Download Transcript
Spencer Levy
Digital infrastructure is a relatively new term, a pretty new asset type in commercial real estate. In all its complexity and reach, digital infrastructure is so much more than servers storing information that makes the world go round. It's a network of architecture that's driving the evolution of computing speed, artificial intelligence and more, as well as newfound needs for power and other resources. On this episode, a global adviser and a digital entrepreneur who specializes in a business that's now much bigger than four-walled data centers and is growing faster than ever.
Marc Ganzi
So when you look at the total wallet size of what digital infrastructure is today, it's a $13 trillion marketplace. And I don't think a lot of people really recognize how big that marketplace is.
Spencer Levy
That's Marc Ganzi, founder and CEO of DigitalBridge, a firm with more than $70 billion of assets under management in five key verticals of digital infrastructure, data centers, cell towers, fiber networks, small cells and edge computing. DigitalBridge is Mark's sixth startup in a career that spans nearly 30 years in the telecom and datacenter industries.
Pat Lynch
I view this opportunity, not just for today or two years from now, but there's a long run here that is really exciting.
Spencer Levy
And that's Pat Lynch, Executive Managing Director and Global Head of CBRE Data Center Solutions. Pat also started out in the telecom side of the business before joining CBRE a dozen years ago. He's helped build an advisory business that now also manages more than 700 data centers in over 50 countries. Coming up, we get under the hood of data center real estate and digital infrastructure to find out how it all works and why the business is booming. I'm Spencer Levy, and that's right now on The Weekly Take.
Spencer Levy
Welcome to The Weekly Take, and this week we are fortunate to have today Marc Ganzi, the CEO of DigitalBridge. Marc, thanks for coming out.
Marc Ganzi
Yeah, thanks. It's a pleasure to be here. Thank you for the invitation.
Spencer Levy
Great to have you. And welcome back to the show, our friend and colleague, Pat Lynch, Executive Managing Director, CBRE. Pat, great to see you.
Pat Lynch
Spencer. Thanks. And Marc, thank you so much for doing this today. It's a pleasure seeing you.
Marc Ganzi
Yeah, it's good to be amongst colleagues and friends. So we'll have a nice discussion today.
Spencer Levy
You bet. So for the benefit of our listeners, Marc, just give us a little bit more specifics on DigitalBridge. What is it? What's your portfolio look like?
Marc Ganzi
So DigitalBridge, today, is the largest global asset manager in the digital infrastructure space. And so today we manage about 75 billion of assets on a global basis. We do that from Singapore to London, Los Angeles, New York, and then our headquarters is in south Florida. We really feel like in the last decade, digital infrastructure has risen to a level of prominence, largely because there's about a $500 billion annual reinvestment back into our sector. So when you look at the total wallet size of what digital infrastructure is today, it's a $13 trillion marketplace. And I don't think a lot of people really recognize how big that marketplace is. And that's, to be clear Spencer, that's just the infrastructure piece. So that's the plumbing, right? That's the physical, hard infrastructure. Could be a tower, could be a fiber optic cabling round, it could be a data center sitting right here in Virginia. But it's a big, big industry today and it's growing and it continues to grow. We'll talk about that today in terms of the implications of what it means from an AI perspective and what's the plumbing associated with AI. But we're constantly being challenged in our industry. What I love about our industry is, Pat and I wake up every seven years and the chessboard changes. There's some innovation that's always happened, and I think what keeps me excited and keeps me engaged is that investment cycle every seven years creates a paradigm shift, and that paradigm shift creates hard infrastructure. It's fundamentally real estate. I mean, at the end of the day, what we do is we acquire land, we get entitlements, we got to go build something. Today we're building for, I think, some of the most important companies on the planet, whether it's Nvidia or whether it's CoreWeave or the stuff that we're doing with Google, the stuff that we've been doing with Amazon and Microsoft, Verizon, AT&T, the partnership we did with Deutsche Telekom last year, it never stops. There's always an opportunity and there's always an opportunity to invest and to continue to keep building. So I think that's why we've gone from literally nothing ten years ago to 75 billion of assets, because we're playing in a tailwind that's hard to quantify, and I don't see that slowing down anytime soon, which is great for commercial real estate.
Spencer Levy
You know, Marc, I'm really glad the way you described it, because I always ask our guests, are you in the real estate business? And I get a mixed answer there, particularly when we're dealing with digital real estate. But it sounds like you are in the real estate business. Would you agree with that?
