I have is what's the ambition to have over 1,200 sites compared to the ones that you have today? What makes you confident that you will be able to deliver this figure? In other words, what's different this time than it was before in October, for example? In this context, could you give us some background on the delays? Was it issues with to finding the sites, material bottlenecks, or other reasons? The second question is could you help us to break down the CapEx, for example, for the antenna capacities and the EUR 120 million which you mentioned for 5G? The last question concerning the network has been expected for a long time. Is it going to come up and when?
Yeah. Well, let me start with the first question, and I'll hand over to Mr. Huhn. We have taken over 94 sites. We're speaking of the passive infrastructure here. We're planning for 1,207. That makes us quite confident. The new first cast was on spot on with the 94 sites. The forecast is looking at a faster ramp-up than initially initiated. Initially, we had planned for more sites in the Q1, and the same goes for the Q2. They'll be shifted to quarters three and four. We don't like that particularly much, but that's the way things are, the reality we're faced with today. Why are we confident that this can actually be implemented?
We have renegotiated contracts, we can now see in the contracts the individual building blocks, even during the construction of the various sites. We won't be surprised at the end of the day, that there won't be a site. Instead, we will realize early on whether, for instance, a lease contract was added with the owner of a roof for an antenna, then there was a walk down, and then a draft plan was made, coordinated with the current operator of the site, and then radiation calculations were made.
For all these building blocks, there is a certain lead time, and we checked it all that we can say, "All right." If we see in which stage the various sites are, then it's realistic that by the end of the year, these 1,207 passive sites can be added. That makes us more confident because we have increased transparency compared to last year. If you ask what has changed, well, that is what we can see now. There are changes in management as well. At Vantage as American Tower in Germany, management changes in Vantage is imminent soon. Another company has taken an interest in Vantage, and we hope that this will bring in a bit of fresh wind.
We hope that this will ensure that we will get those sites. You had asked, how come it didn't work last year? Was it due to the material or what's the cause? You mentioned Vantage as well, particularly. I can't tell you, of course, because I don't have any insider inside information there. What I can tell you indeed is that Vantage or rather, Vodafone at the end of last year had 1,600 5G sites that they had registered. Vodafone uses a Vantage infrastructure.
If you compare that with the number of our sites that we had at the time, then you can see that we had practically none. Vodafone had 1,600 and a 5G location for Vodafone means building an antenna that broadcasts at 3.6 GHz. Here the wind loads and the sizes and statics will be similar to others. Why did Vodafone have 1,600, we didn't? That's maybe a question that the antitrust authorities will look into because they are verifying now whether they admit our complaint. I can give you some inside information. I got a lot of explanations, but to what extent every single piece of information that we got had an influence into the overall picture, I can't tell you.
Yeah. Then Sudev.
Okay. I'm coming to the second part of question, OpEx and CapEx. The EUR 120 million EBITDA for 2023 were mainly driven by the operation, HR costs, electricity costs, also monthly sums that we pay for the connectivity of our core data centers by optical fibers and the amount of space that we need. That's one big area. Besides that, in 2023, we have had initial payments for the migration of our customers. These are in the EUR 120 million as well, also expensive for the integration of our technology, the end customer business into our own network. These are the major factors which are covered by these EUR 120 million EBITDA and OpEx. In the CapEx, the EUR 250 million includes more invest into core data centers.
In 2022, we had invests mainly in equipment of the data centers, software servers, and in 2023, that is going to continue. Of course, in 2023, we're gonna have another CapEx sum for the active infrastructure that is for the antennas, which will be added. That's mainly the CapEx. The third question I got, if I got it right, is the NetZwerkTax. I think you're talking about CMD. As soon as we have a reliable forecast, and in the sense that we see that supplies are actually delivered and we can depend on these figures, then we'll see ourselves capable to publish the details of our business plan. At this point in time, we are of the opinion that this doesn't make sense as yet. This is surely going to be a topic that we're gonna readdress in the course of the year.
Well, thank you very much. The next question, Mr. Ulrich Adler, Societe Generale.
