Pleased to have Digital Realty Trust here with us again this year. We have Jim Huseby from Investor Relations here. Jim's a longtime participant here at the conference from several different companies, very pleased to have him here. Jim, why don't you talk to us. Everyone hears about data centers. Why don't you talk to us about where does Digital Realty Trust fit in to the data center landscape and everything used to be this sleepy industry nobody knew anything about and now is on headlines all the time.
Yeah.
Talk to us about kinda where you fit into the space, and then we'll get into some questions.
Yeah. Thanks. Thanks, Frank, and thank you, everybody, for joining us today. I think I may be one of the few people in this conference room that attended Raymond James conferences at the Hyatt 20 years ago, so it's been some time. Anyway, Digital Realty is the world's largest data center company. We've got over 300 data centers in 55+ markets around the world on six continents. To Frank's point, when I was starting in this industry 15 years ago, nobody knew what this was. I had to explain it to my wife and kids as to what is a data center? I had to explain it to investors even.
Today, it seems like we're in the headlines, sometimes for good reasons, sometimes for more controversial reasons, almost seemingly every day. Digital Realty plays. We call ourselves a full spectrum provider. We've been around for 20+ years. I think the idea behind full spectrum has actually expanded over the last year or two on the largest end to include data centers that are bigger than anything else any of us had imagined. We have two core businesses, two core products, I guess is the better way of putting it. We have a colocation product that is connectivity-centric. We call this our zero to one plus interconnection business. This tends to attract highly latency sensitive applications and workloads.
They are placed in connectivity hubs, otherwise known as carrier hotels, formerly known as carrier hotels back in the day. And these are mission-critical applications that typically enterprises are putting in here to run and evolve their digital workloads. The second side of the business is what we call our greater than a megawatt business. To be fair, that 1 MW .
It's a lot greater than 1 MW.
Is somewhat arbitrary. Historically, it's represented this dividing line between these connectivity-driven applications and larger applications, probably a little bit more akin to what most of you think about when we think about data centers. These are our larger deployments, oftentimes a little less reliant on connectivity, although there's some, and more reliant on just brute compute force. We're not totally into AI training yet. You know, that would be even above that. It's directionally in that direction. We, you know, Digital Realty has locations around the world in the major metro markets around the world. Think New York, L.A., London, Paris, Frankfurt, Tokyo.
We tend to build in and around a connectivity hub, and then we put, more suburban scale locations for the more, compute-heavy applications out and around it.
All right, great. Kinda with that, maybe walk us through, how are you focused on various types of demand for AI and how do you play in that space?
Sure. So, we always think about it as having three big drivers in our industry. AI is the latest. You know, there's digital transformation, there's cloud, and there's AI. And we support all three, and all three are continued ongoing demand drivers for our business. Specific to AI, in our greater than a megawatt business, we provide some of these large compute-heavy AI training deployments. More recently, let's call it in the last six months or so, we've begun to see some early developments on the AI inference side. These tend to be, again, smaller deployments in more connectivity-driven data centers. We're seeing enterprises beginning to experiment with things like Private AI and a few other things. I think we're still really early in this area, Frank.
I think there's, if I were to use the baseball analogy, I think we're in the top of the first. Nevertheless, I think there's a lot of opportunity there to do that. Our platform is particularly well-suited for, you know, to manage both of these needs, both for AI training and for inference. We're not as yet the size of some of these gigantic, what I'll call giga-scale applications that you may have heard about in places like West Texas or Louisiana and a few other locations around the world where the OpenAI and other of the big AI training models are putting in.
We're not yet at that size, but we manage 100+ or multi-hundred plus megawatt deployments, elsewhere in and around population zones rather than in these remote locations where latency could be an issue.
All right, great. On the other side of that, you made a concerted effort in the zero to 1 MW space to really refocus the company on that, on the last couple of years. You know, what's changed to kind of make that shift, and where does that put you strategically versus, the competition?