Marc Ganzi
Well, I am, because the fundamental underpinnings of everything that we do, Spencer, is signing a lease. And so, therefore, if you're leasing space, you're in the real estate industry, whether you like it or not. Maybe it's not sexy to say that today, but that's what I've been doing for 29 years. I studied real estate under Peter Lindemann at the Wharton School. The people that I looked up to were Sam Zell and Dick Simon and people like that. And those were my mentors and people that formed my career. And I think the reason we've been successful is because we've stuck to that. We haven't tried to be something that we're not. I'm not Blackstone, I'm not KKR, I'm not Carlyle. Great organizations, all friends of mine at those organizations. But what we are at the core is we're developers. And I think what sets us apart is the fact that this year alone, we've got $7.8 billion of shovels in the ground right now. And next year we're forecasting we'll have 10 to 11 billion of new construction just to keep pace with public cloud, private cloud and AI. I mean, that's stunning. How are we going to raise the capital to build that is an entirely separate conversation, which we're going to have today. But it's exciting, right?
Pat Lynch
Just to touch on the real estate question, Spencer, it's an industry that changes so quickly, right? And huge credit to Marc and his team. They've been ahead of the curve on that and their investors have benefited from that. I think about it in the context of assets that I was building 20 years ago and how different they are today. It's amazing, and I think the industry is also lacking for talent. We spend a lot of time looking for new talent in the industry and two of my four kids are in the digital infrastructure space, which I think is validation of how I view this opportunity, not just for today or two years from now, but there's a long run here that is really exciting. A lot of the stuff that Marc's organization is building for hasn't been invented yet, which is really cool.
Spencer Levy
Let's take one step back for just a moment. I've been in the business for 30 years. I've done data center deals, I've done cell tower deals. But we're talking about something here that's more expansive than that. We're talking about the fiber, the towers, the data centers, a whole range of different types of real estate. Pat, can you just walk us through that chain of technology please?
Pat Lynch
Yeah, I'm humbled you're asking me that in front of Marc, so I'll take a shot at it, and he’ll keep me on the rails. But the other… I've started to use the comment, Spencer, that a data center without network is a really expensive refrigerator, right? So if you think about it in that context, you need for us to do this podcast today, right? The connectivity required is going to go through fiber, going to go through towers, going to go through data centers. And then the people that are going to be listening to this are going to be using that same infrastructure. So we need, as an industry, and by the way, this is global, right? You've built a tremendous podcast, so people literally around the globe will listen to, and they're going to use the digital infrastructure that Marc’s company has built and invested in to do that. So that is fiber, that is towers. Those are data centers.
Spencer Levy
Well, Marc, I know that you could lease space in a data center. I know you can lease the ground underneath a cell tower. Walk me through the real estate along the way as you go from the fiber. Walk us through the nuts and bolts of that.
Marc Ganzi
Simply put, Spencer, that it's just the capillaries. I mean, it's ultimately the capillaries that connect everything that allows us to communicate. And if you don't have good fiber connectivity, it's really hard to create some of the applications and some of the content and the use cases that we use day to day. So there's this scientific concept called latency, and latency is just the speed at which it takes to bring one packet of data from one place to another place. And everything that we're doing on the fiber sector today, Spencer, is focused on reducing latency. So if you can reduce latency and you can take that from 10 milliseconds down to 5 milliseconds to 1 millisecond, and now we're building for like tenths of a millisecond. And by the way, in the AI world, 1/10 of a millisecond actually really matters from a compute perspective. Now I'm getting a little scientific, but ultimately, if you think about, you just looked at your phone, okay? On your phone right now is probably running 60 applications, no less than 60 apps. If you track the average human being who has an iPhone today, they have in the background 60 apps running. Those 60 apps are ultimately going to an antenna that somewhere within four blocks of here, that antenna receives the signal. It goes into the radio that's on the rooftop, because there's only rooftop antennas here. Ultimately, that radio is connected to a piece of fiber that goes to the basement of the building, which is the MPO, the minimum point of entry. And then that fiber connects to a curb and then that dark fiber eventually goes back to either a data center or CO, a central office that's controlled by in this neck of the woods, it would be Verizon. But if it goes back to a data center, you're ultimately then going to that data center, and then from there you're connecting out to the cloud, the public cloud. Once it hits the public cloud, it then goes and finds the servers for those 60 application providers, transmits the data, then brings it back, brings it back to the fiber, brings it back up to the rooftop, connects the antenna, beams it back to your phone. All that just happened in this [snap of finger] Okay. Now, that was a second. That wasn't in milliseconds. In the digital world, that's an eternity. But none of that happens, as Pat says, without fiber, dark fiber, high speed connectivity, low latency solutions. And to your point, the real estate of that is tricky.