Well, thank you very much. I have two questions. The first one being quite mundane. The service revenue growth of the Q3, that's 1.1%-2%. In the previous quarters, it had dropped to about 0% in the last quarter, so the reference quarter wasn't too challenging. What's your explanation for this? Which are the drivers for this going back to 2% in 2023? That's what you're targeting. That's the first question. The second question is the CapEx budget. Looks like it's relatively limited for 2023. The initial guidance for 2023 had been given and it has been pared back. It looked like some of this has been brought over to 2023 from 2022.
What I'd like to know, what I didn't understand was the cash CapEx. How much additional fixed assets CapEx will there be? Fixed asset additions that will go via leasing contracts that don't go to cash CapEx? Actually, that was a very low volume last year, about EUR 1 million, I think. What is the situation in 2023? Could you refer back to that, please? What will be the total sum? Thank you.
Yeah.
Answering the first question, the service revenue. There is two main issues to be mentioned. One that we saw last year as well. Seasonally driven in the Q4 of the previous year, we had a drop in turnover of about EUR 7 million compared to the previous quarter, simply due to lower international roaming revenues. We are seeing this this year as well. That makes up about EUR 10 million and a non-operational effect, which is due to IFRS balance sheet requirements. We do that by the portfolio. That means toward the end of the year, the correction quarter are checked and adjusted, and we look at the logic and depreciations forward. In the Q4, that had an effect of about EUR 6.5 million, which is going to cover back up over the next months.
That's not a permanent effect, that depends on the balance sheet calculation on the Q4 due to the regulations. In the EUR 250,000 losses to do the new legal requirements. This is premature, that has an effect on our service revenue. In that respect, it's going to see the course of 2023 to manifest the planned growth and correspondingly an increase in the service revenue. For the Q4, these are the two factors that pay into this account, the season and the IFRS effect. As far as the CapEx budget is concerned, when we forecast 2023 and 2022, we forecast about this dimension. Of course, the share is a little lower for the antenna. We're more aggressive in the ramp-up planning for the antenna sites.
Now the value that we expect for this year is a little lower. Beyond that, the investment into servers and data centers we are planning for 2023, they're still the same amount as we had planned last year. That's cash CapEx and the effects according to IFRS, if we have the rental payments, represented as CapEx, that depends on the ramp-up of the antenna sites this year. In this case, I would rather wait until midyear to have reliable figures to give you a guidance here that strongly depends on the actual delivery of the antenna sites over the next months to come.
Well, thank you very much for the answers.
Okay. Questions from the English Channel. Carola Holze from Deutsche Bank first.
Thank you for taking the questions. I have two, please. Firstly, you will lose 5G access on national roaming at the end of 2025. What percentage of traffic do you expect to be covered by your own network at this point? Do you consider lack of coverage outside of this as potential operational headwind? Secondly, a question just relating to your level of commitment to the network rollout. It's understandably, and not to your fault, had a slower start, and you are still waiting for spectrum clarity. Do you see yourselves building under all scenarios, or do you think you'd be open to other solutions, such as a new wholesale agreement or consolidation? Thank you. Waiting for the answer.
Well, concerning the first question, the coverage 2025. We're looking at the household coverage and the traffic coverage. If that is lower, our savings in national roaming will be lower. Mr. Dommermuth talked about the one-nine-four-nine mobile network. It is positive to have the operation costs for our operation centers and for HR that can be refunded by savings that would occur without the antennas already. If we build less antennas, we have less expenses and of course, less savings so that we don't have a fixed cost block building up, and we lose the savings. We do everything possible to build up everything as quickly as possible to generate the savings as quickly as possible. Coming to the second question concerning the rollout speed and the range, I will hand over to Mr. Dommermuth to answer that question.
Well, at the end of the day, the type of spectrum allocation is decided upon by the Federal Network Agency. If you ask what impact does a slow rollout have, then I would say it doesn't have a major impact. We can show that we're doing everything to build the network. The problems that we can see are not due to the fact that we placed our orders too late or that we didn't get any clear commitments by via contract, but that the contracts weren't fulfilled. The problem really doesn't rest with us. At the end of the day, this will be decided by the Federal Network Agency.
