Digital Realty Trust, Inc. (DLR) Earnings Call Transcript & Summary

Welcome to the 2025 Global Property CEO Conference. I'm Nick Joseph here with Mike Rollins with Citi Research, and we are pleased to have with us Digital Realty and CEO, Andy Power. The session is for Citi clients only and disclosures have been made available at the corporate access desk. To ask a question, you can raise your hand or go to liveqa.com and enter code GPC25 to submit any questions. Andy, we'll turn it over to you to introduce the company and team, provide the opening remarks, tell the audience the top reasons an investor should buy your stock today, and then we'll get into Q&A.

Great. Thank you, Nick, for having us back here. To my right is Jordan Sadler, Head of Public and Private Investor Relations. To my left is our CFO, Matt Mercier. I am Andy Power, President and CEO of Digital. Here at Digital, we support 5,000 customers, including the leading technology companies around the world and corporate enterprises in 50-plus metropolitan areas on 6 continents with their digital infrastructure needs, space power and interconnection. We've been a 20-year-old public company and been coming to this conference for a long time now and are on the back of a pretty landmark year for our company on multiple fronts which goes into, I would say, our future: One, first and foremost, we're accelerating our growth. That's at the top line, and that is especially at the bottom line with an inflection in our core FFO per share growth coming into this year on the heels of a record signings year, strong pricing power, record 600 new customers to our platform in the last calendar year. So acceleration to the growth is key, and we're not done yet in 2025. Two, we've been building our backlog to derisk that. So backlog sits at 20% of our revenue base at attractive returns, much of which is going into a $7.5 billion, largely 70-plus percent pre-leased development pipeline at 12 ROI today. In the meantime, we've also had great success at our most core and top priority of our 0-1 megawatt interconnection enterprise business, where we had a record third quarter in that category of new signings, followed by a record fourth quarter, record new logos that have dated back several quarters of 100-plus customers per quarter added to our platform and 25% increase in those new signs of that category year-over-year. In 2026, I believe, we'll be even better than that. And lastly, we have the balance sheet and the funding to support all this growth, having delevered the balance sheet to less than 5x, call it 4.8x net debt to EBITDA, $6-plus billion of liquidity. And on the eve of this conference, we announced the next evolution in our funding model with our first hyperscale fund achievement with the first 800 plus of commitments on the way to a $2.5 billion hyperscale fund.

Great. That gives us a lot to dig into. Just maybe first on the funds that you've created. Can you give us more details in terms of the size of the fund and how you plan to invest those dollars.

So taking a step back, Mike, this has been an evolution of our funding model for the last few years, where we saw an addressable market for hyperscale that was just tremendously capital-intensive and only getting bigger and bigger and bigger. That is certainly the heritage of our company that we've started at the hyperscale and worked our way into the enterprise colocation interconnection success. But we said we need more arrows in our quiver. We need a bigger boat to navigate this capital intensity. Historically, we've done it in more one-off ventures or partnerships on private capital, both stabilized and then development. And this is the next evolution of that funding model. This first vehicle will be seeded with a handful of stabilized assets and also some development assets and will be largely prespecified, but we'll have some unspecified capital as well to deploy along the way. We're super excited to having the first round of new LPs and I think we'll be adding more to that. And again, it's another arrow in our quiver to fund this extensive hyperscale growth. This vehicle by itself really U.S.-focused in nature.

And so $2.5 billion is the target for the fund?

$2.5 billion of third-party equity.

And can you lever the fund as well, so we should think about the capacity as significantly larger than that?

Correct, to double it.

So that's -- and so that's $5 billion that you can -- some specified, some unspecified projects put towards development and then you're able to lever those development projects themselves as well.

When I double it, that would include the leverage of the product.

It Includes, okay. Great. Just wanted to get a sense of the size of the opportunity. So -- and maybe relative to the opportunity, if you can share with us what you're seeing in terms of the size of demand and sales pipeline? And maybe just provide some perspective on how that compares to past years and how much is coming from AI?

