SEGRO Plc (SGRO) Earnings Call Transcript & Summary

Well, good morning, everybody, and thank you for joining us at such short notice. I'm joined here today by a number of my executive colleagues, but particularly by Susanne Schroeter, our CFO; and Andrew Pilsworth, who's our Managing Director in charge of data centers. Plus also, we have our Chair, Andy Harrison, with us in the front row. We are really pleased to be here today to be able to share with you our conviction in SEGRO's strategy and its ability to deliver significant long-term value for shareholders. This is a company that we've built purposefully over many years. It's evolved repeatedly over its 100-year history, and we'll continue to do so. Simply put, it's not a static real estate portfolio. It's a scarce platform with a great strategy and a clear execution path to create significantly more value from here. On a personal level, I've been leading SEGRO now for 15 years, and I've never been more excited about the prospects and the opportunities ahead. So there are a few important messages that I hope you'll take away with you from today's presentation. The first is that SEGRO has an irreplicable portfolio concentrated in Europe's most supply-constrained urban and logistics markets. These positions have taken decades to build, and it would be exceptionally difficult to recreate what we have. Our operating platform is already driving strong performance from this portfolio with sector-leading like-for-like rental growth, but there is much more than that. Within our industrial and logistics land bank, the embedded value is substantial. It's based on sites we already own in some of the tightest and most sought-after corridors in Europe with development momentum now building rapidly. In data centers, we have a differentiated platform for sustained shareholder value growth, built on a portfolio that combines power, planning and land positions that are exceptionally hard to assemble. And we have this rare opportunity precisely as the European data center market is poised for significant growth. So together, we'll show you today how SEGRO can deliver value well above our current NAV and today's share price. We have the operating platform, the power and the capital resources to deliver this plan. So let's start by discussing some of the special features of our business. Our portfolio has been meticulously constructed in the most attractive locations where land is scarce and relationships really matter. Almost 2/3 of it is in Europe's major cities. We also have a super prime portfolio of larger logistics assets and more than 30 data centers in Slough, built by SEGRO, which is already one of the largest data center clusters in Europe. And we have a unique operating platform with deep local expertise, strong relationships and a proven track record of value creation that has and will continue, in my opinion, to deliver our performance in the years ahead. We deliberately overweight urban assets in the largest, most densely populated and congested cities, particularly in London and Paris. These markets are underpinned by population and GDP growth that has outstripped the broader economies of their respective countries. And this drives diverse and dynamic demand. Meanwhile, supply is heavily constrained by a lack of available land, planning restrictions and competing uses. These assets support superior rental growth and outperformance in total property returns as illustrated for London in the chart at the bottom right-hand side. It's taken many years to build the significant positions we have in Park Royal, Heathrow, Slough and Paris as well as in other European cities such as Dusseldorf, Berlin and Warsaw. Our big box portfolio is focused on prime logistics hubs and major transportation corridors, complementing our urban exposure. These are unique strategic sites, helping to transform European distribution networks. We've built this position through a development-led strategy and again, it would be very difficult for anybody to replicate this portfolio from scratch. Our powered Shell data center portfolio is also very rare. Many of you know that our Slough trading estate is one of London's most important data center locations with 0.5 GVA of operational capacity led to some of the world's leading colocation providers. Apart from its proximity to London and an abundance of power and a dense fiber network, the Slough trading state benefits from a unique simplified planning zone, the only one of its kind in the U.K. with development rights currently secured until 2035 and renewable beyond that date. Alongside the assets, our unique platform is also a source of differentiation. SEGRO has decades of experience developing and managing industrial assets with deep local knowledge and a partnership approach to stakeholder relationships. Our customer-focused approach ensures we consistently score high levels of customer satisfaction. And this, combined with a hands-on approach to asset management, has enabled us to deliver strong like-for-like rental growth and significant new income from our profitable development program over the course of the last decade. Our local in-market relationships and expertise enable us to find new opportunities to create value, manage our way through challenging planning regimes and deliver large-scale complex projects. Just allow me for a minute to bring the live with a few brief examples. So at Hays, the Heathrow, we sourced land off market. Worked with the local authority to agree a mixed-use development plan, we sold part of the plot for residential use, which enabled us to recover virtually all of our