Yeah. So, historically, Digital Realty was more of a wholesale player, meaning primarily in that greater than a megawatt portion of our business. About 10 years ago, the company made a strategic shift to try to move into some of the connectivity side. It owned certain buildings, but it acquired a company called Telx that ran a lot of the connectivity within a lot of those, within a lot of those data centers. You know, we've built that business because there are some very, very high barriers to entry in that business. You know, telcos don't love just building huge nexuses of connectivity to random data centers. You've got to go where they are now. We've acquired several of those over time.
We did a few more a couple of years later when we bought some assets from one of our competitors, Equinix. We also acquired a company called Interxion, a Pan-European provider of retail data centers. We've done some other tuck-in data centers in around the world over the last couple of years. You've seen us move into Greece. We've last year we went into Indonesia. Just on Monday, we went into Bulgaria, for example. You see us building out this worldwide connectivity hub.
Now we're at this place where we've got about 35% of our lease roll is in this zero to one, this connectivity-driven business, about 60% of our lease roll is still in this greater than a megawatt or wholesale business with the remaining 5% in a legacy triple net piece.
All right, great. Talk to us about what some of your advantages are versus some of the larger private players that, again, are building a lot of the data centers, getting the attention these days. You know, what are your advantage there? How are you differentiated, and how is that showing up in your results?
Yeah. A lot of the private guys are building the, you know. They are focused very much on large scale capacity, focused on serving the hyperscalers. Digital Realty has a different approach. As I've mentioned, we've offered both connectivity-driven solutions and scale. We create what we call a Connected Campus in each of our metros. A connectivity hub, typically in the city center, suburban wholesale locations. That gives us an opportunity to address big GDP and these workloads that come on and are required by enterprises to work and interact with others.
While a lot of the private guys are focused on simply building large one-off data centers or a series of those for the hyperscalers, we've got a more, what I think of as a more inclusive and broad product set that targets a lot of enterprises, a lot of cloud and AI players. To be fair, there's a, there's some overlap, particularly in the last two.
Okay, great. All right, switching gears a little bit, let's talk about power. It's always been a topic of conversation. You know, how are you thinking about power availability, and can you comment on the situation in Loudoun County with the transmission lines that are hopefully coming up someday, and to your campus? T alk to us about the whole power situation.
Power has become the limiting factor in this industry for future expansion. You know, there's tremendous demand for data center capacity right now, but the supply is being limited by the amount of power that people are able to get or players are able to get. We've been able to negotiating and get locations around, but we always prefer to rely on the grid. We think that's the most reliable. We've been exploring various bridge solutions when situations are very tight. I think microgrids run by natural gas or fuel cells or things like that. But the preference is certainly to go into the grid.
Is that a preference just from a reliability standpoint or a cost standpoint? What, why? 'Cause we hear a lot about that, and a lot of investors ask us about self-power generation and so forth. Where is it that kinda that kinda leads you to the conclusion that that's not a primary or is maybe less of an option than people might wanna think?
Yeah. Again, if you think about where our campuses are located, we tend to be in densely populated locations. Building sort of building grids, to serve your own needs, put in a natural gas generator or even small nuclear down the road or a fuel cell.
That's easy in downtown Chicago, no problem.
It can be challenging. Grid reliability becomes a real asset on that. We've got a lot of experience in doing that. Yeah, as I mentioned, we are exploring these other opportunities as a bridge. You know, power around the world is more scarce today than it has been in my memory, and particularly true in Northern Virginia, where the issue is less around power generation and more about transmission of that power, and specifically it's transmission of that power into a very specific location that happens to have a huge density of data centers. We've got locations both within that constrained area and outside that constrained area, both in Northern Virginia, that we've been able to work on.
We've worked with Dominion, the power provider there. We've got, in addition to the, call it 350 MW of capacity we've got currently under construction, we've got an additional 300 MW roughly to continue to build out into that. That should take us for the next couple years. Beyond that, it gets a little more murky, just given that, given our preference for grid providers, and that we don't have any confirmation yet from those providers about when all of that capacity is gonna become available, which is exactly why we're looking into these sort of bridging solutions.
Okay, great. All right, next let's get down to demand and how that impacts the business. You've had a strong couple years of bookings. You know, talk to us about that and your confidence in the outlook for continuing to be able to see booking similar levels of demand going forward.
