United Internet

Good morning, ladies and gentlemen. Good afternoon. I would like to give you a warm welcome on behalf of 1&1, our CEO, Mr. Dammermuth, and our CFO, Mr. Hoon. The Board will continue to present the company development of 2023 to you within the scope of this investors' meeting. conference. They will tell you a bit more about the forecast for 2024 and some updates on the network expansion. Afterwards, as always, the board is available to answer your questions. Thank you. And now I'd like to hand over to Mr. Dommermuth. Thank you, Mr. Kyle. Hello, ladies and gentlemen. Welcome to today's conference. I'll begin with the company development and then I'd go into a little more details on our biggest and most important investment into the one and one network. Mr. Hoon will then follow with the financial key figures. You know our company. We are dealing with broadband connections and with mobile Internet connections. We currently have 4.01 million broadband connections based on VDSL and FTTH. We produce it mostly with our sister company, 101 Versatel, using city carriers or the Deutsche Telekom for the connection. Besides that, we have 12.25 million mobile contracts, so customers. We have launched the first open brand in Europe, which is fully virtualized. We have a broad market coverage with a number of different brands. We have our primary brand, 1&1, and then, of course, our co-brands, Gmx and Web.de with the co-brands, and then discount brands that we've taken over through acquisitions. Last year, customer contracts have increased by 3% to 16.26 million. This was mostly driven by the mobile internet growth with 570,000 new contracts for broadband lines. We've had decreases in our customer base in the first three quarters of 2023. It has stabilized in the last quarter of 2023. We have now 4.01 broadband lines, which is 90,000 less than in the previous year. Why has it stabilized? We have made progress in our migration from ADSL to VDSL contracts. ADSL is the old technology, has a higher termination pressure than VDSL and glass fiber connections. We have made good progress on this last year. The fourth quarter with 150,000 mobile contracts and the stable broadband ownership was a good finish to the year. The revenue has increased by 3.4% last year. Service revenue increased by 2.1% to 3 billion 243 million. The business with hardware, specifically with smartphones, has increased by 8.3%. But you all know that it's a quite low margin business. Good fourth quarter here as well, 4.5% service revenue increase. And this means we're on track for the new year and for the growth that we have set for ourselves for 2024. EBITDA in the access segment, so in our old business model, so to say, has increased by 5.4%, raised to 786 million. In the one-on-one mobile network segment, where we're operating our own network, we had startup costs of 132 million euros, 80 million more than in the previous year. We're billing it internally that services rendered in the 101 mobile network segment are billed to the access segment. And these are services that we usually would purchase in the access segment. This allows us to see when we cover our costs with our network connections and when it is more profitable than purchasing external services from external network providers. In sum, we're in line with our planning. For CapEx, we're looking at 295.6 million for last year. The major share was for the expansion of the network. And all of that is in line with our expectations. You can see the main KPIs here, 3% more customer contracts, 3.4% more revenue, of which 2.1% in the service revenue area. EBITDA decreased by 5.7%. On the one hand, we've seen growth in the access segment with 4.5%, and on the other hand, we've had higher expenses in the network segment. with 132.4 of course that has an impact on ebit as well we've had ebitda that was lower by 40 million and of course that will have an impact on ebit and ebit also includes depreciations that will be that will be um explained by mr hoon later on Dividend proposal to the Annual General Meeting has remained the same, €0.05 per voting share, which is the minimum dividend that we have to pay according to Section 254 Shareholder Act. Overview on our big investment, the Open RAN network on December 8th, 2023. We have given access to mobile services, so it's fully operational right now. Our network is fully operational. You know the architecture of our network. We're planning to operate four core data centers, 24 decentralized data centers that are positioned around the core data centers, and over 500 regional edge data centers. These regional data centers will be connected to antennas, and one antenna is never further away than 10 kilometers from a regional data center. We're planning more than 12,000 antennas in order to cover as many households as possible. That's our target to achieve by 2030. To give you an idea of where we're at right now, of the four core data centers we're currently operating to, the other two will be operational in a couple of days. The 24 decentralized data centers are all operational. If you look at the regional data centers, we are currently operating more than 100. I will get to the antenna locations in a minute. What does our network look like? What does the Open RAN technology look like? It's mainly three factors. First of all, it's an open system. We have roughly 80 partner companies that support us with or have supported us when it came to the expansion of the network. We have standardized interfaces that allows us to combine software and hardware, and that makes us independent of the dominant manufacturers that usually provide complete networks or large parts of a network. The second factor is that we're ready for real time applications. The far edge data centers and the regional data centers, we cannot only control the antennas, we also have space for application computers. If we need to operate applications from these data centers we have a speed of three milliseconds and only 2.5 milliseconds are actually required for a glass fiber connection so we have a good speed to get the connection where it needs to be And the last aspect is the low electricity consumption. We save roughly 10 to 30% of our electricity. How do we know that? We've measured our electricity consumption in January and then we compared this to traditional networks. There was a study that was conducted on this and our result was confirmed by this study. Looking at antenna locations, we're making good progress. On the left-hand side, you can see the antenna locations that are available to us. By the end of 2023, it was 1,062. At the end of this quarter, it will be roughly 1,350. These locations are passive infrastructure. It's co-location on an existing location or a completely new location. and then we need to install antennas in these locations and we need to connect those to the fiber optic. 