Hello, good evening, and welcome to our Nine Months 2026 Results Conference Call. Apart from our usual participation, the meeting is going to be led by Marco and Raimon, and then we have our CEO for our northern cluster, Nuno Carvalhosa, who will be talking about our new and exciting project in Poland. Let me hand over to Marco, who will give you some opening comments on the results.
Thank you, Maria. Good afternoon, everyone. Let me start with the six main highlights. Looking at our performance over the first nine months, we combined continued organic growth with greater operational efficiency and strong cash generation, supporting continued shareholder remuneration. Let's start with the revenues. Revenues grew organically by 4.9%, 3.6% from co-location and build-to-suit, and 1.3% from indexation. Very important. We achieved this growth with 22% less growth CapEx, reflecting greater capacity and capital efficiency. As a second point, let's focus on the cost base. Our larger tower base does not have to mean a larger operating cost base. Continued operating efficiency reduced OpEx by 4.4% year-on-year, expanding our EBITDA margin to 84.8% from 83.1%. Lease payment increased by just 2%, with land efficiency investment helping to offset inflation and offsetting a 2.5% larger tower base.
This is the operating leverage of our platform coming through in the numbers. As a third point, the operating performance and financial discipline, which are delivering strong recurring cash generation. Recurring Levered Free Cash Flow per share up by 18.3% to EUR 2.09 per share, with share cancellation from buybacks driving additional per-share growth. Free cash flow reached EUR 474 million in the first nine months, EUR 288 million more than a year ago. Leverage declined to 6.13x from 6.28x at the end of year 2025, and our 5x to 6x target remains unchanged. Our average cost of debt remains at 2.2% with around 80% at fixed rate, and the stronger cash generation supports organic deleveraging whilst we continue to return cash to shareholders through dividends and share buybacks.
Staying in the finance, we signed a EUR 375 million loan with an 8.5-year average maturity with the European Investment Bank to support existing plans for infrastructure deployment in Spain, France, Italy, and Portugal. This is the third financing agreement with the European Investment Bank, and it reinforces our longstanding partnership with competitive terms and a tenor that will help strengthen our financial structure and extend our average debt maturity. Finally, we're excited to share with you our new transformational project in Poland, a great example of how we can unlock further value from infrastructure we already own. It is an industrial blueprint on how to create value for all our stakeholders. It brings together new co-location, site rationalization, and accelerated RAN renewal. This project improves all our financial metrics in Poland. The investment we are making brings much higher future revenues, profitability, and return on invested capital.
It is a clear example of our capital allocation discipline, bringing an IRR at more than double of our cost of capital. We will share part of our efficiency with our client, making a new form of industrial collaboration while making better use of our platform. On this, Nuno will drive you through the Polish project. I hand over to Raimon, which will take you through our main highlight, and then we will pass the floor to Nuno. Raimon.
Thank you, Marco. Good afternoon, everyone. The financial overview shows how revenue growth translates into cash flow and per-share performance. These growth rates are on a pro forma organic basis. Revenue growth of 4.9% translates into a 6.5% growth in adjusted EBITDA and 8.9% in EBITDA after leases, reflecting operating efficiencies and proactive land management. The combined strength of our operating performance efficiencies and disciplined financial management delivered 11.2% organic growth in recurring levered free cash flow, and the share cancellation from buybacks driving additional per-share growth to 18.3%. Focusing on consolidated revenues, we grew organically by 4.9%, supported primarily by our core tower business, which we will explain shortly. Co-location and other businesses, together with build-to-suit and fiber, contributed EUR 104 million, equivalent to 3.6% of growth before indexation, demonstrating our ability to grow our business more efficiently with lower investment.
On a reported basis, revenues excluding pass-through reached EUR 3,008 million, up to 2.4% year-on-year. Points of presence continue to reflect healthy commercial activity, with demand for both co-location and new sites across our platform. In the third quarter, we added 1,120 net new PoPs, comprising 956 gross co-locations and 506 build-to-suit. Poland led the growth in the quarter with 673 net new PoPs, already reflecting our new transformational project. Comparing nine months 2026 with nine months 2025, gross co-location and build-to-suit grew by 5,567 and 2,979 PoPs respectively, with churn of 1,094 PoPs. This represents an annualized growth rate in net PoPs of 4.1%. We continue to see demand from our clients to expand their coverage and increase network capacity, as is reflected in the operating growth throughout the year.
Turning to our core tower business, we delivered organic revenue growth of 5.1% on a pro forma, supported by continued co-location, build-to-suit delivery, and contractual escalators. Reported tower revenues reached EUR 2,456 million, up 3.3% year-on-year. Our other business lines complement this performance with distinct growth drivers. Fiber, Connectivity and Housing Services grew organically by 8% on a pro forma basis, adjusted, as you know, for the French data center disposal in the first quarter 2026, and driven primarily by the continued rollout of the Nexloop project in France. DAS, small cells, and RAN-as-a-service grew organically by 5.2% on a pro forma basis, excluding the discontinued O&M activity in Spain. Finally, broadcasting grew by 0.9% on a like-for-like basis. Turning to efficiencies. Operating efficiencies and disciplined land management are allowing us to grow our tower base while reducing the combined cost of operating and leasing our sites.