original purchase costs and then develop the remaining part as a successful urban industrial scheme. In Park Royal, our relationships and understanding of the local planning environment allowed us to secure planning committee approval in just 210 days at a time when others are waiting years to unlock data center planning opportunities in London. And our scheme at Inteporto Bologna, our relationships enabled us to secure an option over the land, obtain a building permit, a [ Grepre-let ] and only then complete the purchase of the land. And then finally, Park de Pete Caro also shows the value of patients and long-term relationships. We purchased this asset through the acquisition of Sofibus, a French-listed company, in which we originally acquired [indiscernible] . Today, it's a great industrial estate, which also contains one of our near-term data center development opportunities. So I've mentioned the importance of local relationships. And the quotes on this page, I think, make an important point. Value creation in urban markets is about much more than buildings. It depends on long-term partnerships with communities, local authorities, and other stakeholders. Our unique platform built up over a long period of time is a source of real competitive advantage for SEGRO. But let's now turn to look at the substantial embedded value that we have within our business and which is not reflected in our adjusted NAV. And I'll start with our industrial and logistics development pipeline. We have an exceptional land bank for industrial and logistics development. We have land on the balance sheet, expanding almost all of our existing markets and which offers GBP 282 million of future income based on today's rents. That's equivalent to almost 40% of our current rent roll. We also have GBP 147 million of potential rent from land options which give us access to further growth opportunities in a capital-light manner. Industrial and logistics development remains very profitable for SEGRO with an average expected yield on cost above 7% and meaning that there is significant value upside embedded within our land and option land positions. Now after a couple of quarter years following the COVID boom and the subsequent disruption caused by war in Ukraine, occupation momentum is now clearly improving. We saw this in the second half of 2025 in Continental Europe with the U.K. showing signs of following suit earlier this year. Those tentative green shoots are now turning into reality. In the U.K. big box market, for example, DTRE have reported that first half take-up this year reached 16.8 million square feet, the strongest period since the pandemic. In our own case, the work we've been doing over recent years to prepare our sites for development is beginning to bear a lot of fruit. As at the 30th of June, we had a record amount of rent associated with developments under construction or in advanced pre-let negotiations even higher than during the COVID peaks. In the last month alone, we've agreed terms on 4 large pre-lets equating to GBP 25 million of headline rent, with a number of further projects under active negotiation. Whilst there's no guarantee, all of those will convert into signed deals, I would go as far as to say that right now, demand feels as strong as in any period that I can remember in my 21 years with SEGRO. So taken together, the industrial and logistics projects planned on our existing land bank and on the land options already in place, represent almost GBP 430 million of potential additional income. We expect to start construction on projects representing more than GBP 150 million of potential rent within the next 2 years. CBRE have calculated the undiscounted value of this pipeline using today's rents and costs to be GBP 1.6 billion. This is not theoretical value. These are real sites in great locations with planning or zoning in place with demand improving and where our teams are equipped and able to deliver. And that's only the opportunity we controlled as at the 31st of December 2025. There is more to come as we replenish the pipeline, add new land, accelerate development if demand supports it. and capture further rental growth Next, we're going to look at the data center opportunity, which I know you're all waiting for with bated breath. And before I hand over to Andrew Pilsworth, there are a few key messages I'd like to share with you. Demand in key availability zones is expected to be strong, driven primarily by cloud and inference AI uses. Land with proximity to availability zones with power and planning are key to meeting that demand. These are in scarce supply and SEGRO is exceptionally well positioned across all 3 factors. We have a growing 3 GVA power bank, which has increased by half of GVA since we last reported, including 1 point GVA -- 1.4 GVA of near and midterm opportunity, giving us strong visibility on future value creation. Through the use of capital-efficient joint ventures, such as announced today for our second project with Pure DC, the cash contributions we will need to make to deliver that pipeline over the coming years are expected to be only GBP 1 billion in aggregate, which is eminently fundable from our own resources and is expected to be very profitable. Based upon our assumptions, this midterm pipeline offers us the potential for GBP 2.5 billion of valuation upside and GBP 460 million of additional income. Now to hear from Andrew, we have a further 1.1 GVA of reserve power in some very attractive markets, which has not yet been modeled or valued by CBRE. So let me hand over to Andrew, who will take you through the rest of the details on data centers.