243 locations were equipped with base stations by the end of last year. By the end of this quarter we will have equipped 600 locations with antennas. Of these 600 locations, 200 locations will already have been connected to glass fiber. So you can see our funnel very well. First of all, we need the location. It can be a rooftop location or a mast location. That's the bottleneck, so to say. And once we have the location, we need to equip it with the right technology and then it needs to be connected to glass fiber. In terms of construction speed, that's a disadvantage of our architecture because we connect all locations to glass fiber. We would be a little bit faster if we did it a bit differently. We could run our location straight away, but now we are connecting it to glass fiber first. This requires some construction works to be done, some construction permits, and of course that extends the duration of the construction time. Also, it depends on a couple of factors on the complexity of the construction and also on how long it takes for us to gain the permit. But once we have taken over the location, so that's the graph on the left hand side again, once we've taken over the location, the implementation is up to us. We can ask subcontractors to install antennas, and we can use different subcontractors when it comes to connecting it to the glass fiber network. So we're quite optimistic. We want to have more than 3,000 locations by the end of the year, and we also want to fulfill the requirements of the Federal Network Agency. we were given some requirements to fulfill by the end of 2022 which we did not manage but we are optimistic to fulfill these requirements by the end of this year that would be it on the course of the business last year and on the status of the network expansion and now i will hand over to my colleague marcus hoon who will tell you more about the financial key figures Hello also from my side. I would now like to continue by presenting the financial key figures of last year to you. I would like to start with the earnings and profit and loss statement. Here you can see our revenue from last year with almost 4.1 billion. Mr. Dammermuth has already mentioned this figure. Cost of sales have increased in um 2023 to 2.937 billion that's a plus of seven four percent this includes the startup costs and depreciation for the one and one mobile network last year they were at 166.8 million after 40.9 million in 2022. gross profit is has therefore increased from 1.229 billion to 1.159 billion, mainly due to the additional costs of the network expansion. We have also included the gross profit without this um these costs where you can see that is increased from 1.7 billion which means a plus of 4.4 percent so if 1.27 billion to 1.3 point 1.3 to 5 million billion. We've had higher marketing expenses, among other things for online marketing. We have administration costs amounting to 115.6 million, which is also slightly increased. This is due to expenses for the one on one mobile network. which are built partially also under administration costs. Other operating income and expenses were at plus €31 million, slight decrease compared to 2022. The impairment losses on receivables and contract assets were at €105.4 million, also a slight decrease compared to 2022. This decrease is partially due to lower corrections to payables that amounting to 6 to 7 million euros. On the other hand, we have higher corrections on assets due to a higher contract volume because we are currently further building up billing according to IFRS. Profit was therefore at 455.8 million euros compared to 534.9 in 2022. Tax expense. financial result was also positive 9.1 million compared to minus 3.4 million. Profit before taxes was at 465.9 million euros. Previously it had been at 531.5 million euros. So here you can see the effects from the ramp up costs of the mobile network. tax expense follow the profit before taxes they have decreased from 164.2 to 149.9 million euros therefore this brings us to consolidated result of 315 million euros for the financial year 2023 I will continue with the balance sheet. Balance sheet has significantly increased from 7.257 billion to 7.740 billion. There are three different reasons. If you look at short term assets. We have a decrease of roughly 70 million euros due to higher trade payables and higher inventories. Long-term assets have increased by roughly 420 million euros. Two reasons for that. On the one hand, our prepayments to Deutsche Telekom for the contingents from the VDSL contract last year this was billed at 210 million euros in this balance sheet and then also our investment in assets for the one and one mobile networks they are also included in the long-term assets and have increased by roughly 200 million euros short-term liabilities have also increased compared to the previous year roughly amounting to 60 million euros Long term liabilities are roughly at the same level as in the previous year, and due to the positive result, equity has increased by roughly 300 million euros. will continue with cash flow net inflow from operating activities were at 225.6 million euros so we were above the 182.9 million euros from 2022 they include 488.8 million cash flow from operating activities a negative effect from the change in trade receivables so increase in trade receivables amounting to 88.3 million euros What I would like to mention is roughly half of that are receivables that are no older than five days and that have simply been billed after the date this balance sheet was created. so by the end of December 2023 they have contributed to this increase in receivables then we have changes in inventories amounting to minus 57.6 million euros here by the end of last year Apple and Samsung we have purchased more from Apple and Samsung therefore and we have sold these assets by the end of by the beginning at the beginning of this year therefore this inventory has further decreased changes in receivables from and liabilities to related parties had an effect of 73.3 million negative effect of 400 249 million euros from change in accrued expenses these are the prepayments to deutsche telecom from the contingent contract that I already mentioned. Other items from working capital have an effect of plus 58.4 million euros. Cash flow from investment activities was at 125.2 million euros. compared to 95.1 million euros in 2022. It includes 295 million capex. This includes mostly the expansion of the mobile network. 155 million euros were Gained from the investment of free cash with United Internet were taken out from that investment and we had interest received amounting to 15.7 million euros specifically from the cash investment at United Internet. Then we get to cash flow from financing activities. That was at minus 101.8 million euros after minus 87.7 million euros in 2022. It includes 12.1 million euros payment in connection with leases, 8.8 million in dividend payments, 61.3 million repayment of liabilities from the 5G spectrum, 14.9 million other payments of an interest nature and 4.7 million interest payments from leases. This brings us to a free cash flow of minus 70.1 million euros by the end of the financial year. We're roughly at the same level as in 2022, where we were at minus 62.1 million euros. On this slide, you can see the bridge from the EBITDA to the free cash flow. We're starting at 653.8 million EBITDA. Then we have receivables and other assets amounting to minus 88.3 million euros. Furthermore, inventories amounting to minus 57.6, deferred expenses at 249, minus 249, receivables and repayables from related parties with 73.3, other working capital with 30 million euros, then tax payments at minus 136.7 million euros and capex amounting to minus 295.6 million euros this then brings us to the minus 70.1 million euros of free cash flow let's now continue with the outlook for the financial year of 2024. Service revenue, here