Operating expenses on a pro forma basis decreased by EUR 22 million, or 4.4% year-on-year. Land efficiency investments are helping to mitigate inflation and the increased size of the tower portfolio. On a pro forma basis, cost per tower decreased by 3.5% year-on-year. Together with revenue growth, this cost discipline expanded EBITDA margin to 84.8% from 83.1% a year ago, and EBITDA after leases margin to 62.9% from 60.7%. On our EBITDA margin, we are one year ahead of our target plan. Let me show you how this strong operating performance translates into cash flow. Operating growth, efficiencies, and disciplined financial management together delivered stronger pro forma organic Recurring Levered Free Cash Flow of 11.2%. Cancellation of shares from our share buybacks provided a further uplift to the Recurring Levered Free Cash Flow per share to 18.3%, combining stronger cash generation with the benefits of shareholder remuneration.
Reported free cash flow reached EUR 474 million, an increase of EUR 288 million year-on-year, as a stronger recurring cash generation and lower build-to-suit CapEx intensity reinforced cash conversion. On a per-share basis, free cash flow reached EUR 0.71, almost 3x on a year-on-year basis. As you all know, our contractual structures and financial discipline help to protect against macro headwinds. Around 65% of our core tower revenues are linked to CPI, while the remaining 35% benefit from fixed escalators of 1%-2%. Exposure to energy inflation is almost entirely mitigated, with around 80% of costs passed through to tenants, and the residual exposure hedged through forward contracts and power purchase agreements. On the cost side, operating efficiencies, economies of scale, and active lease management help keep increase below inflation. This combination creates operating leverage, leaving our net exposure to inflation positive.
Raw material price increases also have limited impact given contractual structures, particularly on build-to-suit projects. On a final note, as Marco highlighted, 80% of our debt is fixed rate, so impacts of the more challenging environment will be progressive as we refinance upcoming maturities. On the balance sheet, we continue to make progress towards our leverage target while delivering shareholder remuneration. At the end of September, committed liquidity stood at approximately EUR 5 billion, including EUR 1.6 billion of cash and EUR 3.4 billion of undrawn credit lines. Our 2026-2028 maturities are largely funded, giving us flexibility to navigate market windows. IFRS 16 leverage declined to 6.13x from 6.28x at the year-end 2025.
Excluding the EUR 1.3 billion of share buybacks completed since 2025 that were not considered in our Capital Markets Day 2024, pro forma leverage would have been approximately 5.76x , illustrating the underlying deleveraging capacity of the business. Our leverage target remains unchanged at 5x to 6x . Our debt profile reflects the funding discipline that supports recurring cash flow generation. Around 80% of debt is fixed rate, limiting the immediate exposure of cash interest to market rate movements, while our average cost of debt stands at 2.2%. With an average maturity of 4.2 Years and a diversified funding mix, we retain flexibility to manage upcoming refinancing.
A recent highlight that Marco mentioned a few minutes ago, at the beginning of November, we signed a EUR 375 million loan with an 8.5 Years average maturity with the European Investment Bank to support existing plans for mobile infrastructure deployment in Spain, France, Italy, and Portugal. This third financing agreement with European Investment Bank reinforces our long-standing partnership with competitive terms and a tenor that will help strengthen our financial structure and extend our average debt maturity. I will now hand over to Nuno Carvalhosa, CEO of our Northern EU cluster, who will give you some more color on our Poland transformational project.
Thank you, Raimon. At our 2024 Capital Markets Day, we set out three priorities: promoting co-location growth, fostering co-location to suit, and pursuing tower rationalization. Our business in Poland is a strong fit for all three, with two independently acquired portfolios of passive infrastructures with a low tenancy ratio. Poland is a country in need of better coverage and higher capacity in the mobile networks, and because of a somehow delayed 5G auction, Poland was also in the need of a rapid rollout of 5G. Working together with Polkomtel, we have shaped an industrial project that improves network quality and strengthens our partnership while delivering better economics for both companies. It is a blueprint of how we can proactively create industrial value for our clients, helping them have the best possible networks and also allowing them to capture parts of the value such a project creates.
Importantly, it allows for better economics for us, Cellnex, namely through a higher asset use and stronger cash conversion. The project combined three main elements through 2030. First, approximately 3.2 thousand committed new co-locations with a further 2,000 committed from after 2030. These new co-locations contribute to increased revenues and EBITDA after leases. Second, site rationalization involving approximately 4.2 relocations, improving the use of our assets and contributing to increased EBITDA after leases through savings in both leases and operating costs. Third, accelerated RAN renewal and deployment of 5G configurations, enabling the wider network improvements and optimization program with an increase in revenues and profitability. We are also accelerating BTS delivery and our existing Play program from 2028 to 2027, bringing forwards the associated contribution to revenues and EBITDA after leases. Together, these elements combine contracted growth, a more efficient footprint, and network modernization.
This project reflects disciplined capital allocation with double-digit unlevered post-tax IRR and driving long-term tenancy ratios over 2x by 2035. Incremental CapEx totals EUR 500 million over 2026 to 2030 versus our previous projections to support an increased scope in footprint and accelerated rationalization plan. The project combines attractive returns for Cellnex with benefits shared with our customer. With that, I hand back to Marco to put this project into the broader group context.