Thank you, David. Firstly, to provide some context, European data center demand is growing very rapidly, driven by the continued shift to cloud and the emergence of AI inference workloads. Much of that demand is focused on established and emerging availability zones where proximity to major cities, fiber routes and existing data center clusters really matters. And that plays directly to see growth strength because our powered land bank sits in exactly those markets. By contrast, latency in sensitive AI facilities for training and some lower value inference workloads can be located in secondary and tertiary locations where land and power are less constrained. We are focused on core availability zones because they have the most enduring value. The 3 key factors in deciding whether a site can become a successful data center location for cloud and inference AI demand are proximity power and planning. Proximity means being close to major population and business centers and fiber infrastructure. Power is increasingly scarce across Europe and is by far the biggest constraint on new supply. And finally, Planning is critical because data centers are large, complex assets and local authority and community engagement matters. SEGRO has local teams on the ground with deep knowledge of their local markets, the planning regimes and strong stakeholder relationships. And this, combined with our centralized data center and energy expertise means we are very well placed to unlock data center opportunities in these fast-growing but complex data center market. Our JV model allows us to execute this strategy in a capital-light resource manner. Our partnership with pure data centers is an example of our ability to harness the strengths of our chosen partners. And through that relationship, we have been able to benefit from a platform of more than 250 employees with a deep expertise of designing, leasing, financing, and delivering fully fitted data centers. We have expertly built powered land bank with over 3 GVA of power capacity across key European markets, which gives us substantial embedded growth well beyond today's earnings base. This is an increase of 0.5 GVA since we last reported our Power Bank back in February. And 0.3 GVA is available to lease today with a further 1.1 GVA available by 2033, creating a visible near to midterm value creation opportunity. A substantial proportion of that near-term opportunity is in Slough where we have land committed power and planning in place through the simplified planning zone. And that matters because hyperscalers value certainty and speed to market and we can offer both. Beyond that, we have a further 1.1 GVA of additional power, and that is either longer dated or it's still being worked through our development plans. This slide shows quite clearly that the power positions that I've just highlighted are not hypothetical. They are allocated to specific projects in very attractive data center market. We intend to deliver the majority of the sites in 2 phases as fully fitted facilities, and we have clearly identified time line as to when power will be available for each site. There are 3 more projects in high-quality flat delocations in the third bucket. And that is only part of the opportunity. We have a further 1.1 GVA of power, again, across flat Dmarket and emerging availability zones, which we are yet to allocate against specific land. And just as a reminder, we intend to pre-let all of our data center projects and we'll be targeting long-term triple net leases with major hyperscalers where we do so. We have heard clearly that you want more detail on the data center opportunity and how and when we expect to create value from it. So today, we are providing that detail. Over the next 7 years, based on our assumptions around planning, power and customer demand, we plan to unlock 14 sites through development. And these could deliver around GBP 460 million of additional rent and almost 700 megawatts of IT capacity. We expect to deliver the majority of the planned sites via fully fitted data centers, mainly in existing flat D markets and selected fast-growing locations. And we intend to do these through joint ventures which allows us to benefit from our partners' technical development and hyperscale leasing expertise. It also allows us to share the capital commitment and risk. We know that our pipeline is extremely attractive to high-quality partners, and we are already progressing the first 2 projects through our joint venture partners with pure data centers. One in West London and as announced today, a second one in Paris, which will be our first data center on the continent. CBRE's view of the undiscounted value upside from our entire DC pipeline is GBP 2.5 billion. That GBP 2.5 billion value creation upside is before the additional 1.1 GVA of power that is not yet included either due to its long-term nature or as specific development plans are not yet finalized. As I hope is now clear, the sequencing and the pace of our data center pipeline is dependent on power, planning and leasing and not capital. It's also important to highlight when we can start to unlock that value. With a fully fitted data center, the income comes through at practical completion, and the first phase can take 2 to 3 years to complete but the value creation starts much earlier. Value starts to be created as soon as you secure power and planning. And just under half our data center land bank is held at industrial values, providing substantial upside. Signing the pre-let crystallizes a significant proportion around half of the total development profit. So whilst these are long-term assets, there is a very real near-term value creation opportunity as we move projects through the development process. Our JV structure is highly profitable and has the potential to generate a very attractive return on capital. It also materially reduces SEGRO's cash equity requirement, which means that we can fund the pipeline through our normal capital recycling program supported by selected powered land sales. This is an illustrative example of exactly how our JV model does that. So say, the total development cost is GBP 860 million, including the powered land that SEGRO has contributed into the venture. And assuming a 70% loan to cost, debt funding is around GBP 600 million and total equity is around GBP 260 million. Our partner contributes half of that equity SEGRO contributes the powered land at fair market value plus in this example, around GBP 60 million of cash. And that is the reason the structure is so attractive. It allows us to crystallize the value of our land and power, access specialist delivery expertise and generate a strong return and substantial value creation from the cash we invest. And with that, I will hand over to Susanne to talk about what this means for earnings growth and the balance sheet.

