MERLIN Properties SOCIMI, S.A. ($MRL)

Good afternoon, everyone. Thank you for joining MERLIN's 3M '26 trading update. As we always do on quarterly results, our CEO, Ismael Clemente, will briefly walk you through the main highlights of the period, and we will then open the line for Q&A. Without further delay, I pass the word on to Ismael.

Thank you, Teresa. Well, MERLIN is off to a very good start of the year with total revenues up 11.2%, owing to basically a 3.5% increase in gross rents like-for-like and the additional revenue brought by data centers as compared to the same period last year. The FFO, however, is only up 3.9%, still better than what we predicted. It was slightly lower than last year, but it's only 3.9%. And as warned already at the full year results, it's a combination of more financial expense, a 10% increase, and much less financial income because in the same period last year, we were still enjoying a significant amount of cash in our bank, so 40% less income, financial income. That puts the company at 8.8% total shareholder return per share year-on-year comparing the same periods. We have seen a very strong activity in all of our divisions, but particularly in traditional asset classes. Of course, Spain continues enjoying good macro, but I would not take out importance to the asset management effort carried out by my colleagues. The -- occupancy remains very high at 95% and quite stable despite the fact that in the first quarter, about 60% of renewals come due, and this is Spanish idiosyncratic. Everything is done either in January or February. So most of our renewals come due in the first 2 months of the year. But the erosion in occupancy of close to 60 bps, I believe, is quite acceptable. We will discuss the guidance for next year probably in the next quarter. But what we said at the end of the financial year 2025 remains true. In offices, we said between 93% and 94% and our models are giving us now midpoint in that range in shopping centers, full stability, so between 90 -- around 96.5%, something like that. Remember, in shopping centers, we are starting to yield manage a little bit the portfolio. You have seen it in the release spread. So we are starting to fight in a good way, the occupancy cost ratio, which now stands at the lowest I have ever seen in my professional life at 10.8%. And in logistics and some of you, I know, are worried about logistics because you are kind of assuming that there is some sort of underperformance in the portfolio. The truth is that simply is the GXO share, which is a big one, it's 48,000 square meters, and it will take time to lease it up. I mean, because it's a relatively big deal to swallow for the taker, we are in negotiations with more than one party. So we are negotiating with multiple parties. And that will basically move the needle of occupancy as of year-end. I mean, if we cannot sign that share before year-end, occupancy will be more in the region of 96%. And if we are able to sign it, it will be 99%. Anyway, I already warned you that 99% is abnormal. So please do not bang on our head every time we go down from 99% because 99%, of course, is not reasonable. More importantly, we continue executing the mega plan in a very satisfactory manner. I think it's -- all the assets under construction are on track to meet the delivery date. And the ones in which we were requesting licenses are now little by little achieving very interesting milestones in terms of construction licenses, et cetera. I will go in further detail in a minute. Noncore sales, meaningless for the quarter, EUR 6.8 million at premium to GAV. But more importantly, we have been signing EUR 123 million for execution in '26 and '27, giving us optionality for receiving the cash flow. As you know, while we continue building data centers, it is critical for the company to keep cash flow. And as such, every time we sell an asset, we try to do it in a type of contract in which we keep a lot of optionality as to what is the moment of execution and what is the moment of loss of cash flow and receipt of the money. At present, we don't need money. I mean we are overfunded. We have too much cash at banks. So what we are trying to do is drag a little bit our feet in terms of sales. And that EUR 122.9 million, it's also a premium to GAV. And I know some of you are not worried of making that question. It's at a premium to GAV. The NTA per share, it's EUR 15.32 with no valuation or no appraisal during the quarter. Another thing that has raised the eyebrows is we have said that we expect substantial revaluation of DCs in 1H. This is no nuclear science. It's simply that up to now, the scope of valuation of our appraisers has been a total potential of 240 megawatts of data centers. And just with Lisbon, we moved into 340. So it's a very significant jump ahead, much more scope, and I'm sure this is going to have an effect in the valuation of the portfolio, together with the fact that there's been a near elimination of any leasing risk in Phase I and increasingly in Phase II. And as a consequence, I'm sure that the valuers will reflect that in the discount rates and also bringing the cash flows closer to present. So this is why we expect some jump in values in the first half. As you all know, in March, we executed a capital increase and ABO to fund Phase III. It was a very good transaction, reflecting particularly your support to the company. It was heavily oversubscribed and executed at strike. Thanks for that. At the end, 100% of the paper was placed with existing shareholders. Of course, begging pardon to the super minority shareholders that, of course, cannot participate in ABOs because of their non-institutional nature. Even from a legal standpoint, you cannot address them. The final distribution after the general approval of the General Shareholders' Meeting of 2022 will be paid on May 25. Some of you have asked us about why the FFO goes down by 5.6%. As you know, we used the absolute number of shares of the company at the moment of reporting, which now includes the capital increase. We don't, we don't use the average. If we were to use the average, the FFO will have grown by 3.1% and the adjusted FFO by 3.5%. So, that we prefer to report what we see. And what we see is that now we have 620 million