Thank you very much. Thank you very much, Nuno. I'm sorry, I did not switch on the microphone. Thank you very much, Nuno. To sum up, this first major tower rationalization project in Europe addresses continued demand for capacity and coverage. Nuno, I think it's the first in the world. I never heard about such a big
To the extent of my knowledge, indeed.
Okay. To sum up, the first major tower rationalization project in the world addresses continued demand for capacity and coverage. It puts our strategic business model into practice to improve asset returns and cash conversion. We're combining network rationalization with an expanded client base to capture higher returns with an IRR more than double our cost of capital. Importantly, we're strengthening our client partnership through additional long-term network commitments that will last at least until 2046. We're proving again our ability to proactively and successfully manage our business relationship, as we have already shown in the likes of Spain and the U.K., and we're on track to demonstrate in France. To conclude, let me leave you with four messages. First, our growth is releasing more cash. Strong organic growth, a more efficient cost base, and a lower BTS capital intensity are translating into stronger free cash flow.
Second, that cash generation allows us to progress towards our leverage target whilst continue to reward shareholders through dividend and share buybacks. Third, our new transformational project in Poland opens an exciting opportunity to unlock further value from assets we already own, combining contracted growth, shared efficiencies, and network modernization with attractive IRR. It's a win-win-win project for Cellnex, for our clients, and for our shareholders. Finally, we reiterate all our 2026 and 2027 guidance on a like-for-like basis. Our new Poland transformation project implies EUR 300 million CapEx impact on free cash flow in 2027, and it is set to bring us significant growth and increased returns in the future. We continue to create long-term value for our clients, our platforms, our shareholders. With that, Maria, back to you.
Okay. Thank you. Thank you very much. Now, let's go to the questions. The first question on the list comes from Akhil Dattani from JPMorgan.
Hi. Good afternoon. Thanks for taking the questions. The first one was just to try and understand how we should think about the economics of this Polish new contract a bit better. Marco, I guess it's interesting and obviously relevant to understand this is a double-digit unlevered IRR, but I guess the challenge for us is to understand a little bit about how that should impact our forecast going forward. I just wondered if you could help us understand how we should try and think about it, and I guess there's two different ways that I was trying to get it right in my head. One is, at the moment, we obviously have certain forecasts that we're modeling for Poland, and I'm sure we have all slightly different assumptions.
Is there something you can help us with to understand maybe what the compound growth rate of revenue or EBITDA is going to be going forward, or maybe some other way to help us conceptualize exactly what this does to the numbers in Poland going forward? So, that's the first thing, just trying to really understand. We understand the CapEx, we understand it's a good return, but I guess just trying to understand exactly what that means relative to existing assumptions. The second thing is, Poland's been an interesting market for you in that it seems to have been a leading market on a number of themes. I remember a few years ago, this was the market where you did an active sharing arrangement as well. I'm just trying to understand how we should think about that. These are two scenarios where Poland's been leading.
When you think about the rest of your footprint, are there other markets where you've already got active negotiations? Is this something that you think certain other markets are better suited than others? Just help us understand a little bit better, beyond Poland, how broader could this theme be? Thanks very much.
Yeah, sure. I tried to drive you a bit more, as you said, the conceptualization. If you wanted to, each of you, if you want to have more modeling details, of course, the IR team is at your disposal for every question in the coming days. Let me put the conceptual framework. Point number one, Poland is a big country in terms of land. It is more than 300,000 sq km, with a population that is one of the large European countries. If you make apple with apple, an average MNO in Poland has a sort of 12,000-13,000 PoPs. You compare with Italy, an average MNO has 20,000-25,000. It is evident that Poland needs growth. The other data point is ARPU in Poland is not very high, so you need growth, and you need to make the growth efficiently.
Continue multiplicating and working standalone is a bad idea because the network will be very expensive. The idea is we have to incentivize our clients to go together and to share as much as they can. Because every time we share something, we can dismantle a tower. Dismantling a tower, you save the lease, you save the maintenance, you save the energy, you save a lot of costs that let's make an easy number. I can tell you we are talking about something between EUR 10,000 and EUR 13,000 per site. You easily understand that if you release more than 4,000 sites, you are making available a pot of saving that is pretty big, and you can reinvest together with the client, a part making, transferring part of the saving to the client, and this is the win for the client.
The client, with this money, self-finance part of the investment in the 5G. This investment comes back to us with a broader network, more PoPs, more towers. Everything is increasing at the price that is good for the client and which help us because part of the saving are retained by us. This is how the mechanic works. Of course, this is a multi-year project. You can imagine that moving 4,000 towers, swapping more than 10,000 antenna, increasing 3,000 PoPs takes more than a year. Numbers. In our original plan, we had EUR 300 million. The new plan costs total EUR 800 million. So, we have a delta of EUR 500 million. I make it simple, EUR 500 million is the delta. All the rest is a different timing. What we do with Play, we agreed to anticipate the investment we had with them by one year.
What is the impact on the total spending? Zero, because it is an anticipation. Simple numbers. We invest EUR 500 million for making higher revenues, using operating leverage in order to have a growth in EBITDA that is more than the growth in revenues, and we deliver to our client a modern network. To your second question, yes, it is a model that can be replicated, but the trick is that you have to make on sites. I mean, if you make a project that have to call the attention not only of the CFO, but the CFO, CTO, CEO of your client, it has to be big. Because if it is not big, ultimately, the NPV of the project, you see the pain of moving, but you do not see the gain, because the gain is too small. Can be replicated? Yes, it can.
