DigitalBridge Group, Inc. (DBRG) Earnings Call Transcript

Good afternoon. It's Michael Funk from Bank of America. I'm a data center analyst. And very happy today to have Colony Capital and their Chief Executive Officer, Marc Ganzi, with us. For anyone not really familiar with Marc, I mean, Marc has been doing digital infrastructure since well before we even had that name, for over 20 years now. He's been creating companies, investing in towers and in data centers. Previously, he was the founder of Global Tower Partners, which was sold to American Tower in 2013 for $4.8 billion. Prior to that, he led investments in towers in Latin America. And he joined Colony Capital after the acquisition of Digital Bridge Holdings and has been CEO now for, I believe, just over a year of the company. A very unique story and a strategy. A bit different than our companies you've heard from today. So I'm going to invite Marc to open up by talking a bit at first about some of the high points of the public strategy over at Colony Capital. So once again, Marc, thank you for being with us.

Yes. Thanks, Michael. A pleasure to be with you again, and great to have a chance to talk with your clients in the BofA ecosystem. So thank you.

Do you mind giving us that high-point overview of the different assets you pulled together at Colony Capital and in the strategy?

Yes. Thanks. Look, the strategy is pretty simple. We've got really 2 core digital businesses that we own and operate. One is what we call Digital Operating, and the other segment is called Digital Investment Management. So first, maybe we could start out with Digital Investment Management. So Digital Investment Management is a series of 18 investments that we've made now with the combination of our balance sheet capital acting as the general partner. And then, of course, with the help of investors globally, we raised billions of dollars of capital on a global basis to co-invest side by side with us, either in SPVs or in a fund structure. So most recently, we've just announced that we fully invested that of our first fund, Digital Colony Partners I. That was about a $4.1 billion fund. And alongside of that fund, we've made numerous co-investments into some of the businesses like, for example, Zayo and Vantage Europe that required incremental capital. But really, those businesses are fantastic from our perspective. And for public shareholders, it's great, too, because it's the chance to marry up our balance sheet with third-party capital and give public investors the chance to own what we call the Digital Colony converged ecosystem, which is fiber, towers, small cells and, of course, data centers. And so that's a rather substantial portfolio now. There's almost close to $25 billion of assets under management in our digital platform. And what's also great for our public shareholders is, obviously, the balance sheet capital at work is a great way to get a total return. But in addition to that, we get management fees, and we get incentive fees. So we share in those incentive fees with public shareholders. And the management fees, given the duration of our funds, most of our funds are 10 to 13 years in duration. We get great consistency in our management fees, which is part of our digital earnings. So that's 1/2 of the ledger. The other half of the business is what we call Digital Operating. And today, Digital Operating, we have 3 kinds of investments all candidly in the data center space, which is great because that's what we're going to talk about today. But first, it started with a significant investment, a control investment in DataBank, which is the largest private edge data center business in the United States. And most recently, we announced that we acquired zColo, which takes us to 66 data centers, 29 markets and really makes us a powerhouse in edge computing. And we're really excited about the combination of DataBank and zColo. So that's one investment we've made on our balance sheet. A second investment we made back in July, Michael, was Vantage YieldCo. And Vantage YieldCo is 12 of the best hyperscale data centers in the U.S. We developed those data centers in one of our private businesses, Vantage. And in turn, once those data centers reached 94% occupancy, we put them into a permanent capital vehicle using our balance sheet capital, alongside of CB Richard Ellis Investors out of Toronto and along with a series of pension funds. So we raised $1.35 billion of external capital. We deployed about $200 million of our balance sheet capital. And what we got in return are long-term leases with investment-grade counterparties, weighted average lease duration of over 11 years, 3% escalators and over 90% investment-grade income. And so this is really the best of the best of hyperscale data centers: incredible exposure to long-term leases with investment-grade counterparties in some of the best hyperscale markets in the U.S. And so that's the second layer of our data center investing at Colony Capital off the balance sheet. The third thing we've done is, through DataBank, we made a strategic investment in the edge presence. Very excited about this. These are those micro data centers at the base of cell towers. And so there, we've made a -- we're not at liberty to discuss the percentage of the company we own, but I would say it's a major and significant stake of the company and one where we have a lot of control over the future of that business. So there, we're using our close to 20,000 cell towers at Vertical Bridge. We've also been open to partnering with SBA and American Tower, where they share in our success with us. And we're deploying micro data centers at the base of cell towers, where we're bringing really that notion of compute and applications adjacent to mobile infrastructure. So bringing that close to the LTE radios and, eventually, the 5G radios and creating what we would term low-latency edge solutions on a micro basis. And so at the end of the day, on the balance sheet, the way we think about it, Michael, is there's sort of 3 concentric rings. The biggest ring, of course, being hyperscale data centers. Then imagine another small ring inside of that ring, which is edge data centers with DataBank; and then a smaller circle -- a much smaller circle, which is edge micro data centers, inside of that. So really delivering the entire U.S. data center ecosystem off the Colony balance sheet, where we feel like that's highly strategic and ultimately where the future is going. So a lot happening today at Colony as it relates to digital infrastructure and, obviously, very active on the balance sheet, deploying our balance sheet capital to own the best data centers in the U.S.