shares. And as such, if we have to divide the FFO by those -- by these number of shares, it's minus 5.6%, which anyway is already below the 10% theoretical dilution created by the capital increase. So with some probability maybe during the year, we will continue taking event eroding that dilution created by the capital increase. In terms of behavior of the different asset classes, offices, as you know, ended up the period at 93.6% occupancy, down close more or less 60% compared to the 94.2% at which we closed the year. The like-for-like was 3.1% with a very interesting release spread of 2.8%. So offices continue to show some strength despite the fact that we have a problem in Barcelona, a problem that will aggravate in the second quarter because in the second quarter, as you know, we have a scheduled exit of the Meta fake news control center in Torre Glòries. So, that will further diminish the occupancy in Barcelona. First, Barcelona has a relatively minimal weight in our portfolio. So no big thing. But Barcelona is, of course, a complicated city at present because of the relative oversupply that we are experiencing in the area of 22@. It will take some time to recover. We have commented on a number of occasions. Barcelona is a strong market. So sooner or later, it will recover, but it's -- we need a little bit of patience there. In fact, even in rents, now Madrid is about to overtake Barcelona in average rents in the portfolio despite being more concentrated in prime CBD, the one in Barcelona. So very interesting in the performance we have observed in Madrid in recent times. Logistics, down about 60 basis points to 95.8% from the 96.4% we were at the end of the year. The like-for-like is weak, 0.6%, but this is mainly due to the net variation in occupancy because much to our surprise, the release spread has been super strong at 6.2%. We still need to see what happens in the second, third and fourth quarters to check whether this is a reflection of something or it simply -- it has happened a little bit randomly. I mean we need to check the consistency of this figure in the coming quarters. Shopping centers is a rocket. I mean, 6.1% rent like-for-like and a lease spread of 7.4%. We told you that we would start managing yield and we are doing exactly that. Despite this, the sales have gone up so widely that at the end, the occupancy cost ratio continues going down. But anyway, we will continue pushing a little bit in rents and trying to normalize more the performance of the portfolio. If we go very quickly through the different asset classes, and I will only stop in data centers because of its importance. In offices, what is -- what strikes the eye is basically the very good performance of Madrid, now reaching 94.6% with more than 1% change in the period and Barcelona, which has gone further down by about 3 percentage points. Lisbon has gone down by 4 percentage points, but don't take that as a reflection of anything. It's simply that we have suffered the conjunction of the exit of BNP Paribas in the Arts Tower and some of the consequences -- delayed consequence of the exit of Galp in Torre A. But at the pace we are reletting space there, this will be somehow normalized towards year-end. I think the local team there is doing an excellent work in rebalancing the portfolio. In terms of as a curiosity, the once demeaned A1 corridor in Madrid is now doubling. I mean, it has moved from Cinderella into princess. So now it is occupied above the average occupancy of the portfolio, owing to the fact that part of the infrastructure problems that were endemic to that area, particularly the traffic jams because all the residential around have been built, but no changes have been made to the infrastructure. Finally, the municipality of Madrid has carried out some changes in the North traffic hub, and this has reduced very significantly the traffic jams in the area. And the proximity of Operación Chamartín is also enticing more and more real estate managers in companies to find a slot there because at the end, it's going to be very close to where the music will play in Madrid for the coming 25 or 30 years. In logistics, well, as commented, very good lease spread and pedestrian growth in like-for-like terms owing to GXO, is commented, as a big shed. We are now negotiating with three, four counterparties, but it will take time. I mean bear with us because it's a big shed and the market is not also -- is not in its best moment. So we need to make sure that we sign that back and go back to a super high occupancy, if at all possible. Good news, however, is that we continue leasing well our work in progress. I mean we are 178,320 pre-let or ahead of terms. It is important to make the breakdown. It's 174 pre-let and 4 head of terms. So it's basically all pre-let, which is interesting. And even in the noncommitted, we are working on reducing the total number of 182 by about 60 because we are in conversation with the tenant and reduce the noncommitted to only 120, always with the idea that has been conveyed on many occasions to you of finishing our land bank in logistics and waiting for the next cycle comfortably full and cash flowing. This is what we want to do. At the same time, in ZAL Port, we were able to get two additional plots there and one has been pre-let to Lidl and the other one, which is in the airport of Barcelona has been pre-let to Logista. So that will add capacity in the super prime location of ZAL Port in Barcelona. In shopping centers, Spain is between the increase in population, the marginal propension to spend and the macro, it's really performing as like a rocket. The sales -- the increase in sales of 10.3%, which will strike some of you as too high. It's -- part of it is owing to the fact that Marineda, the extension of Marineda has been in operation in the period, and that has added about 4 points to the figure. If you want to do it like similar to like-for-like, it will be in the region of 6.2%, much more in line with the market statistics and what some of our peers are reporting in Spain. Footfall, however, is only plus 1.5%, and this is owing to very small things, difficult to explain like, for example, in Barcelona, we have works in Plaça d'Espanya which are complicating the access to the shopping center. In Larios, the