Many thanks, Andrew. And I would like to start with a very clear message. SEGRO is in a position of strength. We are not capital constrained. We can deliver our strategy without raising equity, and we are doing it while maintaining balance sheet discipline. Before we dive into the capital plan, let's look at why we are so confident about the road ahead. We have published a trading update this morning, and occupier markets are strong despite the conflict in the Middle East. We are seeing strong letting activity across our existing portfolio and are making good progress in capturing the GBP 33 million of reversion available in 2026. Our current and near-term development pipeline stands at GBP 90 million, which is the highest level ever, giving us tremendous visibility of future growth. On the back of this, we have now narrowed our CapEx guidance to the top end of the range provided earlier this year. Our updated CapEx guidance for 2026 is GBP 500 million to GBP 550 million. The recent increase in development activity will also have a positive impact on our 2027 CapEx numbers. I would like to highlight that we are self-funding this growth through an active capital recycling program, while crystallizing profit. Year-to-date, we have GBP 308 million of disposals completed or exchanged above book value. We also remain very focused on controlling our cost base. Our EPRA cost ratio, excluding share-based compensation, at half year is below 18%, and down from 19.8% at year-end 2025. Our pro forma adjusted NAV per share amounts to 905p, reflecting movements in asset values between the 31st of December and the 30th of June. It is not our review 30 June NTA as it does not yet reflect our half year profit, dividends and other adjustments like FX movements. These when combined, are not expected to be material. Our updated reviewed NTA per share will be published with our half year results at the end of the month. The small decrease in asset valuations was driven by our U.K. portfolio as our new U.K. value or Cushman & Wakefield, have taken a slightly different view on certain yields in some of our urban assets. Now let me talk about how we fund our growth and what it means for earnings. SEGRO has 3 powerful funding levers, which we can access at the moment, a strong balance sheet, capital recycling and third-party partnerships. We have a strong investment-grade balance sheet that provides us with optionality. The secondary trading spreads of our bonds are the tightest among our industrial and logistics peers in Europe. We have access to the euro and sterling bond and loan markets at competitive rates. Over the past 5 years, we have raised and total GBP 10 billion in euros. And our ratios give us headroom. We have already refinanced all 2026 maturities, and we expect net debt and LTV to fall by the end of the year, aided by disposals and the new big box joint venture. Turning to capital recycling. Since the beginning of 2021, we have recycled about GBP 2.2 billion, returning a 10% gain on disposal and shifting capital from lower return mature assets into higher yielding development. Every asset undergoes a rigorous annual review to guide that process. We have guided to disposing at the upper end of 1% to 2% of gross asset value in 2026 and we have already completed or exchanged GBP 308 million year-to-date. Finally, we have established party capital relationships that allow us to share capital intensity and to reinvest proceeds into attractive opportunities like our data center pipeline. We have successfully partnered with PSP on our European big box joint venture sell and with Pure on 2 fully fitted data center joint ventures. Last week, we have also announced our new U.K. big box joint venture. The new joint venture signaled strong confidence of international capital in the U.K. logistics market, and in SEGRO portfolio in particular. The joint venture is a fantastic example of how we increase investment capacity efficiently. It's a GBP 1 billion structure seeded with 3 of our outstanding sites. Radler, Northampton and Covetry. These sites include 701 million square meter 701,000 square meters of developable logistics space. and 220,000 square meters of standing assets across the U.K.'s most strategic distribution corridor. This JV allows us to deepen our investment capacity and showcases strength of our development and asset management platform. It reduces our land drag by bringing in partner capital and project level financing to fund future development. It will also increase our fee income because SEGRO will act as manager of the joint venture, providing development, asset and property management, together with financial, administrative and advisory services. The key takeaway, executing our strategy does not come at the expense of balance sheet discipline despite higher CapEx. Assuming the industrial and logistics as well as the DC pipelines progress as outlined by David and Andrew and we execute our new U.K. big box joint venture, we remain well inside our LTV and net debt-to-EBITDA targets which support our BBB+ rating by Fitch. The look-through LTV improves in the medium term before returning to the current level in the low 30s by 2035, even after full pipeline build out. Our net debt-to-EBITDA will strengthen over the same period. These numbers do not assume any disposals of completed data centers, which would give us further room to maneuver. David and Andrew have spoken about the value creation from our strategy. Let us now focus on the income opportunity. There is more than GBP 1 billion of income upside on top of the GBP 755 million of current rent passing. We can create GBP 220 million of additional income from our existing portfolio, driven by rent-free roll-off, vacancy leasing and reversion. The industrial and logistics development pipeline is expected to contribute up to GBP 429 million from near term, long term and options land and data centers. Our strategy assumes a GBP 460 million base case with significant further upside from a growing pipeline. The 3 buckets do not yet reflect the upside from further ERV growth selective acquisitions and here's what this means for earnings. Earnings per share will grow in the coming years from 36.6p at the end of 2025 to 50p by 2030. This growth will be supported by like-for-like rental growth, development completions, cost efficiencies and growing fee income. We expect data centers to become a true growth driver moving from 7% of rental income in 2025 to circa 15% in 2030 and to above 30% by 2035. This creates significant structural growth beyond 2030. And to wrap up, we are not capital constrained. Our EPS will grow in the coming years, and our data center strategy provides an additional layer of medium and long-term earnings momentum. This is a credible deliverable growth story grounded in financial strength and operational excellence. And with that, I will hand it back to David.