Spencer Levy
I don't know how you transact in fiber. Do you lease the ground under the fiber? Let's go real estate on this a little bit.
Marc Ganzi
What we sign in the fiber space are really two types of contracts that look like a lease by the way. If you read it, you’d look at it and go, Marc, that's a lease. Come on. But really what it is, is you have a license agreement which license agreements dictate how fiber runs into commercial real estate. Pretty much all of the fiber that sits in the riser space of an office building like this is dictated by a license agreement. There's no more leases. The second piece you have is what are called IRUs, and IRUs are an indefensible right to use agreement, which allows you to occupy a conduit and think of a conduit as the real estate. So a conduit's a big plastic sheath. And inside that plastic sheath, you have anywhere from two pairs to 400 pairs of fiber. And then what you're doing is you're getting a right to use that sheath, and that sheath sits underground, it sits in a right away, it sits along utility lines. Fiber runs everywhere, but it's principally underground. And so a lot of those contracts are 15, 20 and 25 year contracts. And those contracts are really to dictate the long haul relationship, which is taking information from Atlanta to Dallas or taking information from Kansas City to here to Washington, D.C. And those contracts are good contracts. They're with investment grade customers. And I try to explain to people that when you have a dark fiber route and let's say the three of us decide we want to go build a new fiber route from downtown DC out to Loudoun County, to go to connect to a data center. And we build that with 400 pairs of fiber. The way you should think about that is like almost like an office building tilted on its side, right? And your occupancy is, each glass strand dispenser is an opportunity leased to a customer. So if you've got 400 pairs of dark fiber, think of that as an office building where you could lease to 400 tenants. The reality is most customers take 8 pairs, 20 pairs, 50 pairs. So when you build a really good route from downtown DC out to data center alley here in Virginia, you're going to have anywhere from 6 to 12 customers sitting on that route. And if you build it economically efficiently, you're going to find the cash on cash yield on that from a real estate perspective would be about a 12 to 14% cash on cash yield. That's a pretty good office building in this market today. And most of those tenants are going to be Amazon, Microsoft, Google, Meta, principally investment grade tenants signing 15 to 25 year leases. It's a different kind of real estate is what I would put to you. It's just we have to think about it as real estate owners, not think about it as technology providers.
Pat Lynch
And I think Spencer, just to play off of Marc’s comments. CBRE saw this a few years back, and that's when we acquired NEF and brought that network advisory business to our clients, to our landlords, where we're doing rooftop asset sales, where we're doing fiber optic business with a lot of Marc's companies, bringing that real estate component and connecting the dots from an advisory perspective. So our platform has grown and benefited from the exact description that Marc gave, which was fantastic.
Spencer Levy
And so the way we describe it and the way you just described it, Pat, is network advisory. It’s not data center advisory. That is part of the puzzle. Is that a fair way to put it?
Pat Lynch
Yeah. To the point of, you know, we were fortunate to have you speak at our event a week ago. Thank you for doing that as well. We had 250 clients and colleagues from around the globe, including Marc’s companies were represented there, as well. And that conference, four or five years ago, was a Data Center Solutions Conference. The last couple of years, it's now been rebranded as our Digital Infrastructure Conference, to play to the audience that's there and to where the business is going.
Marc Ganzi
And give credit to my friend Bob Sulentic. I've known Bob 23 years. We used to lease the rooftop space for Trammell Crow's real estate and Bob was my first client. So CBRE is one of our great partners, as you know, Pat. They also have a funds business where they go out and they represent third party capital. And I think you guys have well over $1 billion of capital invested with us in data centers, in towers, in fiber. And it really shows you, I think, what Bob's vision of CBRE is, is that to be a global asset manager, you have to not only have the advisory piece, but you also have to have the ability to invest. And this is where I think the future of CBRE is. Part of the future of CBRE is investing in the digital economy. So we're a huge teaming partner to CBRE and I give Bob a lot of credit for that.