Spanish high-speed train got suspended for a long period. And therefore, access to Malaga was completely complicated by the public authorities. And I mean, we have a number of other situations in the portfolio. But we continue to see a very interesting footfall flow in our shopping centers. Data centers. Well, for Phase I, Madrid Getafe 01, Barcelona Zona Franca, and Bilbao Arasur 03, the first buildings are now fully equipped and fully let, and they will reach EUR 97 million gross rental income in the year of stabilization, which is '27 as anticipated to you a long time ago. Going into the details, Barcelona Zona Franca has been repowered or is being repowered. All the machinery has been received, and we are doing now fit out. There's been a change in layout of the client, and it's -- the fit-out is going like 15 days delay that's not relevant for cash flow, and we will start receiving cash flow in the third quarter of '26 this year. Regarding Madrid Getafe 01, well, it's now fully let to 1 NeoCloud and two 2 Hyperscalers, which basically host in the data center POPs, I mean, cloud points of presence and are occupying an interesting part of the data center. There, we are finalizing the power connection works. I mean we are -- at present, we are big in trenches, I mean, in very plain language terms. And we expect to be ready in October '26. So by fourth quarter '26, the main client, the AI client of the center will start paying full rent. A repowering opportunity has arisen that will -- should allow us to go back to the 70 megawatts of Phase I that we told you in April 2022 in our Capital Markets Day. So if we have the opportunity to acquire 10 megawatts of electricity that will give us 6 megawatts extra in IT terms. I know it's a small number, but it's symbolically important for us. If we can pinch it, the repowering should be ready by fourth quarter '27. And that means that in terms of gross rental income, this year, it's set. I mean, we will receive around EUR 66 million in rents. Next year should be EUR 97 million. But then in 2028, there will be a significant jump. If the repowering happens, we will go to circa EUR 110 million, including the fixed step-ups of the existing contracts. So very, very interesting for the company. Then for Phase II, construction is progressing as commented, as planned. And the flow of pre-lets, it's promising. So in Arasur 2, it's fully let. The first batch, 20 megawatts should be producing money by beginning of '27. The second batch, 28 by mid-2027. Very cool machinery, I mean, really state-of-the-art, liquid cooled and what is more important, fully back to back by the client. So they have now final clients for all that power. I mean they are already sold in that. For Arasur 01, construction is underway. Pre-leasing is in advanced negotiations. What this means, basically, that we have done something special in this occasion. We have signed a reservation agreement, which has as an exhibit all the technical documentation and the full form legal documentation, the lease contract. However, the lease contract contains bracketed terms for the delivery date. So three things can happen. If nothing happens at a certain point next year, which I will keep for me, okay? The contract will become fully binding for both parties, and it will be let. If we have to change the delivery dates, and I will explain why, and the counterparty accepts, again, it will become fully binding on both -- for both parties. And if we can find a language which is sufficiently flexible for delivery dates, which will need the collaboration of the final client of our power taker and NVIDIA for the delivery of the chips, if we can find something which is amenable for the four parties, then eventually, we will move into full, let's say, full form contract signing. So it's pretty much done, okay? So that is the situation there, and we are working. Why are we worried? We are worried for a very simple reason because we have our own working morale. We have our own sense of fulfillment of contractual obligations. But when you are dealing with our public administration or other counterparties, when you depend on third parties, of course, you are exposed. And what we don't want to do is screw up. We don't want to screw up not only because we are going to be imposed a significant penalty. Penalty at the end, you pay and that's it. The problem is that we are a relatively small company, and we are just starting in the world of data centers. We do not have the goodwill of the big American incumbents and screwing up on the delivery of one of our facilities is not going to be good for us. So -- and particularly if it is owing to a third-party intervention, it will be a pity. So what we are doing is trying to make sure that we are pretty firm on the delivery dates. Remember that in this case, the electric line comes through a photovoltaic plant that needs to be built -- is being built or will start soon to be built by Iberdrola, by the utility provider, really next door to the data center. But the plant needs to be built and the line will come through the plant through the data center. So two things need to be executed, the generation assets, the solar panels and the line. And this requires lots of authorizations from the Basque Country authorities, from the central government authorities, et cetera, plus the execution of the works itself. So this is why we have been a little bit very prudent in the way we have approached this potential pre-let. Regarding Lisbon 01 and 02, well, basically, construction is advancing significantly. If any of you fly often to Lisbon, you will see that the two buildings are now with columns erected. They will start closing walls in the second half of the year and roofing. So works are advancing pretty well. after finishing all the preparation of the ground, which, as you know, is a complicated ground in Lisbon. So very happy with the way it is going. For the IT capacity, the primary plan continues to be the adherence to the Portuguese gigafactory. However, this is now being delayed again. The RFP by the European Union might be sent to the different parties at the end of June or maybe July, and they expect to take a decision by year-end. Taking into account that we started in all this process in February 