The frequencies we're talking about here, beginning January first, 2026, will fall back to the Federal Republic of Germany. That's when, Vodafone's, Telefónica's, and, Telecom's, rights over these frequencies, run out, and then the Federal Network Agency will have to decide how to handle this.
Thank you. May I just follow up on the first question, if that's okay? Thank you for the detail. I was thinking more about the customer experience rather than the impact on your EBIT from what you may cover at the point you lose 5G access and national roaming. Would you be able to help elaborate at all, just, you know, what level of traffic you think you'll be able to achieve 5G on at the end of 2025 when it's just your own network which will be able to provide it? Thank you.
Yeah. Virginia. Well, we expect 1,200 antenna sites by the end of the year. That's from 2024 onwards, we can expect 3,000 antennas per year, which means that by the end of 2025 we would have 7,200 sites handed over to us. I believe that as we speak of genuine 5G made at 3.7 GHz, that we then will be fully competitive in terms of our coverage. Also, when compared to Telecom and Telefónica. Telecom now has 3,000 sites at this stage. I think Telefónica, 2,300, Vodafone, 1,600. That's a much bigger quantity than what we have today. I think that by the end of 2025 we'll be able to catch up very well, and that, our experience, our result will be quite comparable.
Sorry, just I won't take up too much time, but just on that second question, would you consider consolidation or a new roaming group with Telefónica, if the spectrum path, for example, doesn't suit what you want from it? Thank you.
We're still waiting for the, standing by for the answer. Sorry. We don't have a Plan B. We decided to build the network. We will build the network. We will not fail due to a lack of passive infrastructure.
That's clear. Thank you.
The next question comes from a representative of UBS.
Hi. It's Polo Tang from UBS. Just have two questions. The first question is, can you comment on what you're seeing in terms of competitive dynamics in terms of both the mobile and broadband markets? How optimistic are you that you can stabilize your broadband base in 2023? My second question is just a follow-up in terms of the Q4 service revenue growth trends. I got part of the explanation in terms of the slowdown in terms of growth. I understood the phasing in terms of international roaming, but I didn't really understand the explanation in terms of what was happening in terms of accounting and IFRS effects. Could you just clarify what the accounting impact was? Thank you.
Yeah. It's okay. Yes, concerning the competitive environment, our view is that right now everybody is reducing their aggressiveness, trying to increase prices. With our offers we will be guided by that as well. We think that it will have a positive impact on the service revenue. We can't see right now our competitors to be more aggressive. On contrary, they're actually decreasing the aggressiveness of the market. Now, concerning the service revenues, my comments on IFRS. What I meant was that in the portfolio approach, a look is taken annually at the question of how parameters such as contracts, terms and or durations and cancellation behavior of customers are evaluated.
More than EUR 6 million sounds like a lot, but that is actually based on more than EUR 3 billion worth of service revenues that we have here. The effect is that this reduction that we see for 2023 will mean that the revenues in the following periods in 2023 will increase. The depreciation period of the contract assets are increased, and that's the effect for Q4 here that we see.
Can I just follow up in terms of the broadband base? Declines were easing somewhat in Q4. How optimistic are you that you can stabilize it in 2023? Can you maybe just give your view on why the broadband base was declining? Was it purely a function of the new telecoms law, or was it something else? Thanks.
Yeah. We have, well, we have seen growth for fast internet access points based on VDSL or FTTH. That's how we make them available. Where we lose customers, where we have high churn is in the old ADSL business. Right now, we still have 1.1 million ADSL connection points among our customer base. At the end of last year, we made a contract with Deutsche Telekom, which allows us to move those 1.1 million customers to a VDSL infrastructure with similar speeds but much higher stability.
These migrations are to happen in 2023 and 2024, leaving only 300,000 customers in the old ADSL business by the end of 2024. We expect, of course, with much fewer customers, we expect the churn from ADSL to decrease, so we should see better figures. For this year, we continue to expect a slight decrease of DSL of the overall decrease of the broadband customer base, particularly driven by ADSL. While with VDSL and FTTH, we will continue to grow.
Thanks.
The next question comes from Credit Suisse, Jakub Stancel.