That's a great overview. And I wanted to hone in a bit more in the data centers before moving on from there. Can you walk us through what your data center ambitions are, both in the U.S. and internationally? And will this largely be organic or inorganic growth?

Well, I think it's a little of both. I think we've demonstrated the ability to step up and do big M&A trades like zColo for $1.4 billion. We've certainly demonstrated our ability to build data centers in the U.S. and Europe at Vantage. I mean, I think we've got a tremendous track record of success there. We entered Europe about 2 years ago with the Vantage brand after spending the last 5 years with Vantage in the U.S. And the key there is really taking a customer-centric approach and being "shovel-ready" for our big logos. And that's really where we've excelled. We've been able to get to those markets early in Europe, really trying to stay out of the flat markets. The flat markets in Europe, between Frankfurt, London, Amsterdam and Paris, have been largely pretty well built. And candidly, power constraints in Frankfurt, zoning constraints in Amsterdam, real estate -- real property interest, land constraints in London have really made the flat markets kind of hard. And so what we decided to do in Vantage Europe is we decided to build and do a little bit of acquisitions, too. We bought NGD, which was the largest contiguous data center in Europe, over 150 megawatts, in Cardiff, Wales. We bought Etix, which had a sizable presence already in Frankfurt. We had already secured land and power. And so we took over that business, integrated it into Vantage, and that gave us our footprint in Offenbach, just outside of Frankfurt. And then we put shovels in the ground in Berlin. We put shovels in the ground in Warsaw, Milan, Zurich and really focusing on those next hyperscale markets. And remember, in Europe, Michael, you've got data sovereignty laws. And so when you have data sovereignty laws, it means that a lot of these web scalers need to have not only application -- high-power density applications in those markets, but you also need storage, right, because of the laws that are resident in Europe. So you've got compute and storage going on at the same time, and we've really kind of hit our stride with the web scalers in Europe. And super active in 8 markets across Europe, a deep pipeline. We have over 278 megawatts in our pipeline today in Europe. I think that would be probably one of the largest European pipelines of all the hyperscale players in Europe. And so that, from my perspective, for Europe, it was a little bit of buy, and it was a lot of build, but it demonstrates our flexibility to do buy and build at Colony. I think in the U.S., really not a lot of hyperscale assets to acquire. So we've been pretty busy with shovels in the ground. We did announce an acquisition 2 days ago in Vantage U.S.A., which is Hypertec, another acquisition in Canada that goes along with our 4Degrees acquisition. And so -- and really, in the province of Québec, where we have low power and we've got a dominant footprint in Montreal and Québec City, we've been very active in those markets and have really built a dominant market position between our greenfield and brownfield. We like Northern Virginia. We don't see a ton of differentiation for us in that market. So we've had a significant plot of land. We have the capacity to go to a little over 100 megawatts. We've already turned on our first 25 megawatts. We've leased about 13 there, got a bunch in the hopper. And so NOVA's good. It's not -- I would say NOVA, for me, is a bit disappointing. I would say Santa Clara has exceeded my expectations. We're the largest provider of hyperscale data center capacity in the Bay Area. And there, we've succeeded a lot. We've just turned on our eighth data center in that market, a brand-new, over-64-megawatt facility in Santa Clara, which we're rapidly filling up. And then in Quincy, Washington, playing in sort of the outskirts of Seattle, a couple of hours outside of Seattle, we've got 2 significant data centers there, the capacity to keep building. And that's been a nice market. It's sort of a low-cost market for the Pacific Northwest, keeping us sort of out of Seattle and Portland. And then we've been -- obviously, we've looked a lot and we have shovel-ready capacity in Goodyear, Arizona. So we're kind of all across the United States. And we look at each market, Michael, as its own market. I don't think we take a sort of shotgun approach. We spend a lot of time evaluating a market to find out if it works for us. And the right parameters have to be resonant for us to go there. So I would say in the U.S. right now, hyperscale, we're more about building. On the edge side, we're buying and building. We just turned on MSP 3, Minneapolis 3, at DataBank. And we'll integrate and close zColo. We'll start integration December 1 when we close zColo. So we're very busy in both Europe and the U.S. on the edge and on the hyperscale side. There's just a lot going on with us right now at Colony.