Spencer Levy
So let's talk about the funding business, the money side of this, whether you're technology, whether you're infrastructure, whether you're real estate. Traditionally speaking, these are different buckets of capital. But today, what we're seeing, in particular with so-called infrastructure capital, a real merger between what would have traditionally been real estate and infrastructure capital. Even CBRE IM has an infrastructure fund. How do you see it?
Marc Ganzi
It's changing. The bucket of capital today is reeling in real assets. If you go around the world and you talk to allocators, you talk to LPs, there's an open admission, Spencer, that real assets is real estate and infrastructure. And a lot of these big allocators, whether it's CalPERS, GSE, Ardea, Aimco, Allianz, AXA, all these LPs are looking at the real asset class. And ultimately, most LPs today will openly admit they're under-allocated infrastructure, and where the allocations are approximately 2 to 3% of an entire asset pie and they're under-allocated specifically to digital and renewables. And so as the CEO of the firm, I spend probably half my time fundraising. I'm literally about to leave this weekend for three weeks of fundraising in Asia and the Gulf and then back to Asia again. And I'll see probably 70 LP's in three weeks. And I'm going to hear the same thing that I think we've been hearing for the last nine months, which is, private equity is out of favor. Unfortunately, real estate right now is a little out of favor, but infrastructure is very much in favor because most people are under-allocated by about 150 to 300 basis points. And so the opportunity for us as a publicly traded asset managers is to continue to raise capital in those swim lanes and be incredibly surgical with our peers about where we're putting that capital to work. If you want to raise capital today, obviously infrastructure works, digital infrastructure works, but you also better come to that dialog with a very big pipeline of ideas and actionable opportunities, which is what we're doing. We talked about the 10 to 11 billion we're going to put to work next year, and I think LPs want to know that you've got active deal flow and most importantly, back to what we said earlier, you've got good customers, good tenants that are signing long term leases. And if you have that development capability and you have that pipeline, you're very relevant today. This for me, this is the toughest fundraising environment I've ever seen in 30 years. And I've been doing this a really long time and I've been raising capital for a long time. It doesn't mean it's impossible, but you better really come with great ideas and unique ideas, proprietary ideas. For us as an organization, the reason we've been able to raise as much capital we've raised this year, we've raised $15 billion in debt capital, we've raised over 7 billion of equity. And the fact that we've been able to form $23 billion of capital in this environment is because we have good ideas and we have customers that are committed long term to the things that we're doing. But it's not easy and the cost is obviously more expensive. I'd say one last thing is credit. Private credit is another area that LPs are super focused on. And I think as we look at debt maturities in real estate and in LBOs, you're going to see that there's a huge opportunity in private credit like we've never seen in our lives, much bigger than ‘08 and ‘09, and certainly bigger than 2002 and 2003, and when we had two environments where there were big course corrections. And make no mistake, we're in the middle of a course correction. So while it's hard work, if you do form the capital, you do get rewarded for it. We're seeing much higher returns, much higher rents and better yields.
Spencer Levy
Pat, what type of advice are we giving our clients today in this complex, higher cost environment of how to succeed in the digital real estate space?
Pat Lynch
Yeah, I'm going to segway to one of the bigger challenges that we've got, Spencer, which is is the answer to your question, both on the investor operator side but also on the occupier And what we see in the near term, and the near term to me is inside of 36 months, is significant demand with very little supply, mostly supply limited because of utility power. And that's a global issue that the industry is facing. So on the operator investor side, there's a, I'll call it a little bit of a land grab for locations that have immediately available power. So markets that we may not have even considered a few years back, but if you're sitting on a substation with 30 megawatts of available power, that's a market that clients like Marc’s companies would consider when before they wouldn't. So the dynamics of that are, find the locations, secure them, acquire them. One of the things in the digital infrastructure space, even with increasing prices, is that the cost of the land is oftentimes a rounding error, right? So if I pay two or three X for the land because prices are increasing and demand is there, if you can deliver power in the near term, the landlord is controlling those conversations. Very different than when we first had these conversations a few years back. So the landlord is now in power. On the tenant side, we have clients that have an immediate, urgent need for space and our advice in some cases is to disregard the RFP, the RFI. We're going to go to the landlord with a letter of intent at a price that's going to be attractive to them. So that's the dynamics that have changed just in the last 12 months. And because of this supply constraint, I don't see that changing in the near term.
Spencer Levy
So let's go back to a point that Marc made earlier on in the conversation, which is latency, which is speed, the speed that you go from point A to point B. But I think, Pat, what you're suggesting may be different, where if you're prepared to go to an alternative location, not Northern Virginia, not midtown Manhattan, not Dallas, aren't you then sacrificing some of that latency that is so important?
