Digital Realty Trust, Inc. (DLR) Nareit REIT Week: 2024 Investor Conference (Transcript)

Digital Realty Trust, Inc. (NYSE:DLR) Nareit REIT Week: 2024 Investor Conference June 4, 2024 8:45 AM ET

Company Participants

Andrew Power – President and CEO
Matthew Mercier – CFO

Conference Call Participants

Matthew Niknam – Deutsche Bank

Matthew Niknam

Okay, everybody can go ahead and please take their seats, we’re going to go ahead and get started with what I believe is the first session of the day. I’m Matt Niknam, Common Infrastructure Analyst here at Deutsche Bank, for those who don’t know me. We’re very pleased to host Digital Realty, President and CEO; Andy Power; CFO, Matt Mercier. Guys, thanks for joining us.

Andrew Power

Thanks for having us.

Question-and-Answer Session

Q – Matthew Niknam

So maybe just to start, set the stage for those maybe a little bit less familiar with the company, can you give us a brief overview of Digital Realty, some of your top priorities for the business and any opening remarks before we get into questions.

Andrew Power

Great. So Digital Realty, believe it or not is a 20-year public company now. This is our 20th anniversary. We’re the largest global owner and operator of digital infrastructure, data center and connectivity infrastructure, supporting 5,000 customers across 50-plus metropolitan areas on six continents catering to the digital transformation, hybrid IT of enterprise customers, cloud computing, some of our top customers and certainly the home of AI and AI infrastructure in the last 12 to 18 months, if not longer. So we are blessed with a sector with secular tailwinds of growth.

We’ve been investing in our platform for many years now and extending our capabilities. I think our priorities right now are probably threefold. One, really accelerate on the momentum we made in the enterprise cost space.

So we’re on a string of at least three quarters, north of $50-plus million of new bookings in that category, entering 130 new logos last quarter to our 5,000 customer base, selling multi-market, multisite, multiproduct to that segment. Two, we’re executing in the hyperscale arena for both cloud compute and artificial intelligence.

In the last quarter, we had record signings, about 50% of those signings were for AI-related use cases. The GPUs often talked about from the NVIDIAs and AMDs and others. They land with inside a server and then often in a fit-for-purpose data center provided by Digital Realty, which we were owning and operating north of 300 around the world.

So we’re catering to both enterprise AI use cases. We had a big announcement supporting the Novo Nordisk Foundation with NVIDIA just last quarter and then other partnerships in AI as well.

And then last but not least, we’re evolving our funding strategy. We’ve certainly come a long way in the last 12 to 18 months, now brought our leverage down to north south of 5.5x. And we’re thinking about what’s next to efficiently chart the funding of our growth and ultimately drive accelerating bottom line FFO per share growth.

Matthew Niknam

Great. Great. So there’s a lot to unpack there. Maybe if we can start around data center fundamentals. I think the industry has seen what many would call a pretty materially positive inflection in recent years, a lot of the dynamics you referenced.

Can you talk about some of the bigger factors driving this from a supply demand perspective and how that’s directly affected Digital Realty?

Andrew Power

What you’ve had is a series of incremental waves of demand that have been building and continuing and not being really exhausted, whether it is still the move from on-prem uses, data centers in server closets, office buildings, not efficient, not secure, not cost effective to our four walls the dawn of cloud computing, the globalization, the build of services, the cloud offering and now Gen AI, spring on a scene as an incremental use case.

And the first of those demand trends is still coming to shore. The second is still building and growing at a great clip. And the third is in its infancy, is maybe an exaggeration of its growth potential. And it’s happened in a time when numerous components to data center supply have been exhausted, power generation, power transmission, supply chains for substation components production slots for data center components, moratoriums, sustainability concerns, nimbyism that is really ground not necessarily a halt but certainly slowed and impeded on the growth of capacity.

Along that way, Digital Realty is now capitalizing on major investments we’ve made for several years in growing our capabilities servicing the full customer spectrum, having a greater value proposition in a time where the customers, whether it’s an enterprise customer or a hyperscale customer really are seeing tremendous value in our partnership and the services that we can provide.

Matthew Niknam

So power constraints. That’s been a huge topic of discussion amidst — especially if it’s ramping AI investment. So how is Digital Realty positioned here across your key markets? And what are you doing to accommodate rising customer needs?