Sure. So maybe split the categories into both enterprise colo and the hyperscale buckets and the latter being the lion's share of the capital at this minute. If you look back, as I mentioned, Matt and team and Jordan have been able to navigate the balance sheet down to a sub-5x leverage, which is quite amazing by itself, but even more amazing when we took our development capacity at the time, north of 50% up in terms of megawatts to now stand at a total project spend of $7.5 billion. So much of that is pre-leased. It's 70% pre-leased. Within that is colo capacity as well. So the hyperscale portion is even higher preleasing, including our North America region, that's close to 97% re-leased. As you turn the page from 2024 into 2025, I think we see a consistency of -- on the hyperscale side, large contiguous capacity blocks in the core markets has been very precious to those customers and then near the delivery, the more precious it is to those customers. And the reason why is, obviously, the larger the capacity block more well suited for their near-term AI needs via training, but in the core markets where we're servicing them, they see fungibility with their availability zone needs for their cloud compute as well. So you call it, get 2 birds with 1 stone when you come with us in these core markets in terms of their customers' needs. The mix of the customers are shifting. We're fortunate to serve 5,000 customers in whole and all the biggest customers that are really out there. And we've had a very diverse set of wins from these customers in this pipeline. And I think it's consistent relatively year-over-year in terms of pipeline.

Great. And then in terms of just recent developments, customer conversations, has anything changed as there's been some press around lower-cost LLMs or possibly 1 hyperscaler canceling some of the capacity that they were looking to take. Can you just share some context of how your conversations are going?

So I think there's 2 elements in there that I'll try to comment on somewhat -- somewhat related somewhat separately. The rumor mill and the noise is somewhat, call it, an accolades to our strategy or mirrors to our strategy, where in the last 2 years of AI becoming more of a main theme, we stuck to the core markets, the markets where we see not only robust but diverse demand. We stuck to a strategy of colocation enterprise, interconnection prioritization despite what looked like great opportunities at hyperscale at the same time. And we also widened out our customer list at the top of the queue. So if you look at our record $1 billion of new signings last year, the 3 biggest quarters had 3 different customers, all having the largest major signing in that quarter, none of which was our top customer and one of which was not even a top 20 customer. So those lists of major hyperscale customers that are coming to Digital has been expanding in recent times, which gives you diversity in demand. And I think that's very apropos to what we're seeing is rumors of one force maybe pulling back in the race, while other names are called pulling ahead in terms of demand. As it relates to the element several weeks ago with the advent of DeepSeek, I think I look at 2 things. I look at what our customers said and have done since that. You have a series of earnings calls with each 1 of those -- executives of those customers had CapEx budgets that were not decreasing or flat but increasing and pretty much unwavering view that we need to continue to invest in infrastructure in this AI arms race. At the same time, I also believe that the -- that news shows that there is probably a potential for greater democratization of AI and chips and GPUs down the road here. And hopefully, pulls forward what comes next and as AI moves, as we move from training, as we move to inference, it was moved to enterprise applications and use cases, which yes, it may have pulled it forward. I believe it's still a long tail of demand that's going to be out there. And I just look at that at our own use of AI at Digital as well, asking many folks at this conference of who's using, how they're using it, the applicability. I just think it's going to take time for those more complicated use cases to really roll out.

We should poll for room for question, seriously.

Jordan wants me to poll the room and how many of the folks are using Open -- or using CoPilot?

Show of hands, who's using CoPilot today?

Not many.

Okay. So a question that we've been getting from a number of clients relates to then taking the opportunity set and then just trying to think about the pace of annual leasing for Digital Realty and it's specifically around the hyperscale, we'll talk about the under 1 megawatt opportunity in a few moments. But just given the wide range of outcomes over the last few quarters in terms of the bookings numbers. And so if you look at this, whether you do it on a megawatt basis or a dollars basis, is there a right level that investors should expect on an annual basis for your model and with the opportunity that's in front of you?

When you dissect it to our most strategic priority, which we've put in data in the 0-1 megawatt and interconnection category, we grew that last year at 25% year-over-year. We set quotas for another double-digit growth in that category. And we see a long runway of that type of growth or CAGR compounding in that segment, given a massive addressable market, a category where we can attract tremendous value to our customers, as I mentioned, we're adding new customers all the time in addition to many existing customers that want to grow with us at Digital. Outside of that segment, as you get to the larger and just much larger capacity blocks, it becomes much more difficult to handicap. At the same time, I, by no means, view that the $1 billion we did in a string of four quarters is the top taking it for our productivity as an organization. Quite frankly, if you would just take the back half of the year, that would have been a record by itself, materially above the prior record. And I say that with a view that this infrastructure that we're providing is -- has a long tail of demand for our customers. We are positioned in the best markets that have the most diverse demand and value-add to those customers in terms of the supply constraints. And every segment keeps getting larger. The smalls are becoming mediums, the mediums are becoming largest, the largest becoming extra largest and some of the size of the capacity blocks are off the [indiscernible] charts. And we have a 3.6-gigawatt owned or under control runway of shell and land bank. So I can't promise you it's going to be the next set of 4 quarters. But I think the overall productivity, we have the organizational bandwidth, the key ingredients in the right markets in a market that needs this infrastructure to put up that productivity over time.