2025 with the intention to have everything decided by April, maybe May, maybe June, it's a 1.5- years delay and which is sometimes not compatible with private activities. I mean we cannot do things in that way. So we are advancing in some alternative conversations pretty similar to what we did with Arasur that you will need to bear with us for long while because those happen to be hyperscalers, and that means basically very, very long conversations, very slow processes. It will take a lot of time to really get to a happy end. But statistics are in our favor. We are talking to multiple counterparties and sooner or later, one of them will end up concluding negotiations. So we are happy with the way it goes. Plus, if at all possible, if we can enter a full lease of the whole campus, it's paradise for us. So if we can really lease the whole campus and accelerate the construction of 03, 04 and 05, that will be perfect. So this is why we are taking relatively long in Lisbon that the demand is very satisfactory. In Madrid Getafe 02, the demolition works are underway, should be finished by year-end. Very happy with the way they are going. And the construction license has been submitted the request to the municipality, and we expect to have it or to get it immediately after demolition works are finished. So maybe with a little bit of luck, early next year. We want to rush in this case again because this data center enjoys a couple of bookings and highly credible. So we believe it will go very well. The current finalization target, which is first 1H '29, it's about one half in advance of the one we told you about it's like 2 quarters ago when we feared that this will go to end of '29 with full cash flow in '30. With a little bit of luck, it should go to more like first half '29 and full cash flow -- well, again, for '30, full cash flow for '30, that partial cash flow already in '29. And regarding Madrid Tres Cantos, it is a small data center, but planning has been completed. We got urbanization permit, and we are now moving around. And that's it because we expect to have this ready in the first half of 2029. And we haven't started commercialization conversations because first it's very small. Second it is in Madrid. So I think it will lease relatively well. And then for Phase III, we are just starting in Bilbao 4 and 5. Well, we are waiting for the execution of the power infrastructure by our electricity provider, which, as you know, is Solaria. We have requested the construction license to the municipality, which in turn, will put in motion the different favorable reports that need to be received from the Basque government. But we are already readying the project in case we need to accelerate it. I mean, in our presentation of Phase III, we said first half '30, first half '31. This is simply an estimate. If we need to accelerate, of course, eventually, we want to be ready in case for some reason, the client or one client wants to get this slightly earlier, I mean, we want to be ready. Regarding Lisbon 3, 4, and 5, the construction license has been granted by the municipality. And we have started doing the groundworks very quickly. I mean we are piloting already, and we will go fast here. So again, we have delivery dates of 1H '29, 1H '30,1H '31. If we get a client, that can be significantly accelerated, but we prefer to keep the existing dates. I mean the guidance provided to you in our Investor Day in Arasur because we still -- we don't know what the future might hold. And if we end up selling or leasing 01 and 02 to different clients on a retail basis or wholesale colocation, then eventually, we will not rush in starting 3, 4 and 5. But if we do just one single lease, campus lease, then yes. And then Zaragoza Wind, well, we have submitted already the DIGA, the Declaration of General Interest for the Community of Aragon. And we are now moving into the PIGA, the Project of General Interest of Community of Aragon. It might sound like Chinese to you, but the interesting thing about the PIGA is that it moves the land from rural status into full construction license in as short as between 9 and 12 months in theory, unless something happens or you find skeletons from the Stone Age in there. But the idea is to continue progressing hand-in-hand with the local Aragon authorities, which are highly collaborative and have the PIGA submitted before summer for readiness of the project in 4Q '29 and the second building 2H '31. But again, it depends on whether we can find a client. And we have a lead for this one, so which is the -- because Bilbao and Lisbon probably have the same client that we have a lead, and we are working on the technical design specific to that lead. Outside the Phase I, II and III, we got the declaration of a project of regional interest for our Navalmoral project in Extremadura for Building 1 only. And we have obtained a little bit of power in there, around 30 megawatts utility, which are good for about 20 megawatts IT. Remember, for those of you who attended the Capital Markets Day in Arasur, that was called Phase IV, okay? Interestingly enough, we got some electricity, not a lot. But given that civil construction is not the lion's share of the budget of a data center, we might perfectly start the construction of Building 1 being ready to equip 20 megawatts IT on it once finished because it's the only way to have an earlier ready for service date in case we get clients. And this Extremadura project is raising a lot of interest in the market because of the sheer size it has. Of course, during the civil construction period, we might get the real electricity, the one that we have requested from the local authorities from the central source, central government authorities. We requested 2.0 gigawatts and good for around 1.4 IT. Whatever they give to us is going to be good. So we -- if we can electrify the first building and if we can electrify the second, because they are twins from a technical standpoint, we might start also the second if we get the electricity from the central government. And that's basically it. LTV, bond maturities, and this is all relatively plain vanilla stuff. So let's move into Q&A. And I'm sure it's going to be much more interesting. I mean I'm sure you will have interesting questions.