Thanks very much, Susanne. And indeed, let's now turn to the final section of today's presentation. And one is that we believe the Prologis' proposal is opportunistic one-sided and inadequate. Firstly, it's opportunistic. It's been timed to take advantage of a period of share price weakness since the start of the Middle East conflict despite SEGRO having started the year strongly, outperforming Prologis up to the start of the conflict. Between the 27th of February and the 23rd of June this year, however, SEGRO share price declined by 12%, whilst Prologis increased by 2%. These relative performances are pretty much in line with what happened to the broader European and U.S. REIT sectors, respectively. And their approach comes just as momentum is improving across our markets. And just before some of the most significant value in our pipeline is expected to come through. As we've already mentioned today, pre-leasing activity within our industrial and logistics platform is very strong right now. And the proposed U.K. logistics venture that we announced last week is a clear vote of confidence in the outlook. We're also now moving from potential opportunities in data centers to execution. Secondly, the Prologis proposal is one-sided. I'm not surprised that they're interested in our business. SEGRO has unique qualities, including very high urban weighting, while Prologis is more focused on larger logistics assets. SEGRO also has, we believe, a much greater relative opportunity in data centers. When measured against market value, SEGRO's 3 GVA of data center opportunity represents around 5x the relative exposure of Prologis's disclosed 5.6 GVA Power bank. That matters because we believe the European data center market is entering a period of exceptional growth whilst, as we've said, sites with the 3 Ps of proximity, power and planning remain scarce. Under the Project proposal, shareholders would be exchanging their 100% ownership of our business with all the qualities and the embedded upside that we've outlined today, diluting it to a materially lower share of a much larger and different business with different asset waitings and one that is -- that has a much smaller relative data center opportunity. So I can see why -- why the proposal is attractive to Prologis, but it's one-sided transfers value away from our shareholders. And then finally, I'm going to explain why the Prologis proposal falls a long way short of our views on value. There are 4 main components to this. The first is our NAV. The next is the value upside we expect to accrue from our exceptional industrial logistics and data center pipelines. And finally, there are a series of additional components of value, which are not reflected in our NAV and which should be paid to our shareholders by any acquirer. So let me explain. We start with our adjusted pro forma NAV of 905p. This reflects, amongst other things, the value of the undeveloped industrial land that we own today. However, it does not reflect the additional value upside that will accrue to our shareholders from developing this land into completed industrial and logistics assets. That's the second block you see here. We showed you earlier that CBRE have valued this upside at GBP 1.6 billion, which discounted to today represent 103p. Next, we have the GBP 2.5 billion of value that CBRE have attributed to our near to midterm data center pipeline associated with the 1.4 GVA of allocated power, and that equates to a further 139p of discounted value. For further 1.1 GVA of power in respect of which we are still finalizing detailed development plans has not yet been modeled nor valued in CBRE's numbers. However, it is valuable, reserve power and represents a significant additional opportunity embedded within our business today. On top of this, there are meaningful other components of value that any acquirer should pay for, which together equate to over 160p of value. The appendix sets out details, but just to highlight a couple, the cluster premium reflects our strength in locations such as Slough, Heathrow and Park Royal, where we have an ability to exercise greater control by virtue of our market position. Then there are certain costs and taxes that any acquirer of individual assets would incur in assembling this portfolio through individual asset purchases, which a buyer of the shares would avoid. Our shareholders should be paid for this. Beyond these blocks of value, we should note this is not a static business. The opportunity set continues to grow with our market position, local expertise and relationships, constantly creating new accretive opportunities. And lastly, I don't need to speculate on the significant synergies that Prologis envisages, but it is value that should be shared with our shareholders. So to summarize and conclude, SEGRO is a unique business. We have an irreplicable portfolio built over decades and a unique operating platform that continues to deliver value. We have a compelling value creation opportunity in both industrial and logistics and data centers. At a time when occupation momentum is building and Europe's data center market is set for exceptional growth. And importantly, we have the capabilities and the balance sheet to unlock that value ourselves. And our team is excited to deliver that opportunity for our shareholders. So thank you very much for listening, everybody. We're going to move to questions, and I'm going to invite Susanne and Andrew to come and join me at the front. Just whilst they're coming up please, everybody remember that since we are in an offer period, we're heavily restricted by the takeover code in terms of what we are able to say beyond what is already documented in our public statements.