It can be replicated almost everywhere, I would say, but it's obvious, the bigger the portfolio, the more you have the possibility of making network rationalization. As always, you need to break the ice. Now, if the Polish example, I am sure that the Polish example will be something that all the CTO will look with a lot of attention, so we are in the hands of Nuno, which is our top gunner in this. I hope that I was clear, Akhil, because this is a great project, it's very new. If some of the colleagues want more details or want to speak with Nuno, more than happy to continue on the topic.
Yep, that's super clear. I mean, just one super quick thing, it doesn't need a very long answer. It's just, would there be a CMD update? Because I guess you gave a CMD up to 2027, so I guess if we're trying to understand the numbers better into the midterm, is there some sort of vision as to when we might get an event like that, which would then help us with those mid-term expectations?
Yeah, sure. We have guidance until 2027, so somewhere in 2027, we have to make an update. It's a topic that we are discussing with the board. The ideal moment is as soon as we have a touch of clarity on France, more than happy to share, because if I do it, I can eventually do it tomorrow morning, and I tell you, I don't know what's going to happen in France. 2027, for sure, we will do it.
Very clear. Thank you very much.
Thank you, Akhil.
Okay, now moving on to the next question comes from Maurice Patrick at Barclays.
Yeah, thanks guys for the call and taking the questions. Just for me, a question on the cost side, please. I mean, you've spoken in the presentation, in the prepared remarks, about the strong EBITDAaL growth driven by your cost focus. You talked about being, I think, a year ahead, I think, in terms of your EBITDAaL margins. So, I guess curious as we look towards 2027 and the guidance you've reiterated today, love to get some more understanding around the ability of taking more cost out in the next 12 months and what that implies in terms of the kind of margin progression we should assume for next year. Just as a quick side question, please.
On the U.K., you've been speaking for some time now about the likelihood of the U.K. catching up with the rest of Europe, and as a U.K. resident, I can certainly vouch for the hope that you'll be right. But we've recently had Vodafone, Hutchison having a capital market event where they've been talking about their extent of wanting to have the best network in the U.K., but then VMO2 and BT saying they don't see a need to kind of really make any big change, and they're really happy with their sort of 5G investments and network slicing. Would love to get your views in terms of when you think that sort of U.K. acceleration will come through. Thank you.
I leave Raimon on the cost, and I take the U.K., but you have to know that Nuno is on the ground in U.K. as well, so I ask Nuno to help me in the second part.
Okay, so hi, Maurice. I am going to start with the cost. As you have seen this year, we are EUR 9 million better in terms of cost, mainly coming from the OpEx, where we have EUR 22 million, and this has been quite recurrent over the past couple of years. We have been explaining you the different initiatives that we have been undertaking, and it is true that we have managed to accelerate by one year the EBITDA margin that we were expecting for next year into this year, this 84.8%. That is a very good improvement. Can we expect further savings in the coming years? The answer is, as we have always said, we always try to keep doing savings that mitigate the inflation in our cost, and it is our target, trying to mitigate. This year, we are doing much better.
It is true that we have accelerated savings, but we cannot expect always to accelerate faster. At least we need to maintain saving the inflation cost.
Okay, thank you. Yes, this year we saw something a bit unusual, that the cost of the 2026 in absolute value are below the cost of 2025. On U.K., I give you my gut feeling. There are two numbers that don't match. One, the target that Vodafone is given in terms of quality, in terms of coverage, in terms of bandwidth at the edge of the network, in terms of speed, et cetera, and the number of antenna. So the 26,000 seems to me a little bit low. So if I look to the traffic growth and a bit of the congestion that I see in the U.K., I think that they should target more 30,000 than 26,000. So how is doing Vodafone? I leave to Nuno to give you because he is on the ground. But I have been very vocal.
I was expecting at least BT reacting a little bit before. There are plenty of activity in this moment. You know that they have been super active on the fiber, which took a lot of their attention. It is also true that they were starting from a comfortable position of being the leader in quality in the U.K., so they could react a little bit later. But it is not to consider for granted forever. Nuno, what is your view?
Yeah. I was going to pick on that exact topic, which is BT has been enjoying for many years a strong lead in terms of perceived network quality. That for the one side. For the other, I would argue that the operational complexity that goes into a project of the ambition and size that VodafoneThree is putting together takes time to pick up, to gain momentum, and to later on start producing results. Indeed, we are waiting to see a stronger reaction from the other operators. As you rightly pointed out as well, BT has been in a very public push to force the deployment of fiber throughout the country. If you want to, they have their challenges of their own.
I would say that these are probably the reasons that help explain the fact that it is taking a bit longer than what we would have expected for the competitors to react.
Do you expect 2027?
Yes, we do.
Yes, we do. Me too.
Thank you.
Thank you.
Okay, so moving on to the next question from Ondrej at UBS.
Hi, everyone. Thank you for the presentation. I had a follow-up on Poland, please. I am just trying to understand what kind of overlap, if any, there is between the rationalization, so the 4,200, and then the up to 5,200 co-locations. What is the relationship between those two? Because you mentioned, Marco, that there are 13,000 sites per operator in the country. Are we going to be basically seeing 9,000, so the five plus 4.2, as being shared out of the 13 in the end, or how does this actually work? That is one question. Then maybe if I pivot to Italy.
