Please take their seats. We're gonna go ahead and get started. Welcome everyone to our 33rd Annual Media, Internet, and Telecom Conference. For those of you who don't know me, I'm Matt Niknam, Comm Infrastructure Analyst at Deutsche Bank. We're very pleased to welcome back DigitalBridge CEO, Marc Ganzi. Marc, welcome back.
Yeah, thanks Matt. 31 years?
33.
33.
For the conference.
That's incredible.
Many of which you've been a part of.
I was at Deutsche Bank from 1999 to 2002. I remember coming to this conference and we were with some investors this morning laughing about that 1999 conference was everything.com, pets.com, this.com, that.com. We've been coming here a long time.
Marc, you've been very busy since you took the helm at DigitalBridge. Maybe we can start by recapping some of the bigger milestones in 2024 and what you're most focused on in terms of top priorities for the company in 2025 .
I think that the headlines coming out of 2024 were, you know, one, pretty significant beat on fundraising at $9 billion. We delivered a really strong fourth quarter, which gave us run rate FRE kind of in that $140-plus zip code. Still a miss to guidance in where we wanted to be in 2024, but a cadence that would suggest there's a strong opportunity to recover and come back, here in 2025. Two other headlines that I think were really important for 2024 were one, we got to almost $100 billion of assets under management. You think about that journey five years ago when we merged into the public real estate investment trust . We were less than $20 billion of digital assets under management now.
We have more than 4x our digital infrastructure assets under management, going from $20 billion to almost $100 billion in a very short period of time, about four years since the merger. We have done a lot of hard work between now and then. The fourth thing that I think was important last year is just the organic growth that is happening down at the portfolio companies. That really manifested itself strongly, with operating results in towers, fiber, and data centers. We will drill into some of that. I think what it did is it set up, you know, leasing backlogs into this year that were record levels. We have a record backlog in terms of where our data center portfolio is today. The backlog at Zayo, over $4 billion, is just incredible. I mean, that has really been a big surprise, standout story.
The early results that we saw at Vertical Bridge in January, the strong leasing in the fourth quarter. When you get all of those swim lanes producing at the same time, that's a rarity. You know, it's the opposite of triple witching. It's like, you know, triple everything going in the right direction. Usually, we have one or two of those swim lanes working pretty hard, like towers is going well and maybe fiber is down and data centers are up. I think all of the sort of DigitalBridge basic food groups across all of our 51 companies are experiencing, you know, really good organic growth. I think it sets up for a really strong 2025. You know, we've got great momentum in fundraising. We have spectacular momentum in terms of putting capital to work.
We announced that the EU this morning had approved the Yondr acquisition, which is great. We got European Commission approval to proceed with the closing on Yondr. That is our fifth big investment out of our third flagship fund, which indicates we are putting capital to work. The other thing that we said in the earnings call that is sort of a subtext is $9 billion of return capital at the same time. I know that sometimes frustrates our public investors a little bit, 'cause we do lose a little bit of FRE. Part of returning capital and generating great returns is what we do as an alternative asset manager. I think it is always kind of like you gotta take sort of one step back to go two steps forward.
I like that setup because as we returned a lot of capital in 2023 and 2024, it's given us the window to go ask for new capital in 2025. I think when you look at the fundraising schedule in terms of the five products we have in market right now and the investment ideas that we're providing LPs, they're really unique. I don't think there's another firm in the world that gives investors that kinda specific access, Matt, to the best digital infrastructure investments, you know, that we see around the planet today. I like the setup for 2025. It's really in a good spot. I also gotta thank my team. I mean, it was a hard year, but our new CFO, Tom Mayrhofer, is doing a great job.
We kinda reframed our guidance, h im, Severin, and I decided to reset the table for 2025, deliver a different kinda cadence, which is back to sort of four or five years ago. We would wanna be the guys that kinda underpromise and overdeliver. Everything is really set up strong for 2025. As I said on the earnings call, there's a strong foundation for where the stock is today to build from here. We now have to go take that case and that argument to investors that we're trading at a massive discount to our NAV.
That's a great segue into my next question. You think about all the milestones and the great setup. What do you think the street's missing or getting wrong when you look at the stock price? I guess maybe if you can be a little bit more detailed and granular around what you can do in 2025 to change that perception.
I think it first starts with what I did on the fourth quarter, which was owning the mistake that we made in the third quarter. We set the wrong guidance. That was on me. I take full responsibility for us missing our guidance in the third quarter. I think that was really a byproduct of we had the right fundraising number for the year, but it was so back-end loaded that we missed Q1, Q2, and Q3. It's actually a lot like towers. If you miss your leasing in Q1, it just compounds and you sort of get whacked by the time you get to the end of the year. If you miss Q1, you miss those commencements that happen in Q2 and Q3. Same thing happens in fundraising.