Pat Lynch
Your are, and what we've seen, and this has actually been happening for the last few years, and again, the digital infrastructure space is addressing this. The clients, be them banks, insurance companies, they're bifurcating their requirements. So there's things that absolutely have to be in Northern Virginia or New Jersey. It's been a market that's actually we've seen a lot of demand over the last few years. Those are low, oftentimes, low latency permits that have to be in a specific location. And I'm using U.S. markets, but I could have the same conversation about London or Singapore. And alternatively, less latency sensitive may migrate to a place that has low cost power, low cost in the data center space for the occupiers, it’s utility cost and taxes, predominantly personal property and sales taxes. So I'll take the opportunity to plug my hometown of Omaha, Nebraska. They just allocated, I think it's a $2 billion utility expansion to accommodate some large Fortune 20 companies that have a footprint there. It's a great location, center of the country, low cost. You're not going to do low latency banking in that market, but you will do large storage, potentially some AI components that are less latency sensitive. So I think that's how we're seeing it play out and how some of these secondary locations that previously were not considered may actually have an opportunity.
Spencer Levy
Well I guess it's, to put it in very basic terms, it's a difference between how quickly do I need pictures of my kid to be pulled up by my phone versus the nanosecond trade that I'm making that could make X dollars for the firm. That nanosecond trade better be as fast as possible. I love my kids, but I can wait an extra second or two to see that picture.
Marc Ganzi
You just basically condensed the concept between compute and storage, right? Active compute needs to be instantaneous and data storage needs to be secure but doesn't need to be to your phone within a millisecond. And this is really what Pat just described, is hybrid cloud, and most sophisticated CIOs today. I know for us, we have a hybrid cloud stack in terms of our infrastructure, DigitalBridge, and we're not a Fortune 500 company, but we have a lot of data. And so there's certain compute resources that we put in a very secure location that's low cost but highly secure in private cloud. And then we have other stuff that's active compute that's much closer to our operations. And the bifurcation between active compute and storage is actually a big part of the value proposition for the tenant, for the user, and giving them good advice and sharing with them that ultimately what you can provide is a hybrid IT stack. This is the essence of hybrid cloud, and when you hear the term hybrid cloud, this is exactly what Pat's talking about today.
Spencer Levy
Well, AI is all over the papers these days. Where are we today from a real estate perspective in terms of the AI usage of our real estate, how much power does it need? Where are we going?
Marc Ganzi
So we've got a pretty good handle on that today. And look, it's going to change. I like giving people a history lesson before you dive into AI, which is the case study around public cloud. So the origins of public cloud are now about 11 years in terms of from a real estate perspective. And if you go back to the early CBRE leasing reports in 2012 and 2013, we had our first cloud absorption, which happened here in Virginia in data center alley. And so we've been building public cloud for 11 years, and today we're at 13 gigawatts of leased space in the public cloud environment. And along that journey, we've spent somewhere around $3.8 trillion building the public cloud. So that's the cap X. And there's the output.
Spencer Levy
Let's back up. $3.8 trillion, Folks, that's about 10% of the size of the U.S. economy every year. I mean, that's an enormous amount of money.