If you can update us, please, just on the, I guess, your point of view on the spectrum renewal process, whether you still think that the requirements embedded in the proposals could translate into something like 15,000 sites, I think you mentioned, Marco, last time in the country, or whether that view has changed a bit. When do you expect all of the stakeholders to agree on the process? What do you make of this insinuations that you could host some of the sites that they plan to move from INWIT, as they said earlier today? Thank you.
If I understood properly, Ondrej, on the first question, you were asking if there is overlapping between the 3,200 co-locations and the rationalization. It is two different things. The 4,000 rationalization, it is sites that we have today with Polkomtel that we are going to be dismantling and moving to sites that we have with Play. So it is moving from one site and moving to another site where we will have tenancy too. The other 3,000, it is growth, new antennas, new co-location on existing towers of Play as well. But it is two different things, not overlapping one with the other.
Yep. Nuno, are you worried about the rooftops?
No, I am not worried about the rooftops. No.
Yeah. For you to know, it has been a big issue in Germany. In our case, when we did it in Portugal, you did it in Portugal, it was, I do not want to say easy, but manageable.
Yes. Well, it definitely was not easy, but it did not stop us from delivering, doing a very successful delivery for a new entrant in the country.
On spectrum renewal, sorry, Ondrej, the line was a little bit disturbed. If I understood properly, you were asking spectrum renewal in Italy. The process has been questioned partially, I would say, questioned by AGCM, so the antitrust authority. Why I say partially? Because AGCM did not oppose to the concept of having an early renewal or a renewal in exchange of investment. This was not the point made by the antitrust. The antitrust made the point, which is, if we freeze today the same level of spectrum that was awarded years ago, it's not impossible that this is unbalanced vis-a-vis the change in market share of the different operators. It suggested to take into consideration not only the need of more investment, but also the need or the possible need of some spectrum rebalancing.
I always like to see the glass that is half full. The good is that the concept of a spectrum renewal in exchange of more investment is considered okay. How many PoPs imply the targets that AGCOM was setting? It was particularly high, so I think that this eventually is a bit too ambitious, because when you make a total investment that is needed in order to cope with these targets, it's more than the cost of the licenses. It's a bit weird that in order to save money, you spend more. I think that some fine-tuning is needed, which means that possibly it will not happen before the year end. We continue to believe that the process will see the light at the end of the tunnel. It's going to be Q1, end of the Q1, beginning of Q2.
I would put it in the springtime.
Okay.
Thank you. If I may.
Go ahead.
If I may have a quick one. Thank you. Making a bit of a link between the two countries. I am just wondering, because I believe Polkomtel has no CPI cap on your MSA. With this renegotiation, has there been any, I guess, attempt by the customer to change that along the lines of what I guess has been raised as an issue by your competitors, anchors in Italy? Basically, is there any change in that regard to the Polish contract?
Well, the Polish contract, like every contract, is made of many things. We have been reopening the dialogue on a few topics. One, for example, was the SLAs. We discussed the SLAs with our client, and we have been fixing demanding but feasible SLAs, that it was one of the strong points of Nuno, of attention of Nuno. The other, we have to find a reasonable cap of inflation. Then we have to consider that we are not in Switzerland, we are in Poland, so reasonable has a specific meaning in Poland, which is different from Switzerland. Sorry if I do not enter too much in the details, but yes, those are the two most important among many others that you have been negotiating.
Yeah.
Okay.
Thank you very much.
You are welcome.
Okay, so now moving on to Roshan at Deutsche.
Great evening, everyone. I have got a couple of questions, please. Just going back to the Poland project, please. I think, Marco, back in 2020, when you guys kind of announced the Augmented TowerCo, you said you needed kind of buy-in from the MNOs. You mentioned a kind of, I think the word used was a share, and they will share in the upside of this project. I guess that mechanism is potentially seeing an anchor customer move to a secondary. Is it possible to get a sense of what the pricing delta is there? Clearly, you get the upside of fewer lease costs or maintenance costs, but it would be interesting to understand how that works. Perhaps tied to that, without going into too much modeling questions, you said you accelerated the Play BTS.
If we look at your BTS tracker, you have got EUR 300 million BTS to deploy. Is there scope to rationalize that 300, something which you previously talked about combining BTS? Could reality, that EUR 300 million not be EUR 300 million, although it is more front-loaded than before?
No.
My second question is, I will be very quick, it is just on Spain. I appreciate it is volatile from quarter to quarter, depends upon what is happening with Digi and their RAN. You did highlight a perhaps pickup in the second half of the year in terms of PoP deployment. Given that Q3 was not anything too special, should we expect a bigger pickup in Q4 then in PoP deployment in Spain? Thank you.
Yeah. I give you the elements for making the math. I told you that every time we dismantle a tower, we save, depending on the case, but something between EUR 10,000 and EUR 13,000. It is obvious that if I am in a rural area, the land costs less, and if I am in the center of Warsaw, the land costs more. But these are more or less the two elements. The idea is we split by two, so it is a sort of 50/50 that we do with our client. When we move an antenna, we move this antenna, and this antenna to the client will cost a price that is reduced by the level of the saving that we are going to share with him. It is not a difficult algebra. Do you want to- Yeah. This is Spain.