No, That's great. It's great color, Marc. And actually, I got a question from a client here that's somewhat related to my last question and a bit more pointed. And it's, in the long run, what general portion of digital assets will be acquired versus built from scratch? And the second part of the question or comment is, with industry multiples as high as they are, it seems like returns from new builds should be higher.

Yes, that's right. It's a smart investor or they've been listening to me talk at conferences in the last couple of weeks. I'm pretty bearish on M&A right now. I think it's really expensive. I mean we see M&A prices and we say, gee, at those prices, we're a seller. So we've been pretty laid back on the acquisition front. I think Hypertec was a great example where we were able to deal on a unilateral basis with the target. And we had the cash, and we acted quickly. I would say, in other situations, we're trying to be thoughtful. I think, of the new deals we're doing in the fourth quarter, a lot of them are smaller platform deals where we're acquiring a small platform, Michael, and we're backing a management team to go build. So we look at what we're doing in Asian data centers. It's all building. We look at what we're doing in U.S. cell towers right now. Obviously, we bought Eco-Site this week inside of one of our portfolio companies, Vertical Bridge. But the reason we bought Eco-Site is because they're a builder. They're a developer like us. So we have a big conviction around building towers in the U.S. We're building a ton of towers in Latin America between Brazil, Colombia, Chile and Peru. We just find where the currency is. There's a huge opportunity to build at historically low USD prices, which sort of counterbalances some FX. So in markets where we're able to build towers for USD 50,000 to USD 85,000, that makes a lot of sense because we look at tower deals where -- we look at the Hutch deal in Europe with Cellnex in that math. From a pure replacement cost, that's almost 4 to 5x replacement cost what they're paying per tower. So that's expensive, right? That feels that feels pricey. So whoever told you that question, that's correct. I would say we're probably spending greater than 50% of our time today on greenfield, and we're spending a lot less time on acquisitions. And we think also, we're heading into a marketplace that's going to get tougher. We think picking up the bar bill for COVID is going to be hard on a lot of countries. We think it's going to put pressure on -- not only on our central bank, but other central banks. And I think the world is going to take a couple of years to recover. I think this will be -- this isn't a magic elixir that gets solved by a new White House. It doesn't get solved by pumping more liquidity in the system. There's -- at one point in time, we're all going to have to answer for what has happened. So it's -- I think it's a good time to be cautionary. It's a good time to be a builder. And the great thing, I think, from my perspective as the CEO of the firm is, given our 26 years of history of being an operator, today, Colony Capital, we can build. We have a great history of building real estate. We have a great history of building digital real estate. And I think what investors get about our story now going forward is we have the capability to pivot between greenfield and brownfield. And that's not a story that all of the digital REITs can tell effectively. And so we can tell that story effectively because we're out executing. And I think it gives us a really interesting spot to occupy in the digital REIT ecosystem.