But with that, who would like to go first?

It's Marios Pastou here from Bernstein. Just a few questions from my side, maybe kicking off with the data center strategy. It feels like this is another partial shift really in the strategy here moving more towards a fully fitted approach and moving away from your -- should we call it, the bread and butter powered shell approach. What has prompted this and the timing of it now? And in some ways, is there enough tenant demand out there, especially across your Slough platform to take hold of this fully fitted approach?

I'll let Andrew just answer the detailed question in a second, but I think it's a really great question. I would actually say we're not announcing anything new today. We're giving more detail of the plan that we already had established several months ago. And if you look back at what we've been saying about data centers over the last couple of years, it's been an evolving strategy. We made it very clear more than a year ago about moving into fully fitted space and that's built further momentum. So the plan isn't new. I'd say momentum has been building and so has our level of disclosure. We've heard loud and clear, as Andrew said, from investors, particularly over the last 6 months, that investors would like more detail on what our plans are. We were planning on sharing a number of details some of the half year results and some in the Capital Markets Day later this year. The only thing that's changed is because of the approach from Prologis we've accelerated the disclosure. But the plan, which was evolving and has evolved over a couple of years, hasn't really changed. It's just we're giving you more detail about it now. But in terms of the specific, what gives us confidence to actually deliver this fully fitted strategy and you asked specifically about Slough. I mean, Andrew, maybe you should talk about that.

Yes. Thank you. No, absolutely. I mean, first of all, I'll refer you to in terms of demand. I have never been more confident in terms of the weight of demand. I mean it is true to say it is concentrated not exclusively, but largely with the hyperscalers. But that -- if we just look at -- take part roll as an example, I'll come to Slough in a minute. But if you take Part Royal as an example, we've announced today that we are -- there's a small pool of hyperscalers, but we are in active discussions with 2 of them, which is obviously fantastic position to be in and a real bit of evidence of the strength of demand. And in Slough also, we have numerous discussions ongoing with regard to Slough, which is that is an absolutely location for data centers. We have the planning already in place through the simplified planning zone, and we've got a very strong position. That comes through the power on date for that is late 2029, 2030. But with the lead time to deliver data centers of 3 years, that means we will be in a window very, very shortly back end of this year, start of next year to start having active discussions in Slough and I've absolutely no doubt from the conversations that I have with DC operators and hyperscalers that there's really strong demand in Slough for fully fitted DCs.

And just a follow-up. I think the plan also now assumes that there's going to be no disposals of these fully fitted data centers or at least with the assumptions anyway. And I think originally, the plan was to crystallize those development profits once it was delivered. Is that also a bit of a shift? Do you now see yourselves as being a longer-term holder of these types of schemes?