Just to dig into your comments just for a few more moments. So just in terms of timing, you mentioned you have 20% of the rental. You're highly pre-leased on the $7.5 billion that you're building. So does that inherently mean that anything you're selling today is really for late '26 and '27. And is that part of what might be complicating the sales process over the next couple of quarters?

The enterprise colo piece is much shorter signed commencements. So that's filling as we -- and that's a moving machine. We're delivering new capacity. We're filling the capacity. So that's happening live. That's the 250-plus bookings we did in last year. But you're correct, anything that's large or really large. It's just building our backlog that is going to go in norther and it's building a further runway of growth for the company.

And then you mentioned about the hyperscale market expanding. Can you give us a sense of how many customers you kind of consider in that core opportunity set that would fit into large capacity blocks of demand and where you might see that in terms of opportunity over the next few years?

I mean we're still in a universe that's probably 10 or less realistically. But you don't need 10 types of those customers to create competitive tension to -- especially in the backdrop of supply constraints to have a healthy equilibrium in terms of rates and returns. And quite frankly, all you need is 2 at the same time. So having more of these customers seeking similar needs and especially at the same or near times, creates competitive tension for our product that we're able to go on a great rates and returns.

And you mentioned the 3.6 gigawatts of runway room, Shell and land. How do you think about the energy availability and any permitting constraints that certain markets may create and the pace at which that could become available as sellable capacity to your customers?

In today's framework, all of that land has a path to power in the reasonable future. Now I'd say, call it, a giga change of it is the most nearest horizon, the 26th, the 27th, the 28th starting to dip into the 29s. But we've been adding to that, and we've been adding -- it's not been a FIFO version. We have -- the most recent land we bought is not the last land we're going to develop. So we've been filling in gaps where we bring the assemblages together in places where we're already operating. And we did that in [indiscernible] a quarter or 2 ago and added in other markets as well. So we're not too idly here in making sure that our customers have this runway for growth, which is us having the key ingredients in terms of land, access to power, ability to build as well as the funding model to execute.

And then maybe just switching gears to the under 1 megawatt plus interconnection. You talked about the acceleration. What specifically is driving that success for Digital Realty? And how much do you look at it as an expanding market opportunity versus the share that you're taking?

I look at it as a combination of both but probably more share taking at the end of the day. If you look at our history as a company, I touched on where we started to where we are. We had to make some tremendous foundational moves of putting the puzzle pieces together to be stretching across 6 continents, 50 metropolitan areas with the capabilities to deliver enterprise colo. We had to unite our brands, we had to evolve our go-to-market. We had to bring in new sales leadership. We've done many things in terms of reorienting our approach to partners. And I think success in that category happened slowly and then more recently, all at once. And at the same time, what's giving me such optimism is the record quarter in the third quarter followed by a record in the fourth, I honestly believe that was a fork and knife way to success where we just rolled up our sleeves. We drove the culture of the company to focus on. This is the priority, and we delivered results. And the things we're doing now in the midst of 2 record quarters are really what's going to pave a brighter future, giving that conviction for 2025 and 2026. In a quarter, where we had good results in our partners' alliances, we brought in a new Head of Partners alliances to take us to the next level. We are evolving our sales segmentation to push more account coverage in a flow business to commercial insight sales, so sellers can spend more time on higher propensity, higher-value, multimarket, multiproduct type accounts. We've reoriented our marketing dollars to not just be focused on new logo hunting, but to mine that installed base. We have 5,000 existing customers. We have over 1,000 key cohort customers that we know can grow with us to the second -- third site or incremental product. Our systems road map, we're investing -- have been investing a lot in our systems. The benefits of a unified 1 customer portal, sales force inventory tracking system, all those things are not even going to deliver to our internal or external customers until 2026. So I look at that as we've got a tremendous opportunity. It's a fantastic category, and we have a major right to win.

A handful of questions come in through a lot of QA, so we'll try to get to some of those.

Balance sheet questions for Matt and Jordan?

No balance sheet ones yet, but if you want to enter one in, I'll ask it for you.

So [ 2 years ago ].

To what extent do you feel comfortable holding back new build inventory to potentially capture better pricing down the road versus pre-leasing more aggressively.

I think this is remain commentary from the call. We're not holding any inventory hostage and we're not treating all inventory the same. We have some markets we're in 50 markets, right? So we have some markets where we do not have the strongest hand in all 50. And we have some markets with truly precious to our customers' capacity. And all we're doing is setting up and being judicious when you have a record year that you've leased so much and you've built so much our development pipeline that's already spoken for and your next delivery doesn't come on until 2026. Let's let the customers all take a fair swing at that opportunity. We're not here to pick our favorite customer of one over another. We're here to support all the growth, and we want to get them to the table and that competition just, we just -- we don't -- we have patients on that precious inventories is what we're doing essentially.