And I wanted [ you to tell me about some of ] the integrations. You mentioned earlier a number of the acquisitions in the operating units. How do you plan on pulling together DataBank and zColo?

Well, look, thankfully, first of all, the 2 businesses didn't compete. They actually happen to be in markets where we were not. I think we only had a grand total of 2 overlapping markets. So from that perspective, the integration from an engineering and sales engineering perspective was quite additive. zColo has people and facilities in places we didn't, which was candidly a big reason why we wanted to do that deal. And so from that perspective, we get a lot of new footprint. We get some talented people that were working in zColo. And now we're giving zColo a lot of capital and a lot of attention. And sort of when it was stranded inside of Zayo, it wasn't getting always the time, attention and care as an independent business unit. It was really used to help continue to extend fiber network into a data center environment. So we're taking advantage of that. There's a huge opportunity to make that business better, make it run more efficiently. We want to give it capital. We want to invest in the people of zColo. And so the integration plan will be to really, first and foremost, integrate the people. And then the second is really just integrating the processes around how we spend our money and where we put our CapEx; third, really bring together a bigger ecosystem cross-connects. We have over 30,000 cross-connects now, making us one of the largest interconnection players in the United States. And last but not least, just really focusing on DataBank's customer set. We have a tremendous customer set that can -- now we can sell into the zColo product. zColo has a ton of customers we didn't have. And so there's a real opportunity for us to scale the customer relationships. So I'm very excited about that. I think there's a huge opportunity there as well. And this is one of those unique acquisitions where, potentially, from a cross-selling perspective and a sales force engineering perspective, we believe it's kind of 1 plus 1 equals 3. And given where edge workloads are going and addressing those Tier 2 and Tier 3 customers that are putting capacity and adding hyperscale and web scale capacity on the edge of the network, DataBank is really positioned well to seize those 0.25-megawatt to 1-megawatt workloads. So a lot to be done right now. It's a pretty exciting time, and we're incredibly well positioned to capitalize on it.

You mentioned zColo was sold a bit different, under the old banner, I guess. Maybe kind of it was presented differently to customers. Is there a period of churn or transition where you have to work through the old customer base and the products they were sold? Or can those very easily be integrated into the Colony Capital platform and your digital -- your data center solution?

So look, I mean, part of the calculus of selling zColo was embedded in the fact that, that was already happening. So Zayo decided to leave the managed cloud product services about 1.5 years, almost 2 years ago. And so when Dan Caruso made that strategic decision to exit managed services and just focus on infrastructure, the customer base was already rotating at zColo. And that was obviously evident in the results when they were public. And since we've taken ownership of Zayo, we've watched that wind-down of those hybrid cloud customers. And now what we see is a big opportunity to sell through to edge customers as well, which is something that we've been quite good at. So we think probably there's some nominal churn at the beginning that we'll deal with, but there's also huge upside, massive upside to sell through to customers that previously hadn't been sold through on the zColo platform. So we're excited about that, maybe a baby step backwards -- if but multiple steps forward as we open up that inventory to new customers.

Got it. What are the main differences between the Vantage and the DataBank market and operations? And how does that -- how do you see that playing out globally?