It's completely -- that one is -- sorry, where you're going to go. Sorry, that's completely consistent with our wider strategy is we develop -- we will develop where it's the right thing to do and crystallize value. And then as part of our wider strategy, we'll assess the portfolio as to whether hold on to those assets for the income or recycle them out of the portfolio, and that will be no different with our fully fitted data centers. David, anything you want to add to that?

I mean we think we think we're going to have tremendous optionality. I think the point that we were trying to illustrate today is we've got the capacity to fund a number of these, but we have made it very clear in previous discussions that we would expect to recycle. We're not going to put huge amounts of capital in without greater evidence and proof of where liquidity and where pricing is. But the point about the presentation today is we're not reliant on that. in order to be able to deliver this plan.

Okay. One more, if I may, just on the upside potential across the logistics pipeline aside from data centers. I see you've also included the option land and you've also not factored in then the potential JV, which could come or it's been announced last week, it should come in the second half. How secured is that option land in terms of the likes of planning and whether you're able to actually unlock that upside potential? And as a side note, how much of that could then be given away potentially in the joint venture because there's probably an element of double counting kind of unlocking joint venture to a like which you then invest in the data center side of things?

Yes. There'll be -- I mean, there will be some of that option land upside that could go into a JV in the future. I mean the JV at the moment is on the existing land bank relates to landbank, that's already on the balance sheet. It's interesting when you talk about options, Radlett is a great example. Radlett is a site that we've been working on for 50 years, the most complex and difficult planning process we've ever been involved in, I would say, but the fact that we had it under option and came to exercise the option once the planning has been secured means that we're able to tie up large opportunities in the future without overcommitting our capital. And that's the nature of a number of the other sites that are in the optioned land, some of them are in the U.K., some of them are on the continent. They usually situations where we've either got an option or a conditional purchase and the condition really always is around getting planning. So that's how those work. But our history and our track record is that these things eventually come through into land on balance sheet and then they get developed. In terms of the impact of the JV, which hasn't yet been closed, of course, we've only announced we've agreed hazard terms on the numbers. They're not modeled, but I don't know whether you want to say anything about that, Susanne?

The numbers are tied back to the CBI valuation report, which is based on 2025 with the adjustment for the June valuation. The impact of the joint venture on those values is not material. So we've obviously done scenario calculations there. It doesn't have a material impact on the numbers of the CBRE valuation bridge.

But it will have some impact on the rental upside, which we haven't disclosed, but people can probably make their own estimates based upon the information we have shared. Tom?

It's Tom Musson at Berenberg. I think as recently as your H1 results presentation last year, you pointed to data centers to be developed between an 8% and 12% yield on cost. Today, in your illustrative fully fitted DC example, you're using a 9% yield on cost. So just below the midpoint of that previous range. Can you just give a bit of color on why is now the right number? And what's changed perhaps to bring that down slightly at the midpoint?

So yes, that wide range that we previously guided to, a big factor in that, particularly with power shelves given the level of length component is whether the land is held at value or industrial values. So the 9% we have given is a really good pivot in the market and it is a cost plus market for fully fitted data can so 9% is a really good is a really good point estimate for a sensible yield on cost for a fully fitted data center, assuming a market -- a market value of land.

Okay. And then just a question again on the NPV calculations of the future value creation in the logistics and data center pipelines. What discount rate is being used in those calculations? And do they differ between the logistics and the data center pieces?

It's an 8% real cost of equity discount rate that has been used the same for both the logistics and the data center pipeline.

And on the data center pipeline, over how many years are you discounting back?

Well, the pipeline of the 2.5, the 2.5 GVA that CBRE have put in their report, we've set that out in one of the slides. You can see the different buckets. But they are across projects. They're 1.4 GVA and they are projects that are contracted up to and including 2033. So 7 years' worth of contracted rent. Obviously, with data centers, some of the comes through a little bit later than that with the lead time.

Just some question on the Page 20, the data center detail. So you have different planning status there ongoing or in progress. If you can just specify what that means, please. You have also in the say, you're talking about sale for some assets available to lease by 2028. So if you can highlight when you envisage these sales and what would be exactly the model here. Third question would be you talked about MVAs on all this slide. What should we assume the key 2-megawatt IT? Is that a fixed key? Or is that something that is evolving from a PE perspective? Let's stop there and there is one more.