we expect revenue growth of 4% to roughly €3.37 billion after €3.24 billion in 2023. EBITDA is expected to grow by 10% to roughly €720 million after €653.8 million in 2023. of these 10 percent we want to gain roughly 880 million euros from the access segment which would be a plus of 12 percent and we expect a minus of 160 million euro startup costs from the mobile network segment this would be an increase or let's say higher expenses or lower result of roughly 20 million euros compared to 2023. And then cash capex will be at roughly 380 million euros after 295.6 million euros in 2023. This is mainly driven also by the mobile network expansion. Maybe some remarks on the development of the contract customers or customer contracts. We expect a growth in customer contracts on the same level as last year. Due to the migration into our own network, we will, however, lose some customers that are hard to quantify. We're expecting something of 200 to 300,000 contracts. So our net growth minus this effect of the customer migration will roughly be at 200 to 300,000 contracts in 2024. Thank you for your attention. And this now brings me to answering your questions. Thank you very much. Let's start on the left hand side Bank of America Titus and then Societe Generale after that, please. Thank you, first of all, for the presentation, and thank you for the opportunity to ask questions. First question, in regard to the 200,000 to 300,000 additional losses in contracts, could you elaborate on that a little bit further maybe? Will these be high quality contracts and what effect will that have on the revenue and what will be the reason for those connections to be switched off for us to lose these or for you to lose these contracts? Second question on white spots. Last week, in the last two weeks, the German press has reported on this issue. Also, in regard to the payments of the spectrum costs, could you please give a statement on that? What is your perspective on that? What should we think about this issue? Thank you. On your question on the contracts that will be lost within the context of the customer migration, it is difficult for us to estimate the size, the scope of this effect. But of course, we have contracts in our inventory that have a relatively low usage. where we will have the situation that the customer decides to no longer use this contract within the context of this migration, or we will no longer be able to reach the customer. This will mean there is a possibility of extraordinary contract terminations. We see a relatively low impact on our result and on our revenue because those are rather old and less valuable contracts. I've prepared something for you. I'm quite proud because I have a backup of this presentation. I said, well, if somebody asks, I have something to actually show them. Because we have, of course, read this article that you're referring to. And it makes sense to shed a bit of light on this. First of all, 1&1 was not a participant in the 2018 Mobile Summit. Nevertheless, we committed to contributing the expansion of the network in white spots. And in return, we promised to invest the interest advantage that we've gotten from the deferral. We were at the time back then where we had rather low interest, 0 point something percent interest rate. So you can see that this amount was not large, but it was a significant amount. This amount would have meant that we could have constructed roughly 410 allocations. but this 400 based on the assumption that we would have mainly rooftop locations. Rooftop locations only costs a small percentage of mast size. you have antenna heights of 10 meters, sometimes higher without needing a specific permit. A mass location is sometimes 40 meters, more than 40 meters high, consists of concrete and therefore needs more complex permits to build. Since we don't have any low band frequencies and because it is here about connecting white spots to the network, so about connecting rural areas to the network, it was said, well, you construct the location and then you provide it to Telecom, to Telefonica or to Vodafone. then they will use this location we simply construct the passive infrastructure therefore we're booking these costs as interests i think that's right right mr hoon is that correct yes the costs for these locations are part of our interest results because they are an equivalent to interests Why did it take longer than planned? Well, Telekom, Vodafone, Telefonica has almost exclusively chosen mass locations, only a handful of rooftop locations. Therefore, the price per location has become a lot more expensive and that has turned 400 locations into roughly 130 locations that you get at the same price. we have a fixed investment sum. We didn't look at locations. We looked at a set amount in euros. And if one location becomes more expensive, then we can construct fewer locations. It has taken us until Q1 2021 for us to be clear on where these locations are to be built. And now all of these locations have specific requirements. Because if they didn't, the others would have already constructed them a long time ago and wouldn't have to save them for us. But that was the deal that we struck. Therefore, we have struggled with delays, which are completely usual when you deal with these white spots. I have here copied a text from the Telecom because Telecom is usually trying to make it well known that we're lagging behind when it comes to these white spots. And it's a text from the end of 2023 that they've released within the scope of the Federal Network Agency consultation. You can read it on their website. And they wrote that in rural areas, the planning processes usually take at least four years already when under optimal conditions and of course we can make the impossible possible and we can do it in four months if telecom says they usually can only do it in four years. Therefore these mass locations are very difficult to build. Where are we at now? 130 locations will be able to be built with the investments that we have available. of these locations 48 have already been finished that's a dark blue column that you can see on the slide further 50 are currently under construction and will be finished in the next upcoming months For 32 locations, we're still waiting on the building permit. So we have already, the plot is already available. We have already leased the plot, but the permit processes are highly complex and these processes are still ongoing for 32 locations. Once this is complete, the location can be realized or the construction can be realized within roughly six months therefore we think that we're actually doing quite well when it comes to this expansion because 400 rooftop locations or almost 400 rooftop locations has turned into 130 mass locations and we need highly complex permits um procedures for that. There are also some environmental factors playing into that and all of these permits simply take a lot of time to be granted. Now you asked what does that mean for us? And I'd like to tell you it doesn't mean anything for us because all of this has already been put in contracts. In our contract back in the day, we made sure that delays that we don't have any control over don't mean that we cannot fulfill the planned timing. All of this is reported to the responsible ministry with an exact list of all locations with a status per location. There's complete transparency in this regard.