[Operator Instructions] We have the first question coming from the line of Marios Pastou from Bernstein.

So just firstly, just on the broader FFO growth of...

Marios, you are breaking off, your voice is metallic. Can you check your line or eventually, if you want, if you send us in writing the questions, I will read in loud voice your question and reply to it. Do you agree with that method?

That's perfect. Apologies, my line is funny.

Okay. So questions. FFO growth, circa 4% in the first quarter. I appreciate that it is early days, but is there upside potential to the flat guidance? Yes. Yes. Yes, there is upside potential. But I don't do this to me every year. I mean, the EUR 0.58 pre-capital increase are EUR 0.53 post capital increase. Let's stay with the EUR 0.53 for the moment. We will review the guidance in the second quarter. And in case we see that we are fine, yes, we will change it. But for the moment, let's keep the 53 because it is not easy. I mean this year from an interest rate perspective has been roller coaster, but it's a roller coaster that only goes up. I mean the 10-year swap rate is going up significantly, and let's see. I mean, financial expenses might give us a headache. The good thing, however, is that however quick, however fast financial expenses grow in the coming years, the top line is going to run much, much faster. So we are -- that puts us in a very good situation. I mean it's compared to what I see around me, I think it's a very good situation. The company -- I mean, it's been fantastic, it's been a big luck to find this vector of growth of the data centers because it will simply send our top line to the roof, and this is going to go much faster than any potential increase in the financial costs. Then an update on Portugal, how the dual track discussions are ongoing with tenants versus gigafactory project. I have already referred to that. I mean we are holding like a dual track through, and we might perfectly end up going through a private solution. The third building in Bilbao under advanced negotiation rather than booking able to provide details, yes, well, I have already given you the details. I will not be specific on who is the client, although it's relatively easy to guess. And then Navalmoral, what this means for Phase III? Well, Navalmoral is Phase IV, as we commented in Arasur in the Capital Markets Day, Phase III, it's a little bit flexible at this time around, could be increased by some things coming from Phase IV, some things coming from Tier 1 pipeline or eventually, if we see that, for example, in Tres Cantos, if we see that it's going to be impossible to meet the deadlines given to market for Phase III, we will move it to Phase IV. We need to get accustomed a little bit to the fact that having so many balls in the air, have so many things in execution, some things will happen. So you need to be prepared for that. Let's be adult in that respect. I mean if something happens, I will clearly inform all of you that we sometimes might need to do some adjustments.

So the next question comes from the line of Celine from Barclays.

Can I ask you two questions, please, on data centers? The first one is, can you give us an update on where you stand with the EU on your Portuguese scheme? And then secondly, what does that mean that Navalmoral has been declared project of regional interest? What does that mean effectively for you in your day-to-day?

Okay. Well, regarding the EU, as commented, Celine, the problem is permanent delays. I mean, as you know, the program was launched by the European Union. They called it gigafactories because they were trying to replicate the Stargate Program in the U.S., trying to find 1 gigawatt project across Europe. Then somebody raised their hand and said it's impossible, then they reduced to 200 megawatts, then to 100 megawatts. So finally, it's a 100-megawatt program. Locations moved from 4 to 5 in order to please more countries within the European Union. Now they have moved from 5 to 7. So little by little, it's starting to look like the lottery of the parish in which I go to mass on Sundays. So the boys do a run. And then the one that ends up last still gets a medal because it's super sympathetic and has a very good smile. So sooner or later, this program is going to be launched. Apparently, it's going to be launched around June, July for decision-making before year-end. Okay. We can no longer be super faithful on that. So in Spain, we are completely out of this. I mean, we were told that we were not needed. So we found our way in the private market. And in Portugal, we have remained loyal to the agreements, verbal agreements reached with the Portuguese government. We remain there to provide them with digital infrastructure in case they need it. If they tell us they don't need it or the situation continues being delayed, we might move perfectly into a private execution because at the end, we have to manage frugally, your money. I mean -- and we cannot play with that. Then regarding Extremadura, the declaration of regional interest basically means technically that all periods for licenses are reduced to half. And therefore, everything gets some sort of like a fast track, and it moves significantly quicker. But in practical terms, more importantly is that the Junta de Extremadura is now clearly trying to help the project, and they have allowed us to get some power, which is testimonial is not significant on 45 kVs. So it's not super high-quality power, but it's okay. And will give us the possibility to illuminate 1 block out of the 5, if we do a B100, if we do a W96, the blocks are 24, so partially 1 block, 4 blocks or 5 blocks. I mean, depending on the type of building we finally build there. But we will have some electricity. And as such, we can start serving clients there, which is important because clients normally deploy system engineers in the site, et cetera. And once they are building -- they are working with you in a building, which is occupied only by 1/5, then the second fifth, third fifth, fourth fifth and five fifth is normally much easier to place with the same client. And Extremadura is raising a lot of interest among the clientele.