It's actually a very parallel path because our fundraising commitments, our investors are with us for 11 to 13 years in a fund, much like signing a 10-year SLA on a tower. If I have a bad first quarter and a bad second quarter, it compounds. That is what you saw was that snowball effect in terms of our miss on guidance in Q3. I think this year we are a lot more poised to do quite well. We have been very prescriptive about where we are, what we are gonna do. You know, the first quarter is finishing up our third fund, finishing up our second credit fund. That will last into Q2. We are launching our second private wealth product. Our first private wealth product was wildly successful. I never could have imagined we would raise $1.1 billion in three months.
and that, you know, we should've probably gone to private wealth a lot quicker. our peers have all gone to private wealth and I think it's a copycat league. we're just backdrafting on what Blackstone has done and what Ares and Apollo have done. I think, when we finally stepped up onto that same stage as our peers, what we found out is private wealth allocators were like, "Look, I don't have that. I don't have access to digital infrastructure on a private basis." what we found is that product flew off the shelves. we're back with our second product. Andrew Cox has done an amazing job. we were really lucky to get him from KKR. he kinda fell in our lap. he's doing an amazing job. I'm excited to get that launched.
That launch is in Q2. We should feel the impact of fundraising in Q4. The two new products that we announced on the earnings call I think people are really curious about, which is, you know, our new data center realty income fund. It'll be our first real estate product, you know, since we did the post-merger with Colony as a REIT. This'll be our first re-entry back into the real estate LP world. Remember, LPs in the real estate world allocate over $4 trillion of capital to private real estate. That's actually bigger than infrastructure if you think about it in terms of where we can go fishing. That is a much bigger pond with a lot more fish. We are competing against the likes of strip centers and apartments and office buildings, CBD office buildings. I like that competition.
I'm happy to go toe to toe with CBD office and strip malls. We think there's a lot of opportunity for us to raise the capital. We're doing it in partnership with Goldman Sachs, which we're excited about. That has a lot of runway. We're launching that product in Q3. It should feel the impact of that in Q4. What we've talked about, power, I think you and I have been talking about power for like two years. We've been developing power adjacent to our data centers with partners. Those partners typically have been publicly traded utility companies. They're getting most of the margin on our power activity. We want to flip that script.
I think we've been really clear about our ability to build adjacent to our data center power sources behind the meter and off-grid power solutions where we can be the provider of power. We can be on-grid. We're interconnected to the grid and it allows us to trade power. That's a big difference because once you have interconnection, much like fiber, a lot of these things we learn these lessons from digital infrastructure. Power's the same way. Once you're interconnected, you can go grab power from many different sources. There what we have found is we own the data center. We've taken the risk. We have ownership of the customer. We should own the economics on power. It's something we actually never did in towers, and we never did it in fiber. In data centers, you can actually move upstream.
You can control the flow of power to your data centers. Microgrids, batteries, other forms of transmission, these are things that we can invest in, that we have confidence in. We've now built microgrids to a couple of our data centers. We understand how the economics work, how energy flows. Most importantly, I really like where battery storage is going. A lot of new things happening in battery storage that are extending the life of batteries. More importantly, we can get to higher power sources. That really helps us deal with intermittence because one of the grievances on solar and wind is it has fluctuations. If we can even out those fluctuations and we can control that source of power through our own battery sources, there's a big opportunity to invest there.
We think there's a $20 billion-$50 billion AUM opportunity to invest in power over the next five years. Remember, again, we're already building the data centers. We've already got $18 billion of CapEx committed to new data center construction over 100 locations where, you know, if you aggregate our 11 portfolio companies in data centers, we're the largest developer of data centers in the world by us. Third-party data centers by a long shot.
It would not be a session with DigitalBridge without asking about AI. Maybe let's start and.
Sure.
How would you characterize, I guess, how far along we are in the AI infrastructure build-out? And maybe we can delve into how this differs across data centers, fibers, maybe early days of whether that's impacting the local edge.
I think, you know, we sat here two years ago and I said we were kind of in the first inning of a nine-inning baseball game. I think last year we were kinda bottom of the first, top of the second. I think this year, you know, spring training, we're probably bottom of the second, you know, staring at the top of the third. We're still in this phase where we're building a lot of LLM infrastructure. And some of that is self-performed by our customers and some of it's performed by us. By and large, it's like I would say 70% self-performed and 30% industry performed. We'll see where kinda the first workloads at Stargate go and who ends up performing the first kinda 3 GW for Stargate.
I think that's kinda the first tell in terms of big LLM leasing where that goes. I think as you move to inference and you move to more specific locations where you're sort of hunting with a sniper rifle, not a shotgun, there's gonna be a big heavy reliance on third-party data center operators like Vantage, like Switch. I think that's where we really see an accelerated cadence of leasing in kinda, you know, 2026, 2027, and 2028 as we move to inference. That's where the DigitalBridge portfolio's really set up to excel. The only reason I say that is 'cause we take contextually the 13 years of building public cloud. We've been along for that journey for the last 11 years.