Andrew Power

So we’ve — our operating portfolio today is just north of 2.5 gigawatts. So two DeLorean’s from back to the future, that’s what I could tore you through physical today. On top of that, we’ve got another three gigawatts of power capacity across the 50-plus metropolitan areas, ranging size and scale, whether it’s suited to colo enterprise customers to hyperscale large contiguous capacity blocks.

We’re certainly navigating new world as are others with certain markets having regulated, unregulated power providers saying, I can’t give you any more power until next year. We’re working with these utility providers and reallocating power, and we’ve shifted power around that we had idled at our substations and our suites to make sure we can deliver on our commitments to customers and also deliver growth capacity.

A sizable amount of leasing we just had in the last quarter was in the Famous Loudoun County, Virginia, which has been a bottleneck until 2026. And that work and being a long-term partner in that part of the world has allowed us to support customer growth in these times of need. We’re also looking at other ways to be innovative to bring alternative power sources, we’ve done bloom cells historically looking at natural gas turbine options.

We’re a leader in the sustainability front. This doesn’t solve the power pinch, but we’re 1.4 gigawatts of PPAs for wind and solar, true additionality of sustainable power to the grid, so trying to chart this in a sustainable framework as well.

And we’re fortunate that we didn’t get into this business a day ago or a year ago, we’ve been in it for 20 years, and we’ve been making long-term investments that essentially are now certainly coming to fruition.

Matthew Niknam

I mean any handicapping in terms of when, if at all, some of these power contracts get alleviated? And maybe to dovetail on that, is the bigger problem just power generation? Or is it more around the limitations of transmission and distribution infrastructure?

Andrew Power

My opinion is this is a very multifaceted and complicated easy button solution problem because we’re dealing with numerous markets where transmission is essentially the bottleneck immediately. And that navigating transmission, even with eminent domain is not easy, right? And we’re talking large-scale transmission lines, cutting through communities in order to power parts these markets. You’re also facing transmission issues. Some of them are on the ground now, like in countries like Ireland or Dublin.

Some of them are going to be around the corner because we are, as a country, depowering called non-green power sources at the same time that we’re trying to ramp this AI and innovation and technology and that is a challenging dilemma. And this is not — I don’t see a massive federal solution, state solution. It’s got municipalities involved, it’s got regulated utilities involved, it’s got an unregulated utilities, data center providers, big customers.

So on the other hand, I don’t think it’s an off– based on where our platform sits today, I don’t think it’s an awful outcome that the inventory or supply of data center just comes in a little bit slower, more deliberate, more planned, more thoughtful approach versus people call it getting overaggressive and create more volatility to the market. So now if I didn’t have three gigawatts on top of 2.5 gigawatts and probably be sitting different him.

But I think this is probably not an awful thing for natural growth of this infrastructure.

Matthew Niknam

So it’s a great segue. Pricing, obviously, has been a pretty significant tailwind for data centers in recent periods. So can you talk about what you’re seeing from a pricing perspective across the business, the outlook for this year and whether the favorable backdrop you’re seeing now represents a little bit more of a newer normal given the dynamics we’ve discussed.

Andrew Power

Do you want to take that?

Matthew Mercier

Sure. So when — it’s probably easiest to call out one of our largest markets, which is Northern Virginia, which is seen not only — it’s come off the trough, but a trough in the last couple of years, but is significantly rebounded from that. So we’ve seen pricing come from where it was in the ’80s and ’90s of KW to now the 150, 160 plus a KW.

So we’ve seen a substantial rebound in where pricing has come in Northern Virginia in particular, again, because that’s one of the largest markets in the world where you’ve seen a lot of supply, but then now you’ve seen that be absorbed — you’ve seen some of the supply constraints leading to that pricing pickup. And you’re seeing a similar dynamic across a number of our global markets.

Now you haven’t seen quite the same level of increase, but I would say the majority of our global major markets have seen price increases anywhere from low single digits to I’m sorry, high single digits to double digits over the last, call it, 12 to 18 months as you’ve seen the confluence of supply and demand factors that Andy just mentioned kind of work through, and we’re only seeing that we’re only seeing that dynamic, I think, continue across our global portfolio.

Matthew Niknam

And so where do yields, if we can talk about development yields, where do those yields on new developments sit today in light of some of these improved fundamentals? And can we maybe talk a little bit about how that varies across your key regions?