The next question comes from the line of Florent Laroche from ODDO.

I would have three questions, if I may. Maybe the first one, you mentioned the valuation of data centers, think that maybe we can expect some substantial revaluation. I understand this is maybe because of changing in the scope with the one. But maybe can you give us maybe more color on how we can anticipate that impact on the valuation for your data centers in Achuana? (sic) [ Arasur ]

Let me wait for 1 quarter. I mean, as you know, I mean we were carrying that land at cost in Lisbon was as close as you can get to 0. So yes, it will be a significant impact. It is not for me now to evaluate the impact. It will be the appraisers. The appraisers will do their job. They will tell us how much they believe. But yes, there is -- clearly, it's going to be an improvement compared to current situation because current situation is 0. So very, very interestingly. And I believe also there will be some revaluations in existing assets because of certainty. I mean, clearly, the certainty now is full. And as such, I'm sure the discount rates, et cetera, might be modified in our favor. Regarding other asset classes, it remains to be seen because the performance is very good, which is -- should move valuations up. However, interest rates are going up. So I don't know which of the 2 will prevail. I mean, only God knows, we will wait. But as commented with you on some occasions, whatever for the next 7 years, whatever happens to the traditional portfolio is going to pale in comparison to the value creation that you are about to witness in data centers. So don't be too worried about the traditional asset classes.

Okay. And -- so maybe my second question would be on Phase II and Phase III about the speed on how you spend the CapEx today. So you have provided us a very detailed review on all the assets. But in terms of CapEx spending, so how is it are you spending this faster? Or in line with your initial plans?

At present, faster. I mean last year, '25, I think we executed like EUR 900 and change million compared to EUR 800 million we had in our budget. So we were slightly faster. And in '20 -- talking -- very importantly, talking about committed, which, again, we have explained in some occasions, we consider committed CapEx like already spent because we don't want to run into the risk of insufficient funding. We don't want to make an equipment request from Vertiv and then discover that we don't have the money at the time it arrives to pay for it. So we kind of block the money of every equipment request we make. And hence, why we have a relatively significant amount of cash. What we expect for '26 is a little bit of the same. I mean we expect '26 to be also fast in terms of CapEx deployment. The cash at banks we have at present, which is like EUR 1.75 billion. If you take out the bond repayment, EUR 850 million change or less. And the speed of deployment of CapEx in data centers, we are going to finish the year -- if you talk about committed, more or less at 0. If you talk about real money at the banks, we will have money at the banks because part of it will be committed, but not yet spent, okay? But yes, we are spending pretty fast. And in that respect, the arrival of David Martinez, it's helping us a lot because he's making a significant effort of industrialization of the construction processes, which is important in many aspects. One of them is, of course, speed of execution. But the other one is to combat potential cost inflation episodes we might see in the future. At present, Europe is relatively tranquil for the moment because there is a lot of bulls***, a lot of noise, but very few people is really building things. But sooner or later, that situation will change, and that might strain a little bit the commitment queues with the main suppliers. So you need to be mindful of that and try to make sure you slot your things in the right moment and you get your equipment and you might need to equip part of your things in advance for which we are prepared. And we have lots of sheds everywhere in Spain. So we have a lot of storage capacity for equipment that we might receive.

The next question comes from the line of Veronique from Kempen.

Maybe first, I was hoping you could give any color on your stance towards the balcony portfolio. You mentioned you're not mentioned anymore in the last four bidders, but you were quite vocal on the past on your interest. So happy to hear if you can give any color on it.

Well, in principle, we couldn't agree on pricing. So we are technically out. And that's it. I mean they are running an investment banking process with long list, short list and et cetera, and we are not participating.

Okay. That's very clear. And then maybe second question, just curious, given what's currently happening in the Middle East, have you identified any new potential risk or already encountered delays or cost inflation for the data center pipeline?

Not for the moment, as I was commenting. The fact that most -- virtually all our supplies come from OECD countries. And the only thing that comes really from a far supplier is Japan, but it's made in Spain, in Zaragoza, the transformers. It's going well so far. Things -- I mean, collateral effects we are seeing from the Gulf crisis is more interest in Europe by hyperscalers because a number of data centers in the area have been hit. Data centers are, of course, centered in many crisis because they host data, which are vital for today's society. So one, two, like three data centers have been hit. So nothing serious has happened, but that has sparked even more interest now in European locations. And that's basically it. I mean I think we are okay for the moment. Of course, we are not okay. I mean I hope the situation there is resolved sooner rather than later, and we can go back to a normal life.