Maybe that's a segue just to broader pricing trends. What are you seeing in terms of pricing? I realize it's market by market, but if you can give us some general sense of how that's flowing?

Do you want to talk cash mark-to-markets? Or are you thinking just market pricing?

Well, let's take both.

I mean you never give them options. So Matt will touch on mark-to-market. Maybe I'll do spot rate trends. Go ahead.

So I mean, mark-to-market, so we guided towards 4% to 6% this year. I think a couple of things that are important to keep in mind are -- and that's coming off of what we -- basically at the same point where we started in 2024 was 4 to 6. Now we outperformed that in 2024, but that was largely due to 2, what we call, packaged deals that we were able to execute on within the year that pushed us to roughly 9% all-in on our mark-to-markets. If you excluded those, we've been down to 5%, which is right where we started. And coming in again to '25, we've got a similar place, 4 to 6. The reason for that is largely you've actually seen our expiring rents are pretty similar, actually slightly higher when you look forward, and I'm talking mainly in our greater than 1 megawatt, slightly the higher in '25 than the same place we were in '24. But I think what's probably more important is, as you start to look out in the outer years, you see those expiring rates start to step down, which puts us in a position where even if market rates remain flat, which we don't assume that they're going to do, but puts us in a position where we see an improving outlook for mark-to-market over the next 4-plus years that should help recruit to our overall, call it, stabilized growth. On the 0-1, which generally tends to be the majority of our role within a given calendar year, we're still seeing good mark-to-market came off of a quarter where we were almost close to 5%. And we see that in, call it, generally in that 3% to 5% areas where we're mark-to-market on the 0-1 category and we see that continuing over the next several years as well.

And on the larger stuff, the larger hyperscale stuff. I think the context is very important here. We're still in an inflationary environment, right? We're still -- so costs are going up. We're still in an environment where cost of capital is staying steady, not eroding, right? All these headlines you see if people want to invest in data centers. They're not running for low returns in data centers. I have no one come banging on my door saying, can you please invest in the low-return data centers, right? So that's a broader backdrop. Two, we deliberately focus on markets where we saw a robust and diverse demand and supply constraints of some shape or form, right? These markets versus markets we do not play in or did not choose to proceed are different and have different characteristics. And then most of these markets, we have established customers, they have locationally sensitive workloads, they have cloud agencies with radius restrictions of growth. They have needs to be there that it's not just about the price that is driving them to these locations. So yes, there may be a market like Northern Virginia, where we're out in the market at rates now in the $200 range that you'd say, hey, that didn't increase from last month. It's a healthy place to be in. Now there's a lot of other markets that never got at that high. They have been coalesce into the 150 area, including the international markets, which did not have a major benefit from AI to date but have the propensity to change as the nationalistic teams go of data sovereignty just like it did with cloud sovereignty. So I think market rents still have room to run broadly speaking in the core markets we are in 2025.

Very helpful. And just one more follow-up on the packaged deals. Is there some room for packaged deals in 2025 to give you a little help or upside on that renewal spread?

Those events are very much star on moon aligning events where you have the combination of an existing relationship with existing contracts that are not new contracts, not already at market. also intersecting with the customer having a need that really only we can provide. We caught lightning in the bottle 2 times last year. I'm not forecasting a third time in 2025. We always -- in good times and back in bad times, we always try to have holistic relationships and conversations with our customers, bringing everything to the table in places we can work together to partner. So that's always on the docket and maybe things won't happen to the brand size and scale. But I can tell you, as we're helping customers in smaller increments, we're talking out further in our exploration schedule about, hey, how we talk about those expirations in 18 months a little sooner. So we're actively on the balls on our feet on that topic, but I can't promise anything.

We've been getting some questions over the last couple of days just on the forward financial prospects and just how the model works for Digital Realty in terms of translating this demand into core FFO per share growth beyond the 2025 guidance that you provided. Can you give us an update on the multiyear opportunity to improve core FFO per share growth.