We've had two years of bookings in excess of $1 billion annually. We've got a backlog that's north of $800 million. $600+ million of that will commence this year in 2026, the remainder in 2027 and a little bit in 2028. We've had a tremendous amount of success. Some of that, those bookings are these large single deals, call it 100 MW at a clip. We've had a couple of those deals and a bunch of other larger hyperscale deals. The real interesting part about what's happened over the last two years from a bookings point of view is the acceleration in our zero to one business.
About two years or so ago, we were sitting at about $50 million a quarter in bookings here. We were at about 100 new logos a year, a quarter. We had put out an effort to try to do what we call double-digit, which is basically we wanted to double those bookings. In fact, within two years we're almost there. We're at about in the mid-90s last quarter. You know, we're moving strong and quickly. Had 155 new logos last quarter, so it's showing the value proposition is really resonating with a broad selection of new customers there.
You know, all of our connectivity plays have enabled us to say, "Hey, now that we're on the cusp of achieving this goal, maybe it's time to even extend it for maybe move the goalposts a little bit." We'll have to see if that happens.
Okay, great. What does it take to kinda hit the high end of the range of the guide for this year what might cause you to be at the lower end? How should we think about that?
Yeah. We exceeded our guidance last year, in 2025 pretty meaningfully. And we've put out a, I think a pretty well-received guide this year. Our visibility to 2026 results is actually very, very high. I mentioned a few minutes ago, the very large backlog that we have, most of which, will be commencing in 2026. That's really the heart and soul of our guide right there, is that, particularly on the growth side, is that we're expecting to meet all of those expectations, from a commencement point of view. To exceed our, you know, the guide that we've put out there's really two drivers on that. Number one, we could exceed in the zero to one category bookings.
They tend to have shorter commencement sign to commence, periods. We could actually have more of that commence within the year. That would help our later in the year results, which could obviously accrue to the full year. The second part of that is, we could overperform on renewals. We've been in the 6%-7% range recently. You know, should we end up outperforming our expectations there, that obviously has an immediate benefit within the year. On the downside, as I mentioned, the backlog, gives us tremendous visibility. There's always execution risk. We could, we could stumble and not complete some of these projects. We've got a great track record in meeting them, but there's always execution risk.
I would say that's the biggest issue that we've got.
Okay. With that, you've always had, part of your visibility on that is having the commitments and so forth. Talk to us about what you do with your suppliers for labor and materials to stay ahead of supply chain issues and get assurance that you've got the raw materials that you're gonna need to be able to complete all these growth targets that you have and meet these commitments.
Yeah. This has become an issue that has become more of an issue since the pandemic, and then with the rise of AI data centers, even more so, it hasn't really normalized. Digital Realty is the biggest player in the industry. Really has, you know, can benefit from that because we've got longstanding agreements with a lot of our supply chain partners. We've got VMI, vendor managed inventory programs with a lot of that. Given the scale of the company and the number of projects that we've got going on, we're able to move inventory around and order in advance with our investment-grade balance sheet.
That, despite all of the issues that we've had over the last five years as with supply chains, Digital Realty has been able to really manage around that very, very well. There are long lead times for certain data center parts, we're well aware of that. We've been moving earlier to procure that. We keep a few things in inventory as well. We've always got some, just because we've got so many projects, I think we're probably in, we have more flexibility than others with fewer projects. Labor has been another issue on this side. You know, we've heard of issues with regard to the lack of availability of skilled labor to build data centers.
Once again, I think our size and the scale of our development keeps us as a favorable contractor for a lot of our subcontractors. We've not really encountered any meaningful issues with regards to labor availability, but it's obviously something that we keep our eye on all the time.
Have you seen labor be an issue for any of your competitors or in any specific markets? I, that with anyone else?
Yeah. It is, it's a challenge. You know, we've seen not everybody is able to deliver their capacity on time. We're aware of some that have not. T o be honest with you, I'm less sure if that's a labor issue, a supply chain issue, or something else altogether.
Great. All right. Yeah, one of the things you mentioned earlier, looking at, re-leasing and so forth, you've had a very good track record for the last few years with re-leasing spreads. Some of that's benefited with escalators moving up last years and so forth. Talk to us about your visibility on the positive re-leasing spreads. How far out does that extend to be able to continue to see that momentum?