Hi. Thanks for taking the questions. A couple questions from me. Firstly, and this may have been said already, maybe lost in the translation, but can you just give us an update on where are you actually currently in terms of site deployment, given it's March? I think you've only given a number for December and what you're targeting for the end of the year. Secondly, can you maybe give us a bit of guidance just on some of the items below EBITDA minus CapEx, just to help us understand, you know, just to model the free cash flow, so to expect further drags on working capital, for example.
Then just one point of clarification, and I think this is probably a little bit lost in the translation. Did you say you expected positive or break-even EBIT in 2025 on the mobile network segment? Can you maybe just clarify what exactly was the guidance for that segment for 2025? Thank you.
Jakob, can you please repeat your first question? Understand there was an interruption in the line.
Sorry. Yes. The first question, I was just asking how many sites do you currently have, deployed? You had 5,000 in December. You're targeting 1,200 by the end of this year. Where are you currently? Just to give us a little bit of sense of where, or how quickly the deployment is ramping up.
Yeah, if we could. I would like to jump back to our presentation and speak about page 14. That's what I would like to look at. Now we have 94 sites that have been handed over to us. Of these 94 sites, 14 are active. Sorry, it's page 12 that I'm talking about. We have it on page 12. Sorry, I got that wrong. We show this on page 12, that we have 94 sites. 94 times passive infrastructure of which 14 are connected to our far edge data centers via optic fiber. We keep adding new sites and connecting new sites as well. Concerning the EBITDA of the mobile phone network, it won't make a positive contribution in 2024. That will take a longer time.
What I said is that by the end of 2025, when our customers have been migrated, our savings from voice and international roaming will be so high that the network operation, i.e. our staff and the core data centers, the edge data centers, and the far edge, data centers
Will be capable of being refinanced at an EBIT basis completely. In building antenna, we will have less than full capacity utilization. Every antenna will mean a deficit because we still lack the low-band capability, and we don't have the infrastructure in place, so we need more roaming. When we have low-band frequencies, the antenna will create more gigabytes, thus saving us national roaming than we have today with the mid- and high-band frequencies. In the long term, the more antennas we build and the denser our network is, the more profitable each individual antenna will be.
In the long term, they will pay for themselves, but it will take a few years before we have all the low-band frequencies until we have built so many clusters, so many antennas that we can actually yield or reap enough revenue to refinance our CapEx. That is our business approach, and that's why in 2023 or 2024, we won't be EBITDA positive.
Yeah, through the slide.
Concerning the second question, the EBITDA guidance, I could share some more information. The EBITDA that we guiding for the current year is made up of the access, the end customer business, where we expect a stable growth with 2% service turnover and 4% EBITDA on to EUR 775 million. The focus here being just like last year on value, on high-value contracts, in existing customers and new customers. This business is to develop stably. In the 1&1 mobile network, we see a drop in EBITDA from -EUR 52 million to -EUR 100 and a bit million. This depends on the start of operation of the network, the operation costs, the data centers, the access of the data centers, the electricity and cost for the migration of the customers.
This is the background for that increase of minus EUR 57 million to a minus EUR 120 million. As far as the working capital is concerned, we have the free cash flow on the level to be expected as we've had it in 2022. We're going to have a little higher CapEx than in 2022, but the net pay of the operational is going to be a bit higher. We had a repurchase of contingents from the Deutsche Telekom contract in 2022 that gave expenses, which is not going to repeat.
In 2022, we also had a decrease of liabilities from supplies and other liabilities, which we are not going to see again in 2023. These two affects higher CapEx, and on the other side, higher net payments from the operation, is going to be balanced so that the free cash flow is going to end up at the same level, roughly.
Great. just to belabor this point around the 94. That's an end of March figure. You've done about 90,000 new sites in Q1. Is that correct?
Yes, that's the current situation. We have, 94 antenna sites which we can use already or we're just starting the operation.
Okay. Thank you very much.
Next question, Usman Ghazi, Berenberg.
Hi. Thank you for the opportunity. It's Usman from Berenberg. Got actually two or three questions, please. The first one was just on the core business, so excluding the 5G startup cost. The guidance is for service revenue growth of 2%, EBITDA growth of 4%. That does imply better operating operation gearing relative to 2022 when service revenue growth was 2%, but EBITDA growth was -1%. Can you just walk us through the sources of the improved kind of drop through of service revenues to EBITDA in 2023 versus 22, please? The second question was on the 5G startup costs, so the EUR 120 million that you've indicated.