Sure. Yes. So let's start with kind of break down maybe 2025 is, call it, a baseline. So we start kind of the foundation. We talk generally about roughly 4% growth in our stabilized pool. So that's coming from the mark-to-market some level of occupancy improvement that we've also guided to that's around 100 basis points. So that gives us like a great solid foundation for growth in '25. Then sort of offsetting that slightly is we do have headwinds from refinancing roughly $1 billion of debt maturities that we have over the next several years. Given today's rates, those are generally speaking, about 1% of headwind that's going to offset that as we refinance that debt in a higher interest rate environment. We've talked about that as being a headwind, call it a year ago, that's going to continue for some period of time. And then on top of that, we've got what we talked about also, which is a healthy development pipeline starting to bring online the, call it, $350 million of backlog into revenue this year. That's giving us basically another, call it, roughly 3% to our growth profile that's getting to that call IT, 6% rounding up slightly at the midpoint to our bottom line core FFO growth. What we see going forward, and this goes back to kind of my mark-to-market questions is within that stabilized pool, we see an improving outlook on mark-to-market that should give us an opportunity to continue to push that forward up and to the right. And we've got still a robust demand environment that is going to allow us to continue to lease into it at healthy yields and deliver that 3% plus growth from our, call it, development pipeline, potentially pushing further, but also balancing that with capital needs that are going to need to go in to spend to bring that capacity online and some of the private capital or JV partners that are sharing in that. So all that being said is we're looking at 6% growth in 2025 and we see an opportunity to continue to improve on that as we mentioned in terms of improving, accelerating growth in '26 and beyond.

And then when you think about capital allocation, what role does -- now you have more financial capacity with the fund that you're creating? Where does capital recycling fit into the priorities for Digital Realty.

I think we're whittled down. Although we are consistent portfolio managers of what we -- where we own and operate a invest, I think we're whittling down non-core to a much smaller component. What's exciting to me is not just the big transactions that we've accomplished in bolstering our liquidity and be able to grow our investment opportunities with a stronger balance sheet, but evolving that into something that, yes, a $2.5 billion first fund is potentially modest in the scheme of our category, but it's a foundational building block of what could be a tremendous, call it, asset management business around our hyperscale. We have all the key ingredients to do that and the track record when it comes to our category. And that doesn't mean we're divorced from doing hyperscale on our own balance sheet whatsoever. But we just see this opportunity. It's all around us, and it's growing rapidly to harness the private capital with multiple different vehicles, both stabilized and development along the way that I think it's going to be scaled to many, many billions quite rapidly.

And maybe just to reconcile that to the guidance. So the stabilized hyperscale asset monetizations into the fund for this year are embedded in our $500 million to $1 billion full year guidance.

Matt, I don't want to leave you with that a balance sheet question. So on leverage, you're below -- I think previously, you've talked about a mid-5x leverage ratio. You're now below 5. What is the right target for Digital Realty? And is there -- as just debt products are evolving, do you see that influencing the level of leverage that you want to employ in your business?

I mean the simple answer is, I mean, we still view the 5.5x is a good target for Digital Realty. We are below that today, as you mentioned 4.8x, but that's with the view that as we look forward, we see -- again, a lot of what we're bringing together a lot of what we talked about today, we see a strong development pipeline that's growing. We see good demand backdrop from our customers, and we want to be ahead in terms of our ability to fund and be ahead of the capital needs that we have today. So we're sitting here today with $6 billion of liquidity, below leverage target and an ability to kind of see through and fund what is our '25 capital needs and likely into '26, especially considering the -- that doesn't even include the -- in terms of liquidity, the $500 million to $1 billion that we talked about in our guide that we expect to come in from some of the non-core asset sales and the private capital initiatives that we've also discussed.

And as you look at the growth opportunities ahead as well as the investments that you want to make in development, what does this mean for the dividend policy going forward?

I mean, so it's a good question in terms of like how we're looking at funding and being able to fund the business going forward. I think our call it, key priorities are around maximizing our operating cash flow because we see that as an efficient use of funding cost of capital for us given the 12-plus percent development yields that we have overall today. Increasing our EBITDA, giving us leverage capacity. On a normalized basis, our EBITDA has grown plus 10%. So that's another avenue for a source of capital the executing on the, call it, private capital and non-core asset sales is another avenue and ultimately, bringing those all together so that we're reducing overall our need reliance on public capital for to fund our growth. So I think our dividend policy in general is to pay out to our -- at 100% of our taxable income, which we've been able to do and also use that as part of the non-core -- or non-core recycling as well as some of the joint ventures. So I think that policy will continue going forward.

All right. And now we have 45 seconds here, so we'll do rapid fire quickly. What will same-store NOI growth be for data center sector overall next year in 2026?

We're guiding to roughly, call it, midpoint 4%, so I think 4-plus percent.

Next year. Great. And then will there be more fewer the same number of public data center companies 1 year from now.

One more.

Great. Thank you very much.

Thank you very much.