I was jumping in there, that feels like enough. If I if I -- and I might lead you to just remind me of those. Let me answer the last one first. So the ratio of MVA to megawatt IT. I mean you're right. Your answer in the question. That varies site-by-site. So you typically see somewhere around 1.2 to 1.4 for a PUE, the fact by which you reduce that is typical, but that really does depend side-by-side got planning for and also various technical factors around how much power is lost between the incoming power and utilized in the building. So it will vary.

Should we use 1.3?

Well, you can also see, if you look at one of the other slides in the deck, if you look at the -- you get a bit of a clear, although some of these are sold. So if you look at Slide 21, you'll see some of the sites there. We've got 1.4 GVA. And you'll see that there's a lower out, you can see the megawatt IT that we're using on the 14 leased sites. So powered land sales, you can see the 14 lease sites. So yes, it will vary side-by-side is the answer. Sorry, can you just repeat the other 2? So planning status, what's the difference between ongoing and in progress. So ongoing -- that's a bit of a softer one. Ongoing is a little bit more advanced. So that means we have an active dialogue where is in progress in the early stages -- the earlier stages of making that application.

The third one was, you have available to lease by 2028. You have a sales model, which is effectively third model, how are you envisaging that to work more precisely?

Well, in terms of -- so they are sites. So these are initial assumptions. We have put some for sale. They are generally -- so it's part of making sure we balance off recycling, but really those sites generally that we provisionally assumed we will sell tend to be in slightly less strong markets. So there's still great sites. But slightly less strong than those where we are planning to develop. And it will be the same process in terms of the trigger for timing. So when we have a firm power and planning position, we will be ready to contract on all of our sites, whether that is developing them mainly is fully fitted or offering them for sale.

And Jon, yes, that's consistent. That hasn't changed. I mean what we're saying is in super prime locations like London and Slough, and probably Paris, we're happy to put a lot of capital in because as Andrew referenced in his presentation, super strong locations, very scarce land, scarce power. In some of the more emerging markets, the working assumption at the moment and what's been modeled is that we just sell some of these land positions to others. But frankly, the market is changing so fast. So I suspect in 12 months' time, if and when we share with you an update on this, the market will have moved and we will have changed some of our plans on these. So it's adaptable, but it's all about putting our capital into the most resilient strongest markets and getting the best return on that capital.

But you would -- essentially, you would sell the land before pre-letting it. So you try to get it..

And we did -- we had an example of that a couple of years that we sold a great site to hyperscaler in Milan. We bought the site for something like EUR 8 million, sold it for over EUR 100 million at the time, it was the right thing to do, having secured planning and power. So it will vary by side according to where we think the supply-demand position is.

I had a last one, sorry, it's very quick. So on the power, you have some power on now and then some is 2028 or 2029 for the year available to lease now and thereafter. When can you realistically be in a position to have a discussion and then potentially sign an agreement with a hyperscaler versus when that power is actually technically granted versus power on?

Yes. That's really nice -- the answer is in terms of actually contracting, obviously, you can start discussions a little bit earlier than that. In terms of contracting, it's 2 to 3 years because you can have a really mean -- and you also need to have planning to have a really meaningful discussion as well. That's the other caveat. But when you are within 2 to 3 years because that's the typical time frame the shell and the fit out, you can have a really meaningful discussion and contract provided you've got planning, and you've got a firm power on date, you can do that 2 to 3 years beforehand because then you only need to have the energization power on at the point that you deliver the first phase that fully fitted and that is about 2 to 3 years from when you put a spade in the ground.

And who takes the risks on the delivery of the power because it's contracted, but then it needs to physically arrive to the site. Are you taking the risk on that?

Yes. That is typically a commitment that the developer will make. So you need to be -- that's why power is so crucial to all of this. So you need very confident and clear on the delivery of that power. And that's one of the reasons we mentioned that now we have a dedicated in-house energy team with really experienced experts who know about this know about the power business.

I thought we might get a few questions about data centers. Anybody else?

Thank you. Suraj Goyal from Greenstreet. So Hopefully, this will be a quick one. But you communicated the potential to develop fully fitted DC in Paris today with Pure DC. But as we know, planning can be extremely difficult in Paris. So how do you think about this risk to the time line for eventual income generation of this fully fitted DC? And I know -- I think you briefly touched on it. But would you target a yield on cost that's in line with Park oil?