You have to remember what we announced when we made the agreement with Orange. They made my life easier, and it is not MasOrange, it is just Orange. When we made the agreement with Orange, there was a sort of a decoupling between the impact in revenues and the impact in PoP decommissioning. The PoP decommissioning was divided in two big tranches, the one in 2025, which has been completed pretty much upfront loaded. It was at the very beginning, and one in 2026 that has to be completed this year. The impact in terms of revenue was not related in price per quantity, but it was negotiated among the two CFOs, having a scale of discount and recovery after the time. The 2026 is still a moment in which, in terms of PoPs with Orange, we are completing the decommissioning. We signed, as you know, the agreement with Vodafone.
The agreement with Vodafone includes the possibility to transfer onto our towers a certain number of PoPs. I am not sure that we are going to be able to start transferring or moving or installing PoPs already in Q4, but believe me that our Spanish team is working full steam. Last, there is the network sharing between Digi and Telefónica. This is fully in the hands of Telefónica that is progressively opening PoP after PoP to Digi. They have their calendar, and they follow their calendar.
That's great. Thank you.
Thank you very much.
Okay, now moving on to the next question from Rohit Modi at Citi.
Hi. Thank you for the opportunity. Two questions, please. Firstly, on the churn and the PoP additions this quarter. The churn was higher than what you have done in last few quarters, and also PoP additions have came down. I understand there's a bit of Spain thing, but then if I see Switzerland, there's a decline in PoPs. If you can give a bit more color if this is kind of a new PoP addition run rate, normal run rate that you're expecting in future. Second question is on refinancing, and thank you for the color in the slide. Just trying to understand, given the refinancing due in 2027, I understand you have credit facilities to use, but if you use that, what kind of impact do you expect on your interest cost for next year?
How that's going to impact your free cash flow guidance range, because it's like a EUR 200 million guidance range on free cash. Just trying to understand where we should incline towards higher side or mid lower side. Thank you.
Cool. Good evening, Rohit. On the churn, we had some churn in Italy from Fastweb, which was RAN sharing on Wind Tre. You can easily understand that being Fastweb, now Swisscom Vodafone, not necessarily they need this RAN sharing. Progressively, they are releasing some RAN sharing. The good of the story is something that I've been blaming for years, so that the RAN sharing in Italy is priced very cheap. Sometime bad news turns into good news because we are not losing significant revenues. On Spain, I already mentioned. U.K., pretty uneventful. France, totally uneventful. On PoP additions, as you know, the big suit is progressively fading out. In this moment, what we saw this year is France a bit light compared to previous years, and it's totally understandable because of the SFR case.
Poland doing great, because in Q3, Nuno Carvalhosa started with the project in Poland, and we will see further acceleration Q4 and Q1 next year. We have to consider that Poland weather is not your best friend, so Q4 and Q1, you have to face a little bit of meteorological. That meteor is not really helpful. All the rest is doing according to the plan, and on U.K., we already commented before that we are here waiting for the D-day.
On the debt side and on the leverage, you have seen before in the presentation that we have, as of today, more or less EUR 5 billion of liquidity between the cash and the undrawn credit facilities. This is giving us flexibility to see when is a good window to exit into the market and tap the market. We have been holding now for some months, as you have seen, but it's true that over the next two years, we need to refinance EUR 5.5 of debt maturities. We are looking at seeing when it's a good moment to go to the market. Clearly, today, if we issue, the cost is going to be a bit higher than what we were expecting before. If we go for five, seven years, the cost might be at 4.5% or 4.7%, more or less, a bit less than that.
It is a bit higher than what was our original plan at 3.75%-4% more or less. We are having certain other ways to try to mitigate that increasing cost. As of today, we have the liquidity, as I said, that gives us the flexibility, and we will see when we exit into the market if we see that there is a good window to secure the financing and increase the maturity of the debt. One of the things that we have done, and you have said, is the European Investment Bank financing. It is at very good terms. It gives us 8.5 years of maturity, so it is helping us to improve, and we will keep on working like that, trying to make sure that we balance the cost.
Also, remember that 80% is fixed cost, and we are today at 2.2%, so any new issuances will not have that much of a significant impact for the coming year. That is why we have reiterated the guidance on a like-for-like basis.
Okay, so moving on. The next question comes from Ulrich at Bernstein. Apologies.
Yeah, thanks very much.
Sorry.
That is fine. Thanks so much. Appreciate it. All of my questions have been answered already at this point, but one question is, you did one refinancing with the EIB, and I think in the release, you have not actually disclosed the terms. Is there any way you can tell us how that looked? I think in the prior answer, you said it is very good terms. Could you just talk about that a little bit, given that you have a very low cost of debt, but most of it trades now quite a bit below par at this point? Thank you.
Well, as you know, our debt is trading extremely well today in the market. This European Investment Bank financing that we have done, we have shared a few terms, just the maturity, 8.5 years. The cost per se, if you know more or less how they work, is Euribor plus a margin, and it is a quite significant attractive margin that is helping us improve our cost of variable debt. So it is improving on a marginal basis the cost of variable debt that we have today.
Thank you.
Okay, now moving on. Fabio from Mediobanca is the next question.