Thank you very much for that question again. The slides had been prepared for that. Thank you very much. Three short questions. On Rakuten there was press coverage. Rakuten is not in a general entrepreneur anymore and that was risk management issue. Could you elaborate on that? How did your relationship change in the meantime? Second question on the network structure. In the past you often mentioned that the structure you have allows you to differentiate yourself from other providers in the market through your product mainly. And I think that probably applies to the B2B segments. Within what time do you think you'll be able to convince first customers of this? There are two follow up questions, of course. because the other network providers will move into the same direction and one-on-one has been delayed in their network expansion. So then the question is when this will turn into a profit for your business. And then third question, you had higher startup costs compared to the guidance for 2023. And lower capex, on the other hand, could you explain why this differs from your original forecast? Okay, let's start with Rakuten and our business relationship. It hasn't changed at all. staff in our press, someone from our press office felt the need to edit a press release where it said that Rakuten is our general contractor and that something had changed. And then some media outlet decided to put that into an article saying that Rakuten is not our general contractor any longer. And that's not true. There are no differences or changes in our contracts at all. Rakuten remains to be in charge, is in charge of building the network for passive infrastructure. We are responsible for high mass, for connections, for the provision of data centers. We are in charge. But orchestrating our partners and their working together, that's up to Rakuten. The individual services are always part of our ownership that's why we have the capex and depreciations but until the location is built and the we are planning to put into operation data centers 3 and 4 and all that is up to Rakuten. They make sure that these data centers work and once they are in operation they are ours. So we benefit from Rakuten's experience of the past years in the open RAN area and only with that we were able to operate at such speed. Then you mentioned other mobile network providers who are moving in to the same direction. I have not actually seen that. I haven't seen other mobile network providers building data centers 10 kilometers away from their antennas. That's news to me. Yes, they are also doing Open RAN, and that's definitely a future technology, I agree. But differentiation in real-time applications is not something I have seen in Germany so far. So I think we are at the forefront here. which doesn't mean that you could also use applications at other data centers. I think Vodafone has four data centers in Germany currently. You probably know that better than I do. So where do we want to differentiate? That's definitely B2B, or you said that's definitely B2B, but I disagree. We will differentiate when it comes to consumers because we will have DNS servers in the regional data centers or also because we'll have regional servers data centers where we can cache content. Contents that are opened regularly at the same location. We can cache them there so they don't have to travel through the internet anymore. Then a video on demand. There are providers that have a huge library but The consumer will not open all of them at the same time. They will have some popular videos they use all the time. So we are thinking about opening these from the regional data centers. So these are just some of the ideas that we have. Then you asked about our timeline. And I don't want to upsell here. Our network does have the opportunities. But an old colleague of mine always said that it's about standing, walking and then running. So it's good that we're talking about these issues today because a few months ago we were discussing whether this could even work. with hundreds of thousands of millions of customers. How could that work? Now we have 500,000 additional customers. We have the first existing customers using the system and we're happy that it's working. So for now, I would make sure to put the data centers 3 and 4 into operation and then make sure the migration works. We want to increased to 50,000 customers per day. That's our maximum capacity. There is a company that will take care of the phone numbers, importing the phone numbers too, and the maximum capacity for that is 50,000 per day. But that's quite an enormous amount because that's a million a month. So these are our current challenges. So as I said, putting into operation the core data centers, making sure the migration processes work. And it's not that easy because some customers have an eSIM card, others have a normal SIM card, then they have different smartphones. Others may have smartphones that are lacking the newest software update. So there are many issues making this very difficult. So this is our focus for now. And then after that, we'll take the next step. Let's come back to the last question, the expenses for the mobile network that were higher than expected last year. It's mainly expenses within the tests that we had to do that were higher, device testing in particular. We may have underestimated that in the beginning, but we had a high approach to quality here. we wanted to identify all end devices in our customer base and test them and certify them. It's not possible for all legacy devices, of course, but that was six to seven thousand end devices that we tested over the past months with our network in order to be sure that they would all be working and also to know what legacy systems wouldn't be working so that the customer can be informed. That was the main driver of the higher expenses in the last financial year. Let me elaborate on that and give you a sales pitch. if you have an old phone it doesn't mean you can't use our new network we will give you a new configuration for your phone and then you can still use it and i think only four percent of the phones will not work we can we can update 96 of all phones online and for the last four percent that's long tail our core data center will need to help customers with manual configuration of the settings and if it doesn't work at all we'll provide a new phone to the customer. So we're doing everything we can. Rakuten, when it came to the launch of the network, they didn't have an Apple certification. It took a long time for Rakuten to be able to use Apple phones in their network. We had our Apple certification from the first day. We had six and a half thousand tests that are expected by Apple before you receive your certification. Or Dish in the USA, I don't want to give you any wrong numbers, but I think Dish started with four phones. And we are starting with our existing customer base, all older legacy phones that our customers have. And we are confident that we can completely migrate them to our new system without losing any essential data. And of course, that cost high expenses in the last year. Next question will go to Politan. But I had another question. We are asked whether for the fine by the federal network agency that we are expecting, we made any provisions, and if so, how high. Mr. Demmermuth said in the last quarter that we had provisions and that they will not impact our forecast. We didn't give any detailed number in the past. Yes, we do have a provision from 2023 and we think that it will be sufficient, but please understand that we will not give you the exact amount here today.