Marc Ganzi
It’s a lot of money. It's a lot of money, but it shows you how much the cloud guys have invested in their own infrastructure and in third party infrastructure. Now, now let's frame the AI opportunity. And keep in mind, 13 gigawatts, which is 13,000 megawatts of data center capacity, has been absorbed in that 11 years. There's three different reports I've looked at and you can believe the bottom end of it, which is 33 gigawatts, and you can believe the McKinsey report, which I think is 36 or 38 gigawatts, kind of doesn't matter, right? It's a rounding error. It's 33,000 megawatts is the baseline to build AI, and I'm sitting there as an owner of infrastructure saying, wow, I thought the last ten years was hard and now I'm being tasked with ultimately where we're going over the next ten years, which is going to be harder with a backdrop of what Pat said, which is we are capacity constrained in terms of power. Look, the most important data center market in the world today is here in Northern Virginia. And Dominion undershot their transmission infrastructure and we will not have new power delivery in for another two and a half years. And largely because Mark Warner and Governor Youngkin got involved and that was going to be a five year hiatus from new power. But now the governor and the senator have done a great job of truncating that to three years. We've got some solutions now. So basically, this market is off the playing field. Northern California, off the playing field. Pacific Gas and Electric, PG&E has no power. Silicon Valley power, which is the primary provider of power in Santa Clara, their next upgrade to the grid is 2028. And so Pat was talking about these new markets that have emerged like Omaha, which is a great storage market. Here's the next great compute market. Reno. Why is Reno working? Reno, from a data center perspective, is working because the state of Nevada put in place these tax incentives that Pat was talking about. We have close to 5 million square feet in Reno. We've leased about a million and a half of it. We've got another three and a half million that we can lease at switch. We have a big tier five facility there next to Tesla's Gigafactory. We're less than 2 milliseconds away from Silicon Valley there, and we're building more fiber to create more of a low latency environment. But why is Reno winning? Taxes and low cost renewable power. We're 100% solar in that 5 million square feet. So not only are we providing a lower cost solution, Pat, but we have the low cost power. It's renewable, which is what the cloud guys want. They want to go green as well. And the future of data centers is ultimately going to be, can you deliver that intensity, that compute for AI? And can you do it with renewable energy? This is the next paradigm. We think this is a ten year build. We think it happens in, kind of, four different quarters. So if we're playing a commander's football game, we're probably in the first quarter and we're probably 6 minutes into the game.
Pat Lynch
The challenge in the near-term is and oftentimes that commander deficit is growing because the end users have to take the power that's available and more often than not today, it's not a fully renewable source, right? So you are digging a bigger hole and let's hope there's a big comeback in the second, third and fourth quarter.
Spencer Levy
Let me put a real estate spin on this for a second. There's all different types of renewables out there, and some of them are more controversial than others. But let's just take solar and wind.
Marc Ganzi
Yep.
Spencer Levy
I was looking at a presentation that Pat showed me just a minute ago to create one megawatt of power. It takes hundreds of wind turbines, and I can't imagine how much real estate requires for solar panels. So I'm just going to be very practical about this. How are we going to create that much renewable power to be able to power these types of facilities?
Marc Ganzi
It's a daunting challenge. The blueprint for what we did with Tesla in Reno is interesting. In the Gigafactory, I think it’s about 8 million square feet of space. And our data centers are small compared to them. But that solar farm is on close to 2000 acres of land. But it's desert land. It's mostly land that didn't have a great highest and best use. And we've turned it in the highest and best use. We're turning out enough power there that we actually have excess power. So we're selling that back to the grid, to Nevada power, which I think is the blueprint for the future. I think solar in certain geographies works great. I think wind in certain geographies works great. I've never understood why wind in the middle of the United States makes sense. I do understand why in the Irish Sea, wind makes sense, because you have constant flow to that. But we do need a fourth leg to the stool. I put hydro in there as the third leg of the stool. We've got a data center business in Brazil where we have 12 hyperscale data centers serving the cloud, 100% powered by hydro. And there what we did is we made an agreement with two hydro providers in the north of Brazil, released our own transmission infrastructure. We had to build our own substation in Campinas and we had to register as a power company in Brazil. So we are a provider of power in Brazil. So we have our own transmission agreements and we have our own distribution agreements. And ultimately, when the power comes into Campinas, we're not only selling power into our data centers, we're selling power into our competitors' data centers.
Spencer Levy
So this is the science fiction portion of today's discussion, because I've been reading a whole lot about semiconductors and their ability to transmit power from point A to point B without the loss of electricity. And that's really the question, because it'd be great if we all put 100 million acres of solar panels in the middle of nowhere and then transmitted it everywhere. The problem is you can't do it.
Marc Ganzi
You can’t do it, because of loss.
Spencer Levy
And so the challenge is transmission. How do we deal with that? Pat, any thoughts?
Pat Lynch
A lot of this is regulatory related. I'm going to take it just a different direction, Spencer. What I think is, we're going to start see happening – this is back to the comments about newer markets – is I have a great job and a great seat working with a great company. I get to see a lot of really interesting early stage projects, and they are in places oftentimes maybe the middle of nowhere. But I do believe in some cases we're going to see the clients bring the site to the power, and that could open up markets where wind and solar, you still have the problem of battery storage or something to make them a viable site. But I think transmission is a problem, whether it's what we used to build federal highways years ago, I don't know what the solution is, but we've got to make it easier. But I do think one of the near-term results of this is we're going to see people bringing the site to the power, which is a different concept.
Spencer Levy
So creating the power on site or, go to a place that has the power already.