Yes, hi. Good evening. Two questions, if I may. First one is a follow-up on Italy. I was wondering if you agree with the need to link spectrum renewals to the actual availability of passive infrastructure rights. If so, I was wondering if, in your view, regulator also shares this view. Second question is on what's going on in the U.S. with satellite D2D developments, as you managed to put a slide in your presentation deck. Can you, Marco, share with us your view on what's going on there? Thank you.
Yeah, sure. Well, on Italy, I would say that we have to avoid to make confusion between what are the investment that an operator commits to make and what are the content of specific contracts between operators and tower companies. I continue to believe that tower companies have to be considered a strong partner to the operators in order to let them invest at very reasonable terms, reducing their investment. We have seen what has been the content of contracts that has been negotiated in the past.
I think that this is a very important principle, that if you want to continue to join, to benefit from the investment coming from a specialized company like the tower company with a specific cost of capital, which is an infra cost of capital, well, at this point, the important thing is the stability of the legal and the regulatory framework, not only for the MNOs, but also between MNOs and TowerCo. I stop here because I think that whoever wants to understand, understood. On satellite, it's super interesting what's going on. Let's oversimplify. The news flow coming from Starlink and SpaceX is telling several things. Point number one, satellite for delivering carrier-grade quality in urban areas is not okay.
If a satellite company wants to provide an integrated service, which is D2D plus a service in dense area, there are only two ways to do it. One is you build your own network, or two, you make an agreement with an existing MNO on terms that have to be non-predatory. It has to be good terms. If I see what is happening in the U.S., the move that SpaceX seems to embrace is they wanted to build a, let me say, relatively small network. Of course, it's 80,000 towers, 80,000 antennas, a very good number for my colleagues in the U.S. But if you change your hat for a second, these guys are going to have 80 MHz versus AT&T, Verizon, who have 300 MHz. So, the difference of the capacity of the two networks will remain very material.
80,000 antenna vis-a-vis something close to 200,000 antenna, which is the MNO coverage today. Is it true that they will complement with the D2D? Yes, but pay attention, because today, with the D2D, you send an SMS. You do not even send a WhatsApp, okay? This gives you the possibility of looking to Europe, what the European large MNOs are doing, that they are saying, "Okay, let us work together." Orange, Deutsche Telekom, Telefónica, "Let us work together in order to find a common ground for having an intelligent way of cooperating with a satellite constellation in a friendly way, and not in an opposition way." Is it going to work? I do not know. What I tell you is that today, replicating the SpaceX strategy in Europe needs the spectrum, which is going to come with the DNA. We are here to see.
Very clear. Thank you, Marco.
Thank you.
Okay, moving on. The next question comes from Abhilash Mohapatra at Exane BNP.
Hi, good evening, and thanks for taking my questions. I have a couple on Poland, please. I think when it had been announced, it had mentioned low double-digit IRR. Has that changed in the meantime? Is this additional returns? Given that the Polish risk-free rate has gone from 1% to 6% in the meantime, is this still a good deal? In other words, is this a deal that you wanted to do, or is this about honoring an old deal at a time when the counterparty there seems to be facing challenges, as we have seen with Cyfrowy Polsat and talk of refinancing the debt and the share price numb?
Also related to that, I think in the original agreement, there was a mention of up to EUR 600 million of additional CapEx for build-to-suit, and the RAN equipment upgrades that Cellnex would pay for sounds like that is now gone up to EUR 800 million. What is the difference between these two figures? Related to that, should we think of additional revenues versus the originally announced run rate financials? Thank you.
Yeah, sure. The return, and please, Nuno, correct me if I am wrong, the return in Poland is in the, I would call it low- teen. It is a very good return if you take a group WACC that I see several analysts use a 6.5%. It is something that goes in the direction of being 2x . Am I happy investing the shareholders' money at 2x the WACC? Yes, I am. I would like to find a good project like this. Of course, the beauty of this project is that it combines a very good return for us with a very attractive proposition for the client. Because if you speak with our client, he is not going to tell you, "Oh my God, these guys are killing me with their prices." On the contrary, we are making possible, exactly the contrary, exactly the contrary.
Now, we are making possible for them to anticipate and very quickly roll out a brand new network, brand new. By the way, full using only non-high-risk vendors. It is a full brand new network, way bigger than before, because at the end, it is going to be twice as big as before. It is a very good point. How to split, how to reconcile the EUR 800 million with the remaining EUR 300 million? The EUR 800 million are made of three components. One is reinforcing all the sites that have to host two antennas instead of one. We are completing the build-to-suit program with Play, which is the famous EUR 90 million that we are anticipating this year. You have a second component, which is, the first component is preparing the network to host more antenna. The second is dismantling.
Dismantling doesn't come for free, so you have to make an operation dismantling. The third is renewing the RAN. I don't know. Yeah. That's it. More or less, it's- Yeah. Did I answer?
Thank you. Yeah, thank you. Just, if I may quickly squeeze in a follow-up on the U.K. Vodafone, in their recent sort of U.K. Capital Markets Day, had alluded to this idea of additional synergies from leases that were not very specific, so additional EUR 100 million of savings or so by 2031. Could you please confirm, does that have any impact at all on your Hutch MSAs, or is that unrelated and we assume no impact on your business?
Well, since they have the synergies before 2031 and my contract expire in 2037, I would say that if they want us to discuss with them, more than happy, but not because I have an obligation, but because we are always happy to sit with our clients and to discuss with our clients. My contracts, I sit on contracts that are lasting until 2037.