I guess I'm struggling a bit to get to EUR 120 million because, you know, like, as you've indicated, you know, you're building 1,200 or you're taking control of 1,200 sites, mainly back-end loaded this year. I'm wondering whether there's any front loading of the Versatel payments agreed between 1&1 and Versatel. Where, you know, the agreement is for EUR 170 million in cumulative payments to 2025. Is there any, I mean, are those payments front-loaded in, you know, 2023, at all? Are those payments in line with, you know, the deployment of the antennas? Yeah, I'll leave it at that for now. Thank you.
Yeah, Ushdev Singh.
With respect to the development result, increase by 4%, 2% in service, we have two facts here. On one side, as I said, we're going to have a stronger focus or even stronger focus on the value of the contracts, and that means the hardware subsidiaries, we're going to take out some aggressivity and offer less subsidiaries. The second point is that in our contracts, we can generate improvements with customers which are on an MVNO contract now. Vodafone or E-Plus are step-by-step going to be migrated into the MVA contract, which is a purchase price benefit. For us, we've had that in 2022 compared to 2021 already, and that is going to carry on in 2023. That's a stable operative result development in 2023.
As far as EUR 120 million is concerned, that has no front-loaded or down payment included in this. We pay our sister company, Versatel, only what we actually take into operation and services that we use. The only thing, and I've mentioned this before, which is put on top is the expenses for the migration of the customers, which in 2023 occur for the first time.
Would you be willing to say, you know, how many customers you expect to migrate this year or, you know, what that cost actually is or, you know, what this, what the sack is per customer in terms of migrating, migration?
The.
The expenses that we see this year have less to do with the number of customers, just rather the implementation costs for all processes and interfaces to be established so that the customers can be migrated. The variable costs for a migration call center, eMBB costs, for example, are costs that will mainly emerge in 2024, 2025, because only then a relevant number of customers will be migrated. this year, the number of customers in the Q4 is still overseeable.
Thank you.
The next question comes from Anna-Maria Skoglund from Goldman Sachs.
Hi. Thanks for taking my question. I have one, please. Just wanted to follow up on the market environment and pricing. With mobile pricing has begun to rise across the German market, do you expect this to continue? Do you believe you could take prices higher in 2023 or do you plan to focus on maximizing volume gains? Thank you.
We see a clear indication in the market that the competitors price, increasing the prices and the offers, I would put it that way, that the offers are gonna be changed in a way that there's less discounts in the initial months, that will then lead to an increase in prices or rather if more, performance is bundled into a contract, there's gonna be higher fees as well. We are seeing this with nearly all our competitors. As I have mentioned before, we will also look to it that we are going to have a stronger focus on profitability by subsidizing less hardware in the beginning of a package or as a start of a package.
The next question is an additional question from UBS. You have the floor.
Yeah, thanks for the, allowing the follow-up. It's really just about just like the cost in terms of the network build, because you've laid out a target of 12,000-13,000 sites longer term. Is there any color that you can give in terms of the OpEx per cell site or the CapEx per cell site? Alternatively, can you comment on the trajectory or quantum of CapEx for next year, 2024? Thanks.
At this stage, we don't want to give exact OpEx or CapEx figures, which is also due to the fact that it depends on whether we have the site built ourselves or whether we can lease it. If we have to build it ourselves, we have an increased CapEx because we will have to pay for the antenna aerial carrier. If we lease the site, then the carrier doesn't have to be financed and CapEx will be the antenna only. Now that we don't have the exact mix and we don't know about the delivery timeliness of a supplier, we can't speak about this yet.
Is it possible to make a comment?
Operator, here's investor relations again. I think that was the final question, so we may now come to an end. in German.
Thank you very much for your attention. As always, we are ready for more discussions. I hand back to the operator. I wish you an interesting call later on from 2:00 P.M. with United Internet. Stay well and see you soon.
Ladies and gentlemen, the conference is over now. Thank you for participating. Have a nice day.