Pat Lynch
Exactly. Maybe there's a large solar wind site that's going to create a gigawatt or more power. Some aspects, back to the latency question, for a large storage site, why couldn't it be in the Midwest or in the mountain states somewhere where that's readily available? So I think we're going to see that. And then one other topic that is of high interest to me is nuclear. We have a massive, however you want to define it, we have a utility problem for the digital infrastructure space, but for other industries just in general. And I do believe that we need to look at ways to make nuclear more accessible and to take the regulatory component down to what I'll call a more reasonable level, because I think that's our path out.
Marc Ganzi
Well, I think where we've tried to be is, again, just owning a street corner. I think one of the tactical decisions we made after merging with Colony when we were a REIT and then not becoming a REIT is that we knew the way we had to differentiate ourselves was to stake our claim on a street corner and be the expert and be the expert by a long shot. So I think the word we use is just playing at scale. And so our ability to play at scale dwarfs that of a Brookfield or a KKR or Blackstone, who all now, in their public commentary, they're saying, look, we're going to be big and digital, we've got to be big in digital. Nice thing to say, hard to execute. And so even though we're small as a $75 billion asset manager, in the digital space, where we're 2 to 3X bigger than anybody else. So I made a conscious decision to do that, but I'm also hedging, and when I say I'm hedging, we bought AMP Capital a year ago, which was the infrastructure business of AMP out of Australia, and there we now manage two different funds. We're managing about 11 billion of assets in traditional infrastructure, where our focus is middle market digital, renewable energy and digital logistics, which is transports. And so, well, what we see is some of the bigger general listings for GP's drifting into digital, we're now drifting into transports and we're drifting into renewables because we know we have to have a renewable source of energy to power our digital infrastructure. And we also know because we're so involved in transports, whether it's positive train control, whether it's doing private 5G networks at airports, the stuff we're doing at the Port of Long Beach, digital is changing transports. And you can use digital technology to change a shipping port. You can change traffic flow at gates, you can turn ultimately reservation systems quicker. You can change the security pattern in an airport. There's so many things that we're doing in transports where we're using our digital knowledge to create a comparative advantage. And no infrastructure investor has used weaponized digital to change stuff like shipping ports and airports and toll roads. And these are things that we can do that others can't. So we're taking our digital playbook and traditional infrastructure. We've made a nice investment. We changed AMP Capital into Infrabridge, and we're growing that sleeve. So, we think we can do both at the same time, and ultimately, as a publicly traded asset manager, we've got to keep creating new products. That's the game I'm in, unfortunately. So we have a credit team, we have our core infrastructure team that does digital, we have Infrabridge, and we're growing. And we're proving that we can go against the bigger guys and compete, and that's what I have to do. I have to go out and compete every day for capital.
Spencer Levy
A couple more quick questions. We've been speaking for a while. You mentioned this briefly. Switch, you just did a large acquisition, $11 billion. What did you do and tell us about it?
Marc Ganzi
Well, look, we saw something in Switch that the public markets didn't understand, which was that the fastest growing vertical in the data center space is private cloud. Everyone talks about public cloud and AI, but there was a big trend that we spotted three or four years ago, which was called Reverse Cloud. So a lot of organizations rushed into the public cloud. They got to Azure, they got to AWS, and then they realized some of those workloads weren't as secure as they thought. And they said, hmm, maybe we need to rethink our I.T stack and bring some of that back into our control. But what had happened was all the airlines and all the big banks and all the big pharma companies had sold their data centers and they completely outsourced to public cloud. So, what Switch had that was very unique was, it has the most secure environment for data in the world. There's only two tier five providers of data center infrastructure in the world. That's Equinix’s big data center out here in Virginia, and then the five big campuses that Switch owns. Switch is the only data center operator in 20 years that's never had one minute of downtime. That is impressive. So 100% uptime is what they deliver to the customer. And so big Fortune 100 users use Switch. The other thing that's interesting about Switch that we saw was they had renewable power, back to this renewable issue. They've got 11 million square feet of usable data center space. They've only leased 5 million of it and they had 1.6 gigawatts of power and they had only used 550 megawatts. So to Pat’s point about the commodity being power, we saw this 1.1 gigawatts of excess power they had that nobody else had as the key commodity. So, no shock, we had a five year business case. Within nine months, we've done three and a half years of leasing in our plan. So they've absolutely knocked it out of the park. And I think investors are really happy with what's going on at Switch.