Andrew Power

If you look at our total portfolio, which is close to 0.5 gigawatt that’s under construction today, we’ve moved those yields, which these are large-scale projects and we’ve had already done significant leasing. We’ve moved those yields, call it, hundreds of basis points in aggregate to north of 10%. And there are certainly projects in that mix that are well higher than that given the running the rates in some of these markets.

Matthew Niknam

Let’s dig in a little bit into the AI theme. I mean it’s been very, very topical. I know it’s a real estate conference, but it’s obviously been a big theme in the markets over the last 18 months. Can you talk a little bit about the benefits, AI-related demands driving for the business today and how other parts of the business, for example, the traditional colocation or enterprise business could benefit over time?

Andrew Power

I can tell you definitively, and I put this in the context of our industry has certainly had a lot of thematic buzzword bingo come through that have not been as impactful, whether it was edge, computing or some of the other technologies.

But I would say AI has been incredible driver of incremental demand, not dissimilar to cloud. It is certainly been prevalent in large continuous capacity blocks. It’s been the buying amongst the biggest companies on the planet. The hyperscalers have been a major piece of it.

We are seeing new verticals come out as well and be buyers in addition and sizable capacity blocks. And we’re starting to see enterprise use cases, whether it’s my example Novo Nordisk or a partnership we announced with Oracle, that’s going to be supporting enterprise AI use cases. And fortunately, we are able to intercept this across the full customer spectrum. So today, it’s the large language model, the next leg of this as we move to inference, which I can’t guess what the size of that is going to be, but it’s a multiple of the first leg, and that multiple keeps increasing. It feels like every time you ask with some LNL steeped in the demand.

We got the move from public datasets to private datasets and move from consumer use of AI to enterprise use of AI. And it seems that it’s going to have long-term gravity around where the data sits, whether the data sits in our four walls in cloud compute post sitting within Digital Realty, Enterprise Hybrid IT sitting with Digital Realty or in major markets on devices than a proximity of the eyeballs and the mobility or offices or computers. So I think we’re going to be in this first phase that’s going to be large growth incremental to cloud, incremental to hybrid IT for a while. And the next phase is going to be down the road, but I don’t see how it’s not additive.

Matthew Niknam

This didn’t happen overnight. I mean, you talked a lot about the work you’ve done, let’s say, over the last decade plus to evolve the portfolio, broadening it from both a geographical and product perspective. So can you talk a little bit about how that’s helped improve Digital’s competitive position and differentiation in the marketplace?

Andrew Power

So we again, we — this company was built on a strategy that the strategy has evolved. It obviously started around the intersection of technology and real estate. It evolved to data center connectivity. It evolved to the full customer spectrum. It evolved to global approach being very important to add value to our customers.

We reviewed having size and scale as a key differentiator. And we’ve had to make some near-term painful but long-term beneficial investments along the way. And that’s from where we deploy capital to grow our capabilities, and that’s where we exited and recycled out of noncore pieces of our portfolio that we didn’t see longevity or robust and diverse customer demand. And I think all these things are now starting to provide incremental value. It’s certainly accruing to the customers that were growing with us to multimarket, to multi-countries, to multi-regions.

It’s accruing to us in our pricing power in both the enterprise zero to one and the north of one megawatt category. And it’s certainly these large capacity blocks that we’re helping customers with, we added bar pain for years to accumulate these. I mean we — next to the Dulles Airport, we bought a parcel land for $250 million. We — it’s produced pretty much Zero FFO since we bought that in 2018. We just did an easement to help the power company delivered a substation critical area for $90 million.

And the first building, which is 100 megawatts almost, we’ll probably have north of $1 billion of profit on for just 1/10 of that slide. Zero FFO in 2018 and now it’s coming to fruition today. So yes, we’ve been certainly blessed that these we were playing in a digital transformation, hybrid IT lens, cloud computing seeing the long-term growth, and AI has just been like an incremental massive leave window backs here. But it’s certainly fortuitous that these long-term investments we made are now certainly come to fruition.

Matthew Niknam

Maybe if we can pivot a little bit to the colo and enterprise business. I mean sub-1 megawatt leases, I think, right now make up a little over 35% of your annualized rents today. It’s an area, obviously, you’ve invested in heavily in recent years. What’s the value proposition you bring to customers? And is the strength in demand and pricing, how does the strengthen demand pricing relate relative to the hyperscale arena?

