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

Our next session, I'd like to welcome back Digital Realty to our conference. And from the far right, we have Jordan Sadler, who is the Head of Investor Relations. And then we have Matt Mercier to my immediate left, the Chief Financial Officer. Welcome, both.

Thanks for having us.

Thank you.

I want to start with some general questions, and then we'll get into things like customer behavior, project economics, talk a little bit about the financial outlook and then balance sheet. But maybe just to kind of kick things off, how is Digital Realty adapting its strategy to navigate through macro uncertainties such as interest rates?

Yes. So we've -- so I think a couple of things. First off, I think as probably most of you have seen or heard, I mean, we have the benefit of being, I think, in a sector that's driven by secular demand drivers. And right now, I think we're at one of those points where we have the benefit of seeing a tremendous amount of demand at a time where supply has been constrained and limited. And I'm sure we'll talk a little bit about some of those demand drivers. So I think at a foundational level, we're benefiting from being in that sector and being able to benefit from that positive demand-supply environment. Second, in terms of the business, we've been working over many years to really evolve the company and the business on a few fronts. So first off, we've been very specific and intentional about being a global business. We think that, that's been hugely important. And you've seen us expand globally over the last many years to the point now where we have 300 data centers. We're in over 50 markets, 20-plus countries, which really gives us an opportunity to really serve our clients' need across that large, embedded portfolio. And second, we've expanded in terms of our product set. So at our core, we've always been a business that's been able to satisfy the hyperscale segment, what we call kind of the greater than a megawatt business. We continue to focus on that business. continue to do well there, again, being able to land and expand across our global portfolio. But we've also been intentional about being and growing our 0 to 1 megawatt interconnection segment of our business. And we've done that over the last decade through -- starting with the acquisition of Telx through to Interxion. So it really gives us an ability to not only satisfy our customers' need across the global portfolio, but across the -- a stack in terms of being able to provide anywhere from a cage or cabinet all the way to multi-megawatt. And last, somewhat, I'll say, more recently, although joint ventures haven't necessarily -- aren't necessarily new to Digital Realty, but we've also been evolving our funding model as part of being able to adapt and really go after what has been a tremendous amount of demand in the market today. So you've seen that recently with deals we've done around our stabilized assets with GI and TPG, more recently with development JV, we've done with Blackstone, which we closed the first phase on earlier this year, expect to close the second phase by the end of this year. So -- and that really is giving us an ability to, again, be able to satisfy what are these larger capital requirements as the demands have gotten bigger in terms of what customers are looking for and being able to leverage outside capital to be able to fund the business and really improve returns for Digital Realty shareholders as well.

I think we hit on a lot of the topics that we were going to get on later. We can double-click on some of those. But maybe just in terms of balance sheet management, interest rates, give us kind of a snapshot and where you're heading.

Yes. I mean, so in terms of the balance sheet, again, a lot of work has been done over the last 18 months. So at the start of '23, one of the objectives we had set out with is to bring our leverage back down to our target leverage area of 5.5x. So at this point, we're ahead of schedule on that. We reported last quarter where leverage is at 5.3x. And I think that, one, that's going to just give us an ability to be back what I'd call a more regular cadence in terms of a capital funding model. Part of that being what you might have saw last week, we did a recent euro issuance that was north -- sorry, south of 4%. So it really gives us that ability to be able to be back in debt capital markets at a time when interest rates are, call it, stable to slightly declining. And overall gives us an ability again to be able to fund the business at an attractive cost and really be more strategic in terms of our outlook and how we attack the market going forward now that we're in a more leverage favorable position.

Development time lines, how are these trending? A lot of complexities around supply chain, energy transmission, other challenges.

Yes. I mean -- so maybe to kind of set the stage on that a little bit. I mean, overall, our development time lines on data center that's under active construction has not really changed. I mean, so if we're under active construction, we're going to be -- we're able to bring to market data center capacity within, call it, 6 to 12 months. Today, just to give some data points on that, we're underway on 430 megawatts of data center capacity currently. Right behind that, we've got over 700 megawatts of shell that we've got started. And then sort of last is we've got probably 2.4 gigawatts of land capacity available. So depending on where you're leasing within kind of those 3 major buckets, it gets longer in terms of when you sign to commence. And so you'll see that elongation in terms of development time lines. But I would say we've been in this environment where our power capacity has become tougher to manage, to source. You've got to be much more in front of that procurement process. But that's something that we, again, have been in that business for a long time. We've been very intentional about planning ahead of our capacity needs. So there is -- in some cases, there is an elongation in certain markets around development just because you can't get access to power today. But we've got a number of markets where we've got available capacity that can satisfy current demand quite easily.