Well, look, I think, first of all, they just -- from a workload perspective, they cater to very different workloads. So that's the first difference. Everything we're doing at DataBank today is sort of 1 megawatt below. And everything that Vantage is doing is kind of 1 megawatt and above. So there's a nice bifurcation from a workload perspective in terms of what the customer is looking for, and we've been able to really isolate that quite well and create proper white space for both logos. I'd say DataBank is very focused on U.S. Vantage is now a multinational player. Vantage has been able to prove itself in Europe and proving itself in the U.S. and Canada, where DataBank really focuses on hardcore focus on edge computing in the U.S. And we think that edge compute market in the U.S. is massive. I think we mentioned that on our quarterly call. We think that the marketplace today is less than 20 basis points of workloads around the globe are on the edge. And we think that shifts to about 10% of global workloads will shift to the edge over the next 5 to 7 years. So there's close to -- when you think about that and you just calibrate that for a second, there's about 1.6 million servers that are going to move to the edge. That's a lot. That's a lot of workspace that needs to be fulfilled and a lot of real estate that needs to be leased. So we're obviously pretty excited about that, and that's why we're putting so much energy into the edge.

And I want to talk about the edge a bit more, if we could, Marc. I find it hugely interesting [ that commentary in ] most of my presentations today. And you just mentioned that you see a significant growth in workloads at the edge. And so I mean just curious, what drives that? Because if I think about edge as being a solution for latency, if that's correct, you can solve for a lot of wireless-based latency simply by going to 5G. You don't necessarily have to be out at the bleeding edge. You can be kind of at the near edge, right? So what are the applications in your view that are going to drive that growth in the edge compute?

Yes. Look, there's a bunch of applications, right? I think the case studies that we're seeing are no longer science fiction. I think our customers are driving those types of case studies. And so I think some of the first things we're seeing is certainly around IoT. A lot of device-driven, device-to-device connections that are ultimately fueling efficiencies in consumer electronics, fueling efficiencies in the home, these things are no longer science fiction. And so that linkage between the cloud players and their ability to access those IoT networks and then the ability to access, obviously, homes and cars, so areas that are most impacted are not only what I would call semi-autonomous cars because cars aren't fully autonomous yet, but cars are literally taking down gigabytes of data. And all of that is being enabled by edge compute network loads where you've got base stations and applications sitting not in the CBD but sitting in the suburbs and in the corridors. And so all of that is enabled through IoT. And simple things like diagnostics to your refrigerator, trying to figure out what your next Amazon order is going to be delivered to your door based on your consumption patterns in your refrigerator, thinking about your thermostat and energy efficiency and how that ultimately kicks back to the grid and how that ultimately kicks back to your utility bill and ultimately kicks back to the devices your using in-house -- in your house and how much power that's being consumed, all of that can be managed virtually on the cloud. But to do that, there's a physical element that has to be close. And so we look at that and some of the things that we've been doing on the autonomous driving vehicle with NVIDIA and some of the major car manufacturers and Uber. Cars aren't fully autonomous yet, true. But cars are taking in data. They're taking in traffic patterns. They're observing the conditions of the roads, and they're helping the driver make decisions. And we've seen that on some of the new General Motors cars that are very sophisticated. Once you get them on major highways and you've got stabilized conditions, the cars can virtually drive themselves. And that's not a magic trick. That's something that's rooted in a lot of IoT infrastructure using mobile infrastructure and, obviously, using the intelligence from the data center. But for the car to make those decisions quickly, you've got to have a low-latency environment where that quickly gets back to the car and the car gets back to server, server iterates back to the car, and decisions are made in milliseconds. So this is the world we're living in. And to enable these things to happen on the edge -- and we're just starting, right? This is just, if we're playing -- imagine for a second, Michael, we're playing a 9-inning baseball game. We're probably in the top of the first in terms of edge computing. And so these use cases are only going to evolve. The cloud players understand the importance of mobility and how the connected device or the actual device itself is what is the hub for our commerce activities, for our communication activities, for our data activities, for our entertainment. And ultimately, to enable those applications to perform at the best level, you've got to have great network and you've got to have great network infrastructure. So that's how we're thinking about it. I can spend hours on this topic, by the way, but I know we're limited today. But I wanted to try to keep it at 50,000 feet and easy for the audience to digest today.