Andrew Power

So even with certainly incredibly attractive backdrop, serving hyperscalers in these pinch point markets. We have not changed our strategy going after full customer spectrum, growing our 5,000 customer bases each and every quarter selling them more connectivity solutions. We’ve been incrementally investing in our service fabric with Service Directory, private AI exchange. We’ve been investing in our high-performance compute, high-density colo. We just had now 170 of our 300 data centers, can go up to 150 watts per cab.

So we’re in a category where enterprises are looking to have global approaches to the infrastructure. We check that box where we’re differentiated versus a really small arena of competition is our infrastructure was not built the same way universally from data center one to data center 300, right? We came at this from the hyperscale lens with higher power densities. More flexibility for our customers to get today’s innovation and solutions.

And we’re looking to continue to invest in that arena and grow that arena, but also not forsaking these hyperscale customers, we’re selling — building and leasing 20s, 50 megawatts, 100-megawatt locations for what the trick or the next leg of where comes hyperscale is make sure you have the appropriate long-term funding model. That’s where we harness private capital over the last 12-plus months, and we’re going to continue to evolve that to efficiently make sure we’re growing our bottom line along the way.

Matthew Niknam

We’ll get into funding the business. And as I just want to follow up on the colo enterprise front. Any pockets of macro-related softness whether it’s related to higher interest rates, any effects on customers? Any effects just broadly on macro that have impacted that side of the house?

Andrew Power

So certainly in that segment, and maybe I was too bearish early on the concerns of an economic slowdown would have in the enterprise segment. But we’ve executed quite well for numerous quarters down in that segment, adding incremental customers selling more product installed base, putting up quarter after quarter of great results.

And I think the acceleration in that category is just getting started in terms of what we can do. We’d not see — it’s been very broad-based, financial services, manufacturing, healthcare, numerous segments. And even if we turn to an eco-backdrop where you do see some slowdown, you do see tighter budgets.

One, this is mission-critical necessities. It’s been — we are in a race of a limited competitive set, where we had a tremendous amount of value of cost performance and efficiency. And you look at the IT surveys, today, what’s number one and two on your list, data centers in AI. Those are the priorities for an IT executive, CTO.

Matthew Niknam

So I’m going to loop in Matt here. If we kind of put all this together and bring it back to the financial model, can you talk about — this is probably a favorite question for you guys on any conference call, can you talk about the longer-term growth algorithm for Digital Realty all the way from a top line growth perspective down to core FFO per share looking out over a multiyear basis.

Andrew Power

You’ve got eight minutes to do it.

MatthewMercier

Sure. Yes. I mean so we yes, that is one of the favorite questions and we talked a little bit about this on our last earnings call. But so I’ll call out a few things. First, if you look at our — if you look at our normalized revenue and EBITDA growth for this year when you exclude some of the asset sales and capital recycling we’ve done, we talked about talked about revenue growth and then call it the 9% area EBITDA growth in the, call it, 7% area.

So I think from kind of some of those fundamental metrics, that’s something that we think is repeatable from a top line down, we’ll say, closer to the bottom line. You then kind of take a little bit of a different lens. The way we’ve talked about sort of a baseline level of growth and how you build up to that from the way that we talked about some of our key metrics.

What we talked about is, again, from a normalized kind of baseline perspective, we would look to have our stabilized same capital cash NOI growth in the 4% area, that should translate roughly to GAAP, although there’s going to be some differences there when you go from cash to GAAP. So that provides a solid baseline of growth from what is, call it, 80-plus percent of our portfolio today that’s part of that pool.

And on top of that, the call it, $2-plus billion of development spend that we’re doing a year. Obviously, that can fluctuate, but given where we are this year, what we see in the outlook in terms of what we have already in place from some of the development JVs and other partnerships. We’d expect another, call it, baseline of 2% growth on top of that. And then somewhat of an offset to that is we do have debt maturing over the next couple of years. That’s obviously at rates that were at a much more favorable time given where the cost of capital especially on that side is today.

So we look at that kind of knocking off about 1%. So that all blends to basically a point where we’re at, call it, a baseline of mid-single digits which we expect should be able to go higher to as we modulate some of those items. That’s kind of the baseline that we would expect sort of long term on a normalized basis is mid-plus single-digit growth.