So development yields is a metric that a lot of folks have talked about over the years. I think in general, the trend has been upward. People may have different specific answers around unlevered yields on cost. Curious given the demand pipeline that you're seeing and the projections that one hears about continued double-digit growth in rent in Europe and in the U.S., how do you see development yields trending going forward?

Yes. So I mean, we -- if you look at our development pipeline today, which we give a decent amount of disclosure around. So today, we're a little -- overall, globally, we're a little north of 10% on our overall development yields. And I would say, generally speaking, that's what we're targeting at a broad level. We're looking at 10-plus percent development yields is where we target. Now you look and you kind of break down at least where we are today within the different regions, you're seeing slightly north of that in APAC. In EMEA, you're seeing slightly under 10%, but it's pretty close to that. I think it's around 9.7% currently. And North America, where we've probably seen the most movement, if you look back, call it, 12 to 24 months, that's now approaching 12% in terms of our returns within the North America region. That's somewhat a function of -- that's where we're seeing the majority of AI-related deployments in use cases today. But broadly speaking, we're looking at 10-plus percent on development yields is where we target across our global portfolio.

And is that the same for your on-balance sheet investments compared to some of the joint ventures that you would have in Brookfield, which spans a number of different continents as well as Blackstone as well as others that you've entered into?

Yes. I would say, yes, in general, that 10-plus percent kind of fits pretty nicely across the global portfolio. I mean there's going to be certain more emerging markets where we'll target north of that. And then I would say part of it, again, back to how we're allocating capital and our returns. Part of the benefit of bringing in outside capital within the business is that, that provides a boost in terms of the return to Digital Realty and the capital that we're putting to work as we're able to earn fees not only from -- generally speaking, from the development side, but also as those assets come into operation, we're earning property management and asset management fees in most cases, which adds incrementally to our overall return profile.

So you're not new to hyperscale leasing. The market has gone through what I think is sort of a V-shaped recovery now, but we're on the other side of the V prior to the pandemic largely speaking. And as we think about re-leasing spreads over the coming years, given that dynamic, should we expect to see kind of a similar trend as we've seen this year compared to the prior year in terms of upward movement in re-leasing spreads for your greater than 1 megawatt category?

You're talking just the greater than 1 meg?

Yes.

Well, I'll talk a little bit about both, but I'll start with the greater than a 1 megawatt. Yes. So I mean, again, as a result of what has been a very positive sort of demand-supply equation across most of our global markets, we've seen pricing pick up. We've seen that benefit us within our renewal spreads. So we recently guided towards 5% to 7%, and we ticked that up from where we started the year. I would say when you look at 2025, you are seeing in terms of our expiring rates, those are probably still -- those -- the rates of where our leases are expiring still relatively elevated. So the expectation is you're not going to see a huge pop in terms of where '25 is. But if you look out to '26, '27, '28, you start to see expiring rates continually step down where I think you'll see a more pronounced opportunity for better mark-to-market as you get into those outer years as a result of when those leases were originally signed at an interest rate in an overall environment that was lower. So we see better opportunity for improved mark-to-market as you get to outer years, especially within that greater than a 1 megawatt segment. And you've seen -- we've actually seen some of that in terms of the ability to pull some of that forward. We did that in the first quarter of this year, where our mark-to-market was closer to 12%. Again, that's not something that we can do every quarter. But I think that shows what the potential is as you get to these outer years on what the mark-to-market opportunity is, especially within that greater than a 1 megawatt segment. And then I know there's a lot of discussion about the hyperscale segment. And rightly said, I think that's where -- what everybody sees and reads about. I mean we're also -- we have, call it, 40-plus percent of our revenue coming from the 0 to megawatt enterprise, more retail colocation. And we continue to see that business do very well. In terms of mark-to-markets, those contracts because they're generally in the more 2- to 3-year time horizon, they're closer to market, but we're continuing to see those every year step up in the 2%, 3%, sometimes 4% area. So it really provides a nice solid, stable foundation for growth within that segment of our business.

And maybe in that sub-1 megawatt category, which doesn't get as much attention, what are you seeing around demand signals, competitive behavior as a lot of your formerly listed companies have been taken private? I think it's fair to say the CapEx budgets have gone higher and to move that capital, they're focusing on bigger deals. Does that have any ramifications in what you're seeing in kind of the retail enterprise cross-connect part of your business around competition and then share basically?

Yes. I mean we're -- I think you're seeing that there's -- globally, there's really 2 players within that segment of the market ultimately that can satisfy again across from a global perspective, from the product perspective. And we've had an intentional focus on continuing to grow that 0 to 1 megawatt segment, and we've executed well on that. You've seen over the last I think, 4-plus quarters, we've had over $50 million in signings across that segment, and that's continued to step up. So I think we've shown ability to execute within that part of the market. And I think that's one a function of just being focused on that segment of the business. But also maybe to your point, I think others have started to move away from that and focus solely on the hyperscale business. And I think we've been able to take advantage of that, take share in a number of markets and as well as part of us just continuing to add to our product capabilities and being able to provide a value proposition to those enterprise customers. You started to see us, I think, win and grow within that segment.