Okay. And one last question. So you now received the 30 megawatts of power for Extremadura. Is it somehow linked? Or does that give you any view on when you can obtain the remainder that you requested? Or is it something completely separate?

It's completely separate. It's completely separate, Veronique. It's a pity, but it's completely separate. This is electricity obtained through the distribution network through Iberdrola, who is the electricity supplier there. And the one we have requested in Arañuelo 400 kV is electricity requested to the transport system to the national grid authority, to Redeia. So, two different sources of electricity, and one thing has nothing to do with the other. We hope the power contest is called sooner rather than later. We provoked it. I mean we are ranking first. There is enough electricity in the substation. So we are hopeful that we are going to get a very significant chunk of electricity there. But the truth is that we deposited the bank guarantees on February 25, and we are still waiting. And hopefully, during the year, we should know more or the power contest should be called upon because it will be paradoxical that we end up receiving a lot of electricity earlier in Portugal than we are receiving in Spain. In, as a Spaniard, it kills me, but it is probably it's the situation.

The next question comes from the line of Ana Escalante from Morgan Stanley.

I'd like to ask a question specifically on construction cost inflation. I know that you said that so far, there's nothing especially that you should -- that you are seeing or that you think you should mention, but just thinking maybe a bit more long term. I was wondering if we were to see that inflation in oil prices feeling into construction costs, how does that impact your CapEx plans? Would you rather increase of CapEx because probably you would be able to transfer that to higher rent for data centers, meaning -- or would you rather just do less megawatts and keep to your original CapEx commitments?

Well, it depends on the relative impact. I mean, so far, first, Phase I is completed. Phase II, I would say, 80% to 85% is ordered at fixed price and being received as we speak. And in Phase III, of course, yes, it will have an impact. That impact in our opinion, for the moment is confined to steel, but not significantly concrete. And then the oil cost, I mean, our fuel costs for groundworks, et cetera. But as you know, the civil construction is around 25% of the total budget of a data center. So unless the movements are gigantic and we move into a 20% deviation, it shouldn't impact that much our total figures. And rents continue playing in our favor. I mean we are seeing our rents in the market above what we have underwritten for Phase II. So rents might give us a hand if that happens.

That's super clear. And maybe also on kind of supply chain bottlenecks. Based on the conversations that you're having with your prospective tenants, would you say that they are a little bit worried about not receiving their equipment in time or maybe the market just not being able to supply to all the demand for equipment that all these companies are placing in?

Not for the moment. In fact, NVIDIA, if any, they have normalized a little bit the delivery times that at some point were really, really long, particularly with the initial deliveries of GB200, there was a significant bottleneck. But now they are little by little normalizing. They have increased their production capacity. So this is good. What the clients are not getting is IT. I mean, IT power, data center capacity. So we are seeing now that in some cases, they are trying to push for longer-term contracts, which is a clear sign that they are under certain stress to secure IT capacity. I don't surrender. I believe at some point in the future, we will be able to charge back some of the common expenses. At present, as you know, the system is they pay you a lump sum and they forget. So there is a gross to net conversion, which is significant and much higher than in other asset classes. I believe this is -- eventually, this is one of the things that could change in the future because unless the rent inflation continues going up and up and up at some point, maybe the rent inflation stops, but they start accepting some chargeback of part of the expenses, which is the same. I mean the net effect is an increase in NOI, which will be good for all parties. But the -- I mean, from what we see, there are no signs of abatement of the demand. In fact, up to now, everyone was saying, what do you have ready for '26? Because '27 and '28, I still need to see whether I get clients or not. Now people are starting to ask what do you have ready for '27, '28 because they already have visibility on back-to-back clientele up to '28. And I'm sure next year, that benchmark is going to move to '30 and eventually '32 because particularly in Europe, we are not building capacity at enough pace to cope with the current demand.

The next question comes from the line of Stéphanie from Jefferies.

I would have a couple of questions actually. The first one would be a follow-up question on shopping centers. We clearly see that there are great spots currently in Spain. I was wondering about your strategy for this segment going forward, looking at acquisitions. Do you contemplate more acquisitions, so more opportunities outside of the balcony portfolio there? And what about disposal in front of that? Do you have disposal budget?