Yeah. You know, contracts come up for renewal, and our business is very sticky. Once you're in our data centers, you tend to renew at the end of your contract. One big difference between those two product sets that we've been talking about is that the zero to one-plus interconnection business tends to have very short-term contracts. One to three years, initial contracts, oftentimes they move into auto one-year renewals. You end up with you know, despite it being a smaller part of our overall business, those contracts, come up for renewal much more frequently. Really about two-thirds of all of our renewals last quarter were in the zero to one business, and that's not unusual.
This is a very steady eddy kind of business. We see typically, renewals in the 3%-5% per year range. There last quarter, I think we were a little north of 4%, 4.1% maybe, on that. That is really sticky. That, because it's 2/3, it's gonna drive the overall total. In the greater than a megawatt side, which is most of the rest of the 1/3, we exceeded our re-leasing spreads. We were north of 8% last quarter. As I look out two years, I see that our expiring rents decline over the next four or five years. Even if prices stay flat, we should get a, you know, a widening delta between the expiring rents and the current market rents.
You know, that presents an opportunity for these mark-to-market to price them up, even without any ongoing price increases, which we've seen a lot of over the last couple years. Nevertheless, we're pretty optimistic on that part of the business.
All right, great. I don't know if anybody has a question in the audience, glad to take those if anyone has something they'd like to ask. If not. Yeah, go ahead.
Yeah, what's your thoughts on hardware refresh? I seem to remember a few years ago there was a question about appreciation events versus the actual maintenance costs for your hardware. Do you have any thoughts on that?
Just refresh rates for hardware costs.
Yeah, the question about the cost of refreshing for hardware. There was an issue a few years ago about depreciation and so forth, but talk to us about how we should think about that.
From Digital Realty's perspective, the IT hardware that's within our data centers, is owned by our customers. Whether it's servers, GPUs, et cetera, et cetera, the urgency to replace those and refresh that hardware will lie with our customers. We provide the infrastructure around that. We maintain the infrastructure to run all of that hardware, in tip-top condition all the time, but the IT equipment is, generally speaking, not ours. You know, the laptop in my bag is some of the biggest IT equipment we've got.
All right, great. You had another question? Yeah.
In the new geographies that you're expanding to internationally, how are those being powered? What are the, what are you using as power sources? Could you comment on some of the newer ways of electric power generation that you're currently looking at? You hear a lot about small nuclear reactors being, you know.
Yeah
The next thing. You know, seems like that's a long ways away.
Yeah. A lot of what we're doing, enter into new markets is that we're acquiring existing assets and then building on. We're obviously leveraging the connections and the power that the target has had. Over time, obviously, we wanna build these out and add more capacity in a growing market. You know, we'll leverage the expertise that they have developed and the relationships that they worked with their local grid companies to go ahead and expand that. We've done some rather innovative things in certain markets where you've found challenging grids. Good example is South Africa. You may be aware of the grid issues within South Africa.
Our South African subsidiary, Teraco, has built a utility scale solar plant, next to them, feeding that into the grid and feeding both our data center, but also, helping to provide additional load in there. We're exploring other things. We don't wanna be a power company, that's not our business. We have to look at other options. You mentioned small nuclear reactors. We agree. That seems like a great solution, but it's probably just still over the horizon a few years down the road, and we've got more near-term issues. We're looking at the bridging solutions that I talked about earlier. Ultimately, they could become things like SMRs, but we're thinking more around natural gas, fuel cells, and the like right now.
All right, great. Rapid fire, I mean, very quickly, what's the one biggest misperception about the company that you'd like to address?
I think, Digital Realty, this is not. T o paraphrase an old car commercial, this is not your father's Digital Realty, right? We're not this. You know, 10 years ago, 12 years ago now, we were a wholesale provider in North America. Today, Digital Realty is a global player offering the full spectrum, albeit maybe not that super high, super ultra large capacity any, at least not yet. Not yet. We're able to provide solutions for cloud, AI, and enterprises around the world.
Great. Thanks, Jim. Really appreciate it. We got a breakout session after this if anyone wants to join.
Thank you everybody for joining.