Some of your peers shifting back to kind of wholesale hyperscale have shifted a bit towards powered shells and you offer that product as well, hasn't really been the engine of your growth. But any observations about the mix between turnkey versus powered shell appetite for yourself as well as by customers?

I mean, in our -- I mean, to your point, we've had -- that's been -- I mean, we've offered powered shells for a long time. It's been a part of our business. What I would say in most cases today, our customers are looking for that fully fit-out product. I think, they're looking for that turnkey solution that we've been able to provide across, again, that global portfolio. And I think they're looking for sort of the expertise and being able to operate and manage within that broader environment. So the majority of the discussions that we've had have been on our customers looking for the more fit-out turnkey solution. But we have more than the ability to offer powered shell within our portfolio. And we've done that on in certain limited cases where customers have asked for it. But for the most part, our experience and our discussions have been customers seeking that fully fit-out operationally ready solution.

So a lot of your KPIs have shown a lot of improvement. We talked about balance sheet and the delevering, the pricing dynamic, renewal spreads, same-store NOI growth. Given the investor focus on FFO per share, what are the outlook and growth targets for Digital Realty? And how do you navigate from what's currently fairly modest growth to what looks to be much more significant next year and beyond?

Yes. So we've done a lot of work, as we've talked about through some of these earlier questions around balance sheet through capital recycling, stabilized assets and really putting ourselves in a position now at the start of 2024 to really be in a position where we're growing bottom line FFO per share growth. I've already -- at the start of '24, almost -- I had to give 2 years' worth of guidance, which is always fun when you're the CFO. So I've been pretty consistent almost through the entire year that '25, we've been talking about a baseline growth of 5%, and we see opportunity for that to continue to accelerate as you move into the outer years. And so that -- the reason we're -- we feel comfortable about that is, again, back to a number that the dynamics that we've talked about. So we're seeing our same-store growth be positive and improving. We're in -- we talked about being able to achieve 3% to 4% within that area, and that's a result of a positive pricing environment where we're seeing our mark-to-markets continue to improve. In addition, you've got the development yields that have been also improving. You'll see more of that development that we've done over the last year start to come online in '25, more in the back half of '25, then setting ourselves up for an acceleration as you get to '26, '27, where the mark-to-market opportunity continues to improve each year, the development yields and the sites coming online give us that incremental growth to be able to start at that 5% in '25 and continue to grow it from that point forward.

So putting that all together, it sounds like a pretty promising trajectory. What are the biggest operational or macro risks that would stand in the way of seeing that kind of growth? Is it basically just delivery and execution? What should we be mindful of in terms of risk factors?

Yes. I mean it's -- so I think -- ultimately, I think there's -- we've done a lot of the work through -- again, if you look at our -- we've got close to a record backlog. I think we had a record backlog last quarter, and it came down slightly because we had a high commencement quarter. But we've got a record backlog. So that -- again, that really just sets us up for '25 and really starting that foundation of solid growth. I think really from what we're seeing, the main toggle that we're looking at in terms of future growth is really back -- largely back to how much development do we want to start because in today's environment, development is generally a dilutive proposition initially. And this all circles back to kind of how we've been intentional about our funding of the business, bringing in outside capital partners so that we're able to not only attack and be part of the hyperscale segment, but not have that continue to be an outsized portion of our overall business and make sure that we're focused on delivering that per share growth next year and growing that every year beyond. So I think it really comes down to, I think, one of the factors is our ability to manage the development starts and any associated dilution in bringing in capital to support continued growth.

So a growing portion of your commencements will be significantly AI weighted. With that comes some tweaks to your design. And HVAC infrastructure, liquid to the chip and so direct-to-chip liquid cooling and so forth. Any ramifications to think about in terms of customer SLAs, operations and maintenance practices as you move further into that arena?

Yes. I mean we're...

When compared to, say, cloud data centers...

Yes. We're -- I mean, we're still -- I would say, generally speaking, we're still in the early innings of deploying liquid cooling technology today. We've been able to satisfy the majority of our -- of the AI workload needs through our, call it, more traditional cooling methods, largely air oriented. Now we've supplemented that with things like rear or heat exchangers to help improve the efficiency. But to date, that's the way we've been able to satisfy just given customers' urgent need, time to market, wanting it now. So that's how we've solved it. Now as we look towards designing new facility and deploying new facilities where most of that liquid cooling technology is going to take place and where we expect that next wave to be, we're not seeing a significant shift or increase or decrease. It's been pretty consistent in terms of how we're going to operate it, the cost to maintain because you're really swapping out different forms of equipment. Same type of people. There's maybe -- we'll have to train some on the operational side, just given the liquid nature of it. But overall, based on what we're seeing in our initial designs, we're not expecting any incremental cost or complexity really in order to satisfy that demand. And on the return side, I mean, again, given the macro supply-demand dynamics that we're seeing, the improved pricing, the returns on AI versus cloud are also similar just given the demand backdrop that we've got in front of us.