Matthew Niknam

Great. Great. I want to dig into leverage for a little bit. You’ve actually — one of the points we didn’t talk about, but I think the company has done a great job of is successfully lowering leverage. I think you were sitting north of seven turns at the start of last year.

You’re now sub 5.5 turns. Can you talk about some of the milestones achieved to date that helped get you here and where you see optimal leverage for the business in a currently sort of higher for longer rate backdrop?

MatthewMercier

Sure. So yes, I mean it’s — I mean we — we’ve had some of an incredible journey in a relatively short period of time on this front. So and that’s a broadly team effort started with Greg and the investments team really kicking into gear last year. Closing the number of joint ventures, not only of stabilized assets, but also into our foray into development JVs that we talked about and we’re able to execute on. So through that, I think we raised all in last year, ’23, I think close to $5 billion of capital, roughly, I think, three of that from capital recycling efforts from the investment side.

On top of that, we kind of supplemented that with roughly $2 billion of equity capital last year as well. We’ve kind of continued that March early into this year. Whereas a lot of that activity in ’23 was done, call it, post second half. We’ve now already been in a position where we’re at the low end of guidance on our capital recycling, call it, roughly $1 billion, a little north. We’ve then also supplemented that with additional equity raise that we did just recently, a little over $1.6 billion.

So call it overall $2.6 billion of capital raised year-to-date. And I think that’s really put us now in a solid position where we’ve — I don’t want to say achieved our leverage goal. So we’re — I mean, we have pro forma basis, achieved what we’ve talked about in terms of being at 5.5x on leverage.

And I think it’s really put us in a position we want to be able to capitalize on what we see is an opportune time and opportune place within our investment within our investment horizon to be able to capitalize on what is a growing and higher yielding, higher return opportunity in our global marketplace.

Andrew Power

I think there’s something unique here is — this wasn’t just a mission to get the balance sheet back to the target leverage. It was evolving the funding model to make sure we can have a mousetrap that can generate per share growth for multiple years. So Included in that was partnering on the development side, large-scale campuses with a great partner. But what that allowed us to do is that partnership is developing capacity that won’t come online until 2026, ’27, ’28. We obviously are maintaining a minority stake in that.

We’re getting a fee economics along the way. And we also are maintaining on our balance sheet, that three gigawatts I told you about, including the soonest delivering projects that are delivering the 2025s, only ’26s. So all this was thematically about when it comes to hyperscale, which is so capital intensive, so large that we can essentially harness public and private capital and chart a continuous organic growth story.

Matthew Niknam

And so with that in place with leverage now and a little bit more of a comfortable place in the mid-5s, how do you prioritize uses of excess cash from here? Is there an opportunity for more M&A? It sounds like you’re pretty bullish on reinvesting in the business. Just maybe if we can walk through some of the capital allocation priorities.

Andrew Power

Fortunately, our we believe that we’ve assembled to call it, most critical strategic puzzle pieces to our global footprint. And by and large, that we’ve exhausted the landscape. So we don’t see anything like we’ve added to our fold out there anymore.

So yes, we’re adding incremental land capacity adjacent to some of our campuses, which are just logical additions to multimegawatt campuses where we have installed customers, they want to grow with adjacency. We’re investing in bringing online that multi-billion-dollar development in those three gigawatts, but I think that is, by far, our best use of capital is driving returns and growing our customers’ footprint than what we have today.

Matthew Niknam

Excellent, excellent. And so if we sort of tie this all together, just to wrap up here, if we’re sitting here a year from now, what are some key milestones you would have liked to achieve looking back sitting here in June of ’25?

Andrew Power

I think the three things. One, demonstrative acceleration in our enterprise colo zero to one megawatt. So if we’re talking the 50s today, I want to be talking to a material higher number of productivity in a year’s time on a road map that is going to really bring that number to a much higher number of customers, connectivity and signing into that category and growth. Two, this is an environment where we’re seizing upon these hyperscale AI and cloud compute opportunities. So I would imagine a backlog that is much higher than what we had said on today at very attractive risk-adjusted returns.

And honestly, just not even risk-adjusted just very high attractive returns. And three, an evolved funding model any further into the future when it comes to our hyperscale business so that we can keep that — we can keep developing for these customers. We keep growing these customers and then obviously, having the right funding sources that doesn’t slow our growth to our bottom line.

Matthew Niknam

Great. I think it’s a great place to end it. Andy, Matt, thank you.

Andrew Power

Thank you.