Our leasing with the cloud operators really accelerated in 2021, 2022, and 2023, which is really where you saw the cloud begin to move out of the big AZs and into the secondary AZs. The big AZs would be Santa Clara and Northern Virginia, right? Those would be the two primary nodes. As those data center workloads proliferated to places like Reno, Goodyear, Arizona, Austin, Texas, certainly Atlanta, Quebec, Toronto, you really began to see those really what I would say where you get those great 9%-11% cash-on-cash yields where you have those 50 MW to 200 MW campuses, which is kind of the sweet spot for us. I think at Switch, it's just been a different algorithm altogether.
With data sovereignty on the rise, private cloud workloads up, Switch really just found a really unique crease in the marketplace, which was this notion of Tier 5 private cloud. You know, where a customer really wants the belt and suspenders in terms of power, space, and so our portfolio performed incredibly well last year. You know, our leasing backlog is up 22% over last year. You know, we have over 7 GW of what I call lease applications. You can call them a tower business interest in our data centers. That is up year- over- year. I think that is just going to continue to be a steady cadence for at least the next six, seven years in terms of data center capacity. I think as it relates to the ecosystem, the big winner early is fiber.
I think, you know, what we've seen at Zayo this year was a tremendous amount of new activity in terms of bookings, new routes built. I think we announced at Metro Connect how much new fiber we've laid in the last two years. People were surprised by that. We did it quietly. We've continued to turn up network. Where that network is really performing well is for hyperscalers. In data center connectivity, there's very few people that can operate like Zayo does on a nationwide basis where we can connect not just two data centers or four data centers, but given the uniqueness of our footprint and how many data centers we touch, there's not a data center in the United States that Zayo doesn't touch.
That ability to deploy success-based CapEx, deploy what I would call very, very, you know, high strand count, you know, medium-range and long-haul dark fiber routes, that's very unique. That's something that Zayo is very good at. You know, the old days of deploying 12 to 14 pairs into a hyperscale public cloud campus, today we're deploying 124-248 pairs. We're deploying with at least one or two customers. The demand has been incredible. The other thing that's interesting is our data center portfolio got bigger. We're trying to figure out how to weaponize 278 data centers that we have. No better way to weaponize that portfolio than to use Zayo. As we're building new data centers at Vantage, DataBank, and Switch, we're putting Zayo right there.
Front of line, first provider into the data center. That makes it a lot easier for someone to lease when they're essentially the preferred provider. We trust Zayo because they know they can go there first. They can get there early. The other thing that I noticed about the new workloads for the AI guys on fiber is it's not sort of two paths in and out of the data center. We're now seeing four paths. This redundancy and this higher level of redundancy that AI customers want is not only more strand count, but they want more unique paths to create that extra amount of redundancy, but also they're taking those workloads in different directions. I think it's been good. I hope the fiber business continues to be the comeback kid.
That could be a really great story for 2025. Early for near-edge inference, I think we're just not there yet. I'm always listening very carefully to what Masayoshi Son says. I think he's really smart. I think he understands mobility. I think he understands the connection between AI and mobility. You know, a couple weeks ago here in South Florida, he said the real thing about AI that he gets excited about because of his experience in owning a mobile network is he believes mobile data traffic will have a 10x increase to the device once we get to true inference, you know, on the edge. That's a big statement. If you go back to 2013 and 2014 when cloud-native applications started showing up in your wireless devices, you know this 'cause you're an analyst, look what happened to mobile data traffic.
Mobile data traffic from 2011 until 2021 went up 10x. Why was that? It was because public cloud finally got to the device. As you put the public cloud onto mobile devices, Matt, usage spikes. We see a second wave of spike coming on mobile data networks. This is why we need more spectrum. This is why I think Commissioner Carr's got it right. We gotta get more spectrum out. T-Mobile's in a great position. AT&T and Verizon need more spectrum. There are only three ways to deal with this conundrum of whether you believe Masa's number at 10x, I'm more at like 3-5x. I think there's a 3-5x increase in mobile data traffic. By the way, I can be right and Masa can be right. We can both be right.
It's like a huge win for towers. The look-through is that the last time we've had these step functions in mobile data usage to cell phones goes back to really the introduction of LTE, right? Go back to 2009, 2010. Look at what happened to tower leasing at SBA, American Tower, and Crown. Then look at what happened when we started to put, you know, cloud-native applications on mobile devices in 2014, 2015, and 2016. Look what happened to tower leasing. The same thing is gonna happen again here. Ultimately, as we build 5G, 5G's gonna take 10 years. It plays out in three phases just like LTE did. You've got the overlays, which we're done. I think we can safely say that pretty much we've overlaid all the 5G networks against LTE.
You go into what's called infill. Most of you know this in the room that when you turn up a network like an application , like when we turned up 3G and we turned up 4G, ultimately those coverage rings shrink a little bit because of the spectral efficiency in the new radios and the usage case. As you put more use into a cell site, you start to split the cell sites. That creates holes, and that's infill. 5G right now is in infill, and