So big audience in the room. If there's any questions, feel free to raise your hand. I can call on you, a number of alumni at the company here as well. I noticed quite a few.

I'm interested to hear, so you're forecasting no additional capital expense to support the liquid cooling technology. I agree with you in representing tenants that I haven't seen it be a deal breaker if the data center isn't able to support liquid cooling today, obviously. But I would have assumed that there would be additional costs because some preliminary feedback from data center providers as we look to 2025 through 2028 is that those liquid cool deployments would come at a premium. So therefore, I would assume it's because there's a higher build cost.

There's -- I mean, we're -- again, it kind of goes back to and maybe not totally clear enough. I think we're seeing -- it's a marginal incremental cost in order to deploy liquid versus air, right? Because you're swapping out different components. So ultimately, there's a slight marginal increase in the total overall build cost, but we're seeing that more than offset in terms of rate and price that we're able to achieve in today's market so that our overall return profile and margin, we don't see a huge difference between what we're able to achieve from a liquid cooled -- fully liquid cooled design versus a more traditional air handler design.

And I think Matt's referencing de novo builds, right, as opposed to retrofit, right?

Right?

Yes.

Sure.

Go ahead.

So one of the interesting phenomenon we are seeing is SLA fines. And a lot of the times delivering a project late often then just negotiating that way with the hyperscaler and that's not changing, demanding those penalties. How -- it's very hard to collect data on this. What are you seeing in the market on SLA fines? How vicious is this?

So the question is SLA fines and what is Matt seeing in the market and how are they navigating that with customers?

I mean the way we navigate is to make sure that we don't trigger those. So I mean, we've got -- that's where we've got a history of -- we've been building data centers. We've been spending $2 billion to $3 billion a year for the last 4 or 5 years. Before that, we were -- we've been one of the biggest developers, operators of data centers for a long time. So we're -- we have those in our contracts in terms of making sure that we're hitting our RFS dates. But -- and ultimately, it comes down to conversation with the customer. But we're aware of those. We know those are out there, but I think we've been able to manage through those by the fact that we've got a great team involved in bringing data centers online. We've got deep development expertise, and it's ultimately down to a conversation with the customer if and when that ever happens.

I want to ask real quick and then, Dave. So just capital strategy going forward, given target leverage ratios, equity issuance, capital recycling, what are the sorts of things that you're kind of thinking about? And how do you balance those different tools?

Yes. So I mean, you've -- again, we're -- now that we're in a position where our leverage is near our target leverage area. I think going forward, what you'll see is more of a mix of the way that we want to optimize funding the business is, first off, use cash flow from operations. That's the cheapest source of, call it, equity for us. Two, leverage what we expect to be growing EBITDA to be able to put incremental debt funding on the company at leverage neutral. Three, to the extent demand driven, utilize public equity if and when needed, but ultimately, it will be to be funding developments that are in that 10-plus percent area. And then fourth, we're continuing to evolve, as we've talked about, our private capital sources. And those would be similar to the mix that we've seen over the last, call it, 12 to 18 months in terms of a portion of recycling capital from stabilized joint ventures continuing to work with Blackstone as that development JV comes online, where we've raised close to $7 billion of potential capital, which will be spent over the next several years. So putting all that together and again, making sure we do that in a way that ultimately, at the end of the day, results in us growing bottom line core FFO per share at that baseline level for '25 and accelerating going forward.

Dave?

[indiscernible]

Changed your mind.

Any other questions? Go ahead.

[indiscernible]

Yes, you're talking -- GPU as a Service...

So GPU as a Service start-ups and how are you dealing with those customers and kind of behavior you're seeing?

So I mean, we see those in the market. I mean, right now, we're -- I think we have -- we're in an enviable position where we're able to kind of focus on, I would say, the larger customers within -- that we have existing relationships with, better credit quality. Those are the customers that we're serving, primarily within our core markets. And if and when we look at sort of the more GPU as a Service, we'd look to put them in markets that have available capacity that's been there for some period of time. And two, we'd structure in such a way that we were protected either from significant deposits and/or structurally from a legal perspective, able to make sure that we can get back that space efficiently. But as of today, in today's current sort of demand environment and supply constraints, we've been able to focus more on sort of the larger technology cloud companies in terms of satisfying that demand.

Very good. We are out of time. I appreciate both of you being on stage with me.

Yes. Thank you.