Okay. Look, one thing in reality is pretty much tied to the other. In terms of shopping center, our strategy is very simple, is to try to manage our yield through rents. We are trying to push a little bit of rents in order to send back the occupancy cost ratios to, let's say, more normalized levels. So we are trying to basically extract a little bit more cash flow from that portfolio, benefiting from the current federal situation in constructions. Regarding acquisitions, very difficult. I mean it's very difficult to find portfolios that really makes sense for us. Balcony, there were 2, 3, 4 assets that were really interesting for us. But this is just one opportunity that we looked at on an opportunistic basis. It's -- I'm not sure there will be many more like this in the immediate foreseeable future. which again brings the question some of you make to me sometimes that why didn't you buy Unibail leftovers, et cetera, because we run a listed company. And in a listed company, you can be contrarian, but only this much contrarian -- because if you are super contrarian, people will kill you in a public place because you are shopping centers, they are about to disappear in the U.S. And sometimes you need to take decisions a little bit delayed in time because you need to be cleaner than the cleanest. So -- but yes, I mean, we will -- if there is an acquisition opportunity in the market, we will have a look at it. If not, we will simply enjoy our portfolio. And there will be many more cycles in the future. There will be other situations in the future. 10 years from now, the online sales will start going up again. Now they are completely stalled at single-digit growth figures that maybe in 2035, online sales skyrocket again and people start being afraid of shopping centers disappearing, then maybe we venture into some acquisitions. And I will remind you of what you told me about buying more shopping centers because that would have been a good idea had we executed it like 3 years ago. Anyway. And then disposals, we will continue disposing of around 1% of our portfolio per year. I mean the budget for this year is, say there is like 130 or something like that. We will try to comply with our internal disposal budget. And we will have accelerated a little bit in case we have bought the balcony portfolio. If we don't buy the Balcony portfolio, then why selling income? I mean we will keep it. I know some of you say, why don't you sell all the traditional assets and become a mono-asset class pure-play data center company. That will be super cool. We will be the coolest managers in Europe for the next 10 years. But then one day, when data centers hit the wall because at some point, there will be maturity and there could even be oversupply, then what do we do? You get out of our shareholding list and our share price plummets, it will be a pity because the company at present is pretty well balanced, and we will endeavor to continue keeping that sort of balance in our income streams.

Okay. And my second question relates to logistics. So clearly, we can see that one departure can weigh much on operating KPIs. So I was wondering if you have any other tenants like GXO that will leave in the coming quarters. And actually, I will have the same question on offices. What are your main leasing challenges in the coming quarters in terms of tenant departures?

Okay. One very important point in this respect, Stéphanie, is that we calculate occupancy based on GLA square meterage. So as a consequence, any departure in logistics, of course, is meaningful for the total occupancy of the company. However, from a rent perspective, it's like leasing 1,000 square meters of offices, okay? So don't be carried away by what happens in logistics in square meterage terms. We do not expect to have any further departures, particularly this year. But at some point, of course, the portfolio is a moving portfolio, and there could be in the future departures in logistics that we will try to replace. When the cycle is helping us, it's easier to replace. When the cycle doesn't help you, it's more difficult to replace. But remember, always, we calculate occupancy it's not economical occupancy. It's not financial occupancy. It is space occupancy. We calculate based on the space. In fact, as of year-end, we will start reporting the occupancy in data centers. And believe it or not, it will not move the needle that much because in the square meterage terms, the data centers are relatively minimal. So even though we bring into the equation at the end of the year, Barcelona 1, Arasur 3 and Madrid Getafe 1, that is like 60,000 square meters, 100% occupied. They will make up for the GXO share, but that's it. However, financially speaking, it is super important, okay? So likewise with offices, I think in offices, -- the team did a wonderful preemptive work in the past years, securing extensions of the biggest leases we have in the portfolio because we still have like 70% of our portfolio is multi-tenant. -- about 30% of our portfolio in offices is headquarter leases. And of course, they can mean a lot in case you lose the client. But most of that job has been already done. I mean we renewed Pricewaterhouse. We renewed Técnicas Reunidas. We renewed Técnicas Reunidas -- we renewed in due time Endesa. So we are in a relatively comfortable position. And we are pretty mindful of the fact that for some of those headquarter leases, there's a bilateral relationship with clients. So we highly value the bread and butter nature of their cash flow. So if at all possible, our first idea is always to renew unless we want to vacate the building for conversion into residential or things like that. We normally -- our first intention, our first idea is always to renew. But we don't expect any exits this year in offices in our portfolio. For next year, we'll need to check the forward office report, but not for this year. I have been checking everything recently on the occasion of our Board yesterday, and there is nothing significant going out this year, neither in offices nor in logistics nor in shopping centers. I think the year, as commented on the full year results should be a year of relative continuation of the performance of 2025, in which you will see a significant hike in top line that will not be matched by a significant hike in cash flow owing to the financial expenses line. But other than that, it's okay. Then next year, you will see a much more significant jump in the top line that will be accompanied by some cash flow growth.

Thank you very much. There are no more questions. We thank you all very much for being here with us in this 3M '26 trading update. You know where we are at your disposal if you have any additional questions. Thank you very much.