That's clear. Thank you.
All or nothing at renewal.
Okay. Just conscious of time. It is getting quite late in the evening. Moving on to Andrew Lee. Hi, Andrew, from Goldman Sachs.
Yeah, good evening. Just a quick question for me then. The equivalent points of presence growth remains suppressed at 3.5% this quarter, and I think part of that is due to you tapering build-to-suit, as you mentioned during your presentation. It is also partly due to temporary, particularly lower build-to-suit. Just wondering if you could give us any kind of forward-looking expectations on re-acceleration on that equivalent points of presence growth. Do you expect that to accelerate in Q4? Have you got any visibility on a further acceleration or kind of a more definite expectations of acceleration through 2027? Thank you.
Well, on Q4, as we were discussing, we would see Poland further increasing the activity. On France, we tried to front-end load all the work on the build-to-suit that we had. We have still something to do, but not very big numbers. I would say that we are going to progress in line softer and with Poland a bit stronger than what we expected. On build-to-suit, yes, we-
with our client, they are based on the dismantling of these sites. Hence, the reason why I am saying that for as long as we find profitable enough reasons, we will do so. But, just quickly, edge computing, it is absolutely suited for this type of infrastructures. If any, edge computing, at least in Europe, has been elapsing a bit, there has been this recurrent promise that it would start picking up, and until now, it has not taken place. We can use the sites as well, for example, for providing energy services, resilience, for flexibility. These are just, to name a few, some of the types of utilizations that we can do from these sites.
Yeah, no nominal difference. The only-
Nuno, is that in the IRR you have given, or could that potentially be incremental upside?
No, it is not in. It is a potential incremental upside. The only point that I wanted to make is pay attention when you say that we have the energy, we have some 10 source/100 kW, and when you talk about inferencing AI, you are talking about half a megawatt upwards. Yes, it is true, we are already on the grid, which is super good. On top of what you were saying, when in Spain we released approximately 400 towers, we were ready to dismantle 400 towers that were made unnecessary by the MasOrange consolidation, then 100 out of 400 has been reused. Now, do not take, then if it was 400 is 100, is 4,000 is 1,000. I do not want to tell you this, but, yes, your point is correct.
Interesting. Okay, thank you very much.
Okay, I am really conscious of time now. We have got some people in here that need to go and catch some planes. The second-to-last question is coming from Arnau at Bestinver.
Good evening, and thank you for taking my question. We have seen renewed speculation around a potential take private of INWIT, and earlier this year, there were news flow suggesting a possible combination between Cellnex and GD Towers, if I remember well. I appreciate you cannot comment on market rumors or hypothetical transaction, but more broadly, do you see a potential take private or larger industry combination as a credible scenario for Cellnex? Thank you.
Well, the only comment I want to make on our take private is the check is very big. That is the only comment I want to make. It is a rumor, so we tend not to comment on rumors. I take your suggestion on M&A. I think that there are two M&A, two kind of M&A that people like us should look at. One is a big- or eventually cross-border M&As. With this market sentiment, I do not see them very easy to happen. Then you have in-market consolidation also that everybody talks about MNO consolidation, and there is also the tower market that sometime is too fragmented. So, there are countries in which the market is too fragmented, and these make available synergies.
With synergies, again, you can share the synergy with your client, and you end up with a happy client, with a better return, with good business. I think that the M&A season, I do not see really beyond the corner the big cross-border, but the in-market, I think that could represent an interesting value creation opportunity.
Okay. So now closing the call with the last question from Nick at Berenberg. Hi, Nick.
Hello, guys. I hope you can hear me. It was just a very quick one, Marco, on savings, please. Just very good OpEx savings in the quarter. I think your OpEx is down about 8%. Is there anything particular that is changed in the quarter in terms of initiatives, and how much further can you go? Also, how much is this linked with the increased mix of co-location as opposed to BTS growth, please, and just the sort of combining of sites and raising of tenancies? Thanks very much.
Okay. So on savings, this year, we created a transformation project with a transformation office. At a certain point, you cannot make big savings if you continue to do the same things the same way. So you have to change the way you do the things. The second is we started investing heavily on not only on IT, but on IT plus AI. If you put together the two things, so making the things slightly differently, and on the other side, working more with more automated processes, we are working slightly different than before, which delivers more speed. A company that is leaner, it is not only that the company spends less, it is normally faster. So you stop wasting time in activity that give you not enough value. So, this is the reason why. So we changed our approach to efficiency, and this is working.
There is no magic. It is just focus, and when you focus on something, normally you get the result. Of course, the management team approached and embraced the new philosophy, because you can push as much as you want, but if the management team doesn't buy the concept, you go nowhere. I didn't get your second, if there was a second question or not.
No, it was a comment to see if the cost was impacted by the mix of co-location build-to-suit.
No, it is not.
There is no impact. The only difference there is on the leases. If you have less build-to-suit, the lease cost increase is lower, but not on the normal OpEx.
No, interesting. Thanks very much.
Thank you.
Okay. Well, thank you, everybody, for staying on until so late tonight. As usual, we are always available for taking your questions and follow-ups. We know we have given you some exciting new news, so you may have additional questions you would like to cover. Look forward to speaking to you in the coming weeks. Bye.