Spencer Levy
That's great. And so let's talk about water. Water's an issue. We talked green electricity. We did not touch on water and its scarcity, and how water intensive data centers are in particular. Pat, how do we deal with that issue?
Pat Lynch
Yeah, I think part of it will be location driven, right? Hillsboro, Oregon is a place we've seen take off that maybe is a little better positioned. Back to my, this is not intended to be a big Omaha pitch but they have a large aquifer there for that. So I think you'll see that globally. Marc mentioned their projects in Brazil. There's also a trade off with air cooled in places like Arizona or Utah, where water is absolutely a scarcity and you're giving up a little bit on your OpEx expenses to take a different solution, which is more of an air cooled approach. I also think the industry is incredibly focused on increasing our sustainability, be it water utility, more efficient utilities. We're a massive user of utility power globally and I think there's a lot of things on the technology side that are being looked at that will make a difference, Spencer, down the road. I remind people, this industry, had we not consolidated into the massive scale of facilities that Marc talked about and others are doing, our sustainability utility issue would be significantly worse. So while we are using a significant amount of power, we're doing it in a sustainable way, at least as best technology today will allow. But I'm confident that a lot of really smart engineering and IT people are going to allow us to improve as we move forward.
Spencer Levy
We're just about out of time so I'm going to ask both of you for your final thoughts here, and where are we going, right? You are at the Vanguard of thinking on technology, but it's constantly changing. I think, Marc, you used the phrase that every seven years you wake up, and you're waking up in a new world. So seven years from now, what world do you think we'll be waking up in? Marc?
Marc Ganzi
Well, first of all, I think the big trend will be the realization and the monetization of generative AI and whether or not that's going to be real or not. We now have, kind of, acid reflux on AI. Everyone was sort of cuckoo for Cocoa Puffs on it 90 days ago. And now people are saying, well, what does it really mean and how do I weaponize it? How do I monetize it? And I think that'll be played out over the next 7 to 8 years. I don't think it happens immediately. The other thing that I tell people is virtualization of the network. Pat said it earlier, he's in the network business. He's in the network fulfillment business. That's the business you want to be in because when you're in the network fulfillment business, you're accepting that part of that network is virtualized. And what we mean by virtualized is software defined, and we're moving into this new world where infrastructure as a service can be delivered through software and through applications. It's a little bit of thinking in three dimensions, not two dimensions, which we've historically done, and I think A.I. helps accelerate that. But a lot of the new business models that we're looking at in infrastructure are software defined. And software defined networking will be one of the great big trends over the next decade. While everyone's talking about AI I’m like, that's nice. Yeah, that's going to happen, but it only happens through the network intelligence in a software defined environment. And I think that's something that we're spending a lot of time on. If you want to talk about what's next.
Pat Lynch
The imagination of seven years, to me, Spencer, that's probably about the time I'm going to be retiring. And I can't even imagine with our automotives, with our cars, with, you know, people talked about our phones. The devices that are going to be created in that time is so fun to think about and then to think about the digital infrastructure that is going to need to exist to support that. It's a fascinating, almost unfathomable, in a good way, for me to think about. I also think back to the nuclear thing. I believe that we're going to figure out the sustainability issue, whether it's nuclear or something else. But we've got to create a utility infrastructure and transmission to support that. And I'm hopeful in seven years we will have figured that out globally.
Spencer Levy
Wow, we could speak for hours, but we're going to end it here. And I want to thank Marc Ganzi, CEO of DigitalBridge, to talk to us today about all different forms of digital real estate. What a great conversation. Thank you, Marc.
Marc Ganzi
Thank you. I appreciate being with the two of you. And thanks, Pat. Good to be amongst friends.
Spencer Levy
And then our Pat Lynch for his third visit to The Weekly Take. Great job, Pat. Pat Lynch, Executive Managing Director, CBRE.
Pat Lynch
Thank you Spencer and team, and Marc, again, thanks. This has been fantastic. Always a lot of fun.
Spencer Levy
We've got more fun coming up on the show as we head into the home stretch of 2023. We'll continue to talk tech and real estate as well as retail, restaurants and more. If you'd like more information, please visit our Website, CBRE.com/TheWeeklyTake. And don't forget to subscribe, rate and review us wherever you listen. Your comments on Apple Podcasts or Spotify really do help people find the show. We look forward to having you back next week. For now, I'm Spencer Levy. Be smart. Be safe. Be well.
.jpg)