Yes, just have three questions. Can you give an update on the Vodafone national roaming agreement? So when can we expect a final agreement and are you on track to have it operational by October 2024? Second question is, what is your expectation for when the BNets will come to a decision on the allocation of the 800 MHz spectrum? And what happens if you do not get 800 MHz spectrum? Third question is on the recent appointment of Pascal Greider, the former CEO of SALT has joined one-on-one. Can you clarify what his role at the company will be? Thanks.

Thank you very much for your questions. We do have a final agreement with Vodafone. We made an agreement last August, and that's legally binding, but we need investment for that contract. We need annexes because we are speaking of 1,000 pages and more, so we will need annexes for this contract, and we are still working on that. And within that scope, we also said, well, we discussed this contract within a few days, so we may have to make some amendments to this contract. But generally speaking, it's a legally binding contract and we could start on this basis. But as always in life, you can make a contract that has 20 pages or a contract that has a thousand pages that will provide more clarity for day-to-day business. but it doesn't have any other conditions or terms. And you'll also see that in the publication by Vodafone where they always say that they have a contract with us that will be valid from 2026 and cash effective. Then you asked whether we can start with Vodafone National roaming in October. I think we can even start earlier, but I can't give you an exact date. The contract assumes 1st of July, 1st of October at the latest and I can say that we are making good progress and we will then start step by step because Vodafone will have to provide the capacity within its network and we have a plan, a step by step plan on how to migrate the customers over 12 months. Then you asked about the Federal Network Agency and the decision about low-band frequencies. I can't give you an answer to that. I can only tell you what I'm hearing. And I'm hearing that there will be a decision in the summer, but I don't know if that's true. That's just what we are hearing in the market. Then you asked what we are doing if we don't receive any low-band frequencies. And I'd say I'd expect that we will receive them. But of course there could be an auction and someone may have a better bid than we do. Then we will have to buy more roaming than we do today. Our network is working as operational today without low band frequencies. So where we don't have coverage for our frequencies, we will have to buy more roaming. And when we are lacking frequencies, we will have to replace them with roaming services. on the one hand that's expensive because roaming costs money on the other hand we will save the money for purchasing frequencies so you could now do the math using your spreadsheet and try to calculate when it's cheaper to remain at roaming and when it's cheaper to switch to the broadband frequencies. So if it's 100 million, for example, for roaming the frequencies will be less expensive and the other way around. So we will have to evaluate or assess that if we do not receive frequencies but for now we would expect that we will receive them because it has to be free from discrimination. And then our late launch, the question is whether our late launch might lead to us not receiving any frequencies but there are rules and the law cover for what happens when you do not fulfill your requirements but and then there is a fine for example if you park your car incorrectly then there is a clear punishment for that the state can't just choose what how to punish you and that's the same here so we know what the fine would be and we expect that fine, and that's the correct punishment for not keeping your deadlines, but the law doesn't proceed to not give you any frequencies as a punishment. Then you asked about Mr. Grieger. He is responsible for product management and new customers at one-on-one. Our colleague who did that before is now working on our mobile network. In the board, there are Mr. Huynh and I, and then we have Mr. Haver, our chief operating officer. And over the last years I have been in charge of mobile network and Mr Hava was doing product management and new customers. And now. We are at a point where the mobile network is working, so we need good good engineering. We need scaling and he's the best person to do that. So we decided to give marketing for new customers and product management to Mr Greger.

Hi, it's Ben Rickett from New Street. I had two questions, please. Firstly, on the number of sites, so you're saying 3,000 passive sites at the end of this year. Can you say how many of those will be active sites? And can you give us an idea of what proportion of your traffic you could then carry with those active sites? And then second question on CapEx. So you're saying 300 million of CapEx this year. Can you give us a sense for what the shape of the CapEx then looks beyond this year? So I think consensus is expecting another 20% increase in 2025. and then CapEx falls to around 250 million ultimately in 2030. Does that shape look sensible? Do those numbers look sensible? Thank you.

By the end of the year, we want to have 3,000 passive locations. And giving us as much enough infrastructure to have 1000 active sites. Our requirement by the Federal Network Agency from 2022. So our plan is to have these active this year. So 1000 locations. as a minimum 3000 at the beginning we have the mast ready at the beginning of the funnel and then a thousand at the end of the funnel with fiber optic installed and fully operational but of course it would be better if we had more nevertheless that's what we are looking at right now as far as traffic is concerned that's difficult to assess If you have 12,600 to cover 50% of the households, that's 50% of the traffic, let's say, then a thousand would make up for one twelfth of 50%, so 4%. So if we had migrated all of our customers already and we had low band frequencies available. By the end of the year, not all of our customers will have been migrated. The Federal Network Agency gave us two years for that, 2024 and 2025, and we'll need that time in order to migrate 12.5 million customers, though we will have restrictions. The mass will not be at full capacity. and we will not have full band frequencies by then because they will only be given out in January, on the 1st of January 2026. So even if there is a decision this year, we won't have frequencies by the end of the year. And therefore the traffic share will not be enormous. Our advantage from having our own network is that we can migrate a customer and then it starts with national roaming because for minutes we will only pay the data used and that's an enormous sum for us currently we pay each minute and we produce the speech in national roaming ourselves and then we only buy data for that in roaming and speech doesn't need much width of the broadband, so it will lead to enormous savings. Each customer that we migrate, even if we have fewer mass locations and frequencies, we will save money for each customer through national roaming because we will save on the data. And we will also save because we produce international roaming ourselves. Today we are buying it from Telefonica or Vodafone within our MVNO contracts. In our own network, we will buy international roaming ourselves. So we will have much better conditions that we already negotiated. So the more a customer uses the phone and use it abroad, we will save. Even if they don't use our cells. So that was our calculation. We had the basic costs for the data center infrastructure that we had to cover from savings that we make in international roaming and speech transfer. And operating the antennas needs to be covered by the production of each antenna. So that's our calculation. And now addressing your capex question in 2024, it's the same as last year, mainly driven by investments in core and through Rakuten. And there will be one bigger payment next year and it will be much lower next year. So for 2025, capex expectation is to be lower than 2024. There are, of course, a number of sites that we need to build ourselves and then we may sell them. If that works out next year, we will have a lower capex total than in 2024. Adam.

I think Mr. Dominic may just have asked in part my first question, but I was hoping you could talk a little bit more about the breakdown of the cost tailwinds you're going to see within that access segment going forward. Obviously, you're guiding to 12% growth. If you can be a bit more specific about the scale of the benefits you're getting and also to the extent that can help us think about how that progresses into 2025, that would be useful. along those lines, it would also be helpful if you can say anything about how DNA is going to evolve as more of the network comes online. And then on customers, I suppose as you're going to be transferring the customer base over the course of this year and into next year as well, presumably there will be that kind of headwind from kind of nudging customers into 2025 as well. So there's 200,000 to 300,000 customers that you're talking about in 2024. To an extent, that's going to carry over as well until you finish the migration. Is that the right way to think about it? If I may.

Let's start with your last question or comment. By the end of the migration, we will have losses within the migration process that's planned for the end of 2025, so for 2024 and 2025, and we'll see in these years how much it will be. Mr. Huynh talked about that earlier. We don't know the exact number yet. We did several tests with different customer groups, customer segments. We had higher and lower chance we are optimizing our processes but we just wanted to tell you about this we have 12.5 million customers to migrate so we will have a few that are we call them sleepers who have always wanted to terminate their contract with us anyways, for example, and that we will then basically wake up with the letters we sent to them and we'll have to take that into consideration. So we will do that in this and the next year and will be done by 2026. When it comes to the development of the results in the access segment, I can say the following. It's two things. We already communicated last year that we focused on well high quality marketing to our existing customers we didn't just want to chase new customers but we tried to improve quality within our existing customer base and we were very successful with that we also try to keep our customer base stable and we try to optimize operations and technology. And we made good progress last year. We had very successful projects last year, which led to an increase of 12% in the result for the excess segment. In 2025, we'd expect a lower increase because not all the effects will be repeated, but we still expect an increase.

Maybe a remark from investor relations. This forecast of twelve percent is currently comparable to seven and a half point eight percent of consensus. Why am I saying that? Because if we look at the share price, this has decreased this morning by six percent and the feedback was given that we should explain to the capital market how our forecast stands compared to the consensus. The consensus does not believe us that we will achieve the 4% service increase. They are expecting it to be 3.1%. They don't believe that we will have a 12% increase in access. They estimate a 7.5%. So that's just the addition as an addition to many of the questions that we've received this morning. Thank you. Just on the big picture, and please don't get me wrong, please understand this is a genuine question and not as a confrontation. You would have had the choice to continue the business the way you've done so for decades with negotiations with network providers on purchasing, on reselling, but you chose instead to migrate it into your own network operation. now of course that has led to battles on many different fronts would you say that the situation the way it has developed in this project since you've made that decision has maybe put the original decision into question or would you say no this is simply how things go in this business with these matters And at the end of the day, we will reach the treasure trove at the end of the rainbow. So I'm just trying to gain a big picture. What do you how do you assess this decision right now that you made to enter your own network operation? I don't think we had the choice to continue business as usual. Yes, we could have become a second free net that gets more bonuses and commissions than what they earn in profits. But the freedom we had from our MBNO situation in times where there were four network operators, has that situation has changed where there are only three network providers with the same share each. We wouldn't have had to have that many proceedings with Telefonica. There was a rule in our contract that said we discuss prices twice per year. And then we asked Telefonica, well, can we discuss, can we negotiate prices? And Telefonica simply declined. We never, ever spoke to Telefonica again without lawyers being present after that. Just to give you an idea of the choice and the freedom that we had at the time. And yes, we could have pushed through. We could have powered through. Yes, the contract is, there is a contract still ongoing that we could have extended. So we could have pushed through until the summer of 23 with ever more arbitration proceedings, more expert opinions, more proceedings, conciliation hearings. legal processes and so on and so forth. Yes, we could have pushed through with that. That's true. And I don't want to say that this could not also be an option. I studied law for three semesters, so it could have been something for me. I could have benefited from practical expectations. Engel has already offered a job to me, actually. So if I quit here, I can go to Engel and Folker and work for them. But that's not what I want to do. I'm an entrepreneur. I'm not a lawyer at the end of the day. And therefore, if we wanted to grow sustainable business, because it's now my 36th year, that was not an option anymore. This option would have ended on June 30th, 23. You mentioned the battles that this has led to, and I have a different perspective on this. I am saying that these battles have now ended I don't have any legal proceedings ongoing anymore today. I have one more claim from Telefonica that was brought against us that is still, how old is it now? Five years old with a regional court in Düsseldorf. But that's an old thing. That's an old claim. Other than that, I don't need legal advice anymore in my day to day business, nowhere in our business. And that's a nice thing. Therefore, I have I have actually ended or the all of these battles have actually ended. You're asking for questioning original decision. I don't know why we would. we had some struggles in the very beginning we we we were relying too much on vantage powers we believed in all these sides that should have been delivered and that didn't happen the federal cartel office is still looking into that that was stupid but we are doing better now vantage has already fulfilled some of its delivery obligations last year and they started quite well actually this year at the same time american tower is delivering to us at the same time we're building our own sites and um we have roughly 600 600 700 leases that were started last year for our own rooftop locations mr huden just said it we will have thousands we will build thousands of our own locations we need that because we can not have all the 12,600 sites in co-location, but just like other network providers, we will sell these locations and we will then lease them back. And then in our balance sheet, it will be like a co-location because we don't own the site. Therefore, there's not a big difference, actually. The 800 frequencies, yes, we didn't expect this discussion to take place because when we went to the frequency auction in 2019, if you look at the document, it says that... this was only a first step and that in the second and the third step in twenty twenty six and twenty thirty four we can purchase further frequencies In 2026, the low band frequencies will be launched. This is what this document says. That's the foundation of our business. And that's why I do believe that we will not be discriminated. But of course, I can't promise anything. But yes, it's true. We did not expect this part. We did not expect for us to fight for... for this, for the fact that I mean, there always used to be auctions and now this is switched to extensions and they're saying, well, instead of three or four, we will only choose three. There's a consultation of the Federal Network Agency and you can see all statements of all stakeholders on their website. And now we can guess there are 52 statements and 38 said something on the fourth network provider to receive low band frequencies or not. 35 said that the needs of the fourth provider are to be considered. They're mainly municipal providers, etc. Three stakeholders said one and one should not receive any frequencies and you should extend by eight years instead of five years. I don't really want to make any bets because I think you can think for yourself who made these three statements. Therefore, the battle line is very clear. But for me, that's sort of a last hurray. It's a last chance for us. It's a last opportunity for them to slow us down, pretty much. But as I just said, I'm very optimistic. Therefore, our original decision still stands. And then you said you talked about the treasure trove. That's a nice image. I like that. I will take that up. Because look at the EBITDA, for example, that we had in the Axis segment. In the Axis segment, we rarely have any depreciations. EBITDA is roughly the EBIT. This year, I don't want to give you any wrong figures. I think we're looking at 850, 900 million, something like that, just at the top of my head. That's where we're aiming towards. And I'm saying if we can get there, we can get to 1 billion as well. And if we then also manage to really have a network that operates at the same price as the price that we spend right now for external services, then you will actually see the profit. And then hopefully in a couple of years time, we will all sit here and say it was a hard time, yes, with a low dividend and with a high capex. But at the end of the day, it was sustainable business growth. And then this treasure trove is actually there. So I think this was a very beautiful final remark. But of course, if you still have any questions, feel free to ask them. Usman Gassi, please.

A couple of questions, please. On the savings on voice costs that you mentioned from migrating customers, is there a big proportion of that benefiting the 2024 EBITDA guidance? I'm just trying to square the circle between what consensus expected before December and then your guidance came in, which was much higher than consensus. So is that the main difference? And I'm talking about voice migrations from activating your core network and then saving money on voice that way. The next question I had was the provisions for the fine that you mentioned are already in the 2023 numbers. So has that provision been taken through EBITDA? And my final question was a bit more of a strategic question. I mean, like you said, no other operator is building out such a decentralized architecture with far-edge locations. Now, when I look at what's happening with AI, they're all talking about AI inference driving up the value of these edge locations. So the question was, is it possible for you to run, to architect the network in such a way that when you're building the network now for the edge, you can also run AI inference workloads when that opportunity arrives? Thank you.

Let me start with the last part of your question. Yes, for us, it's evident that we want to have applications of all kinds in these data centers. We always want to have a certain backspace in our data centers to control the antennas because we don't have any intelligence at the antennas. We don't have a gray box standing next to the antenna or under the roof or in the basement. to control the antennas the controlling always happens from the closest data centers that's why we're overcompensating in these data centers and whatever fits into the next or next rack is what we'll be able to use so i don't see any restrictions in terms of applications in these data centers On your first question, to what extent savings or optimizations from our own network are already included in our guidance? Yes. In 2024, first effects are included, for example, savings in the voice area. But in 2024, it's still a small figure. The amount will only grow in 2025 when we have a significant or when we have migrated already a significant number of customers. On your question regarding the fine, the fine has been indicated above the EBITDA, so it has a negative effect on EBITDA, so the provisions for this fine. On this fine or exclusion from the auction, these three statements include a counterargument that you surely know. which is that the federal network agency can include companies when they don't have the respective capabilities or they cannot prove the respective capabilities. I don't know the exact phrasing, but that's kind of the direction of the arguments. And all three argument that this is the case. They argue that this is the reason. They don't say that the reason is the slow expansion, but they try to take that slow expansion and argue that this goes to show that you're not a serious network provider and therefore shouldn't get any frequencies. Can you please elaborate on that? I think that we have proven our capabilities as a network provider. Our network is a lot more complicated and complex. Well, let's give you an example. I cannot just call Huawei and say I have an issue. Can you please solve it? Our network has a far higher complexity and we have more value generation and other network providers. Therefore, I think in terms of our capabilities, we don't have to hide behind anyone. When it comes to modern, capable networks, we have actually won legal battles that allow us to say we're building the most modern network in Germany. But even if we weren't capable, they'd have to initiate formal proceedings to prove that. If you as a network, a federal network agency wants to say this company is incompetent, I want to exclude them from the auction, then you have to initiate formal proceedings. You have to initiate hearings and make a resolution that stands from a legal viewpoint as well. You cannot simply make a statement and put a letterhead on it and then sell this as a fact. Such proceedings has never been initiated and the federal network agency has never expressed anything like that towards us. That doesn't mean that they don't think that maybe they write such an informal letter to me tomorrow, but I at least have never heard about it so far. What I think is quite nice, though, is that our discussion is becoming weaker and weaker in terms of argumentation. If you look at the other company statements, which you've done, you can read there that we don't have any antennas, that we don't have a network, that we're not even willing to build that network, whatever. And you can basically cross all of that out. All of these are false statements. Therefore, I am quite optimistic. Thank you. A lot of questions were already asked, maybe two minor questions. First question. Frequency costs and savings compared to purchasing Roman. If we were to make your range smaller, maybe decreasing to 2.5 billion, what would it look like then? And then question concerning the broadband customers. In Q4, the decrease has stabilized and slowed down. I would be interested in knowing why this decrease has slowed down and the customers that are included in this broadband area. How many are actually also getting IPTV? How many customers do you have overall in the IPTV segment? I don't know IPTV from the top of my head. We'd have to look that up, but I can tell you it's less than half a million, but I don't know the exact figure. But we can definitely look that up for you. On broadband and the broadband development, what we're seeing in Q1 is a stable business. I, of course, cannot forecast it for the entire year, but as for Q1, it looks stable. Then you mentioned, when does the frequency purchase pay off compared to national roaming? That always depends on the commitments that go along with frequency purchase. Therefore, you can't just look at the mere amount, but you also have to look at whether we have new expansion commitments that come along with the frequency purchase. Therefore, there is no fixed formula for that. That's why I wanted to give you a broad picture and tell you, well, it does make sense for us at an incredible price. It does make sense for us, however, at a reasonable price. But to find out the sweet spot depends on many different questions. For example, what expansion commitments will we take on? Or when it comes to an extension, how long will that extension go for? The EU Commission says maximum three years, Federal Network Agency has talked about five years, our three competitors are talking about eight years, so all of that needs to be evaluated. Therefore, I cannot give you an exact figure and tell you this would be the right amount for us. Thank you. Does there have to be an auction or can't you just say that we give all four telecom companies the frequency and therefore you have the opportunity to use the money for the bad infrastructure in Germany, the money that would flow into tax liabilities anyway? Well, we're not against an extension. We have made a couple of proposals. Let me start with the auction. The other three always have the argument, well, if we lose the 800 frequencies, then there will be white spots. And you can see that also in our documents, we said that we're willing that in case we get the 800 frequencies, we will immediately exchange them for 700 or 900 frequencies. Our antennas don't care if they run at seven, eight or 900 frequencies. that would mean we could mitigate this sort of drastic approach of an auction. So you can basically replace 800 by 700 or 900 frequencies. So you could kind of mitigate all of that a drastic auction approach. I think that would be a constructive approach, and we have made that proposal. But we're also not against an extension. In France, there was an extension, for example, the so-called New Deal for mobile. France used to have three providers, then with Iliad III there was four, and then they split the frequencies by four. No issue there. We also said that we could even imagine for frequencies to be only extended to the three others, but they have the obligation that they lease the spectrum to us that we need. We're currently using frequencies of Telefonica. We're using twice 10 megahertz. until the end of 2024. So until the end of 2024, Telefonica needs to lease these frequencies to us. So these are frequencies that Telefonica cannot use today because we are using them. So they could say, well, we just extend that lease. We could do the same with other low band frequencies, for example. We could say, well, we give the 800 frequencies to Telefonica, Telecom and Vodafone, but we ask them to lease the 700 or 900 frequencies to 101 to a certain extent. All of these are possible approaches. Everything that helps us to save money is, of course, always beneficial and always welcome. Thank you very much for the very interested questions. Thank you very much for your attention. As always, we are always available for further conversations. I would now like to hand over back over to the operator. And after a short break, we will have a hopefully very interesting meeting with our mother company, United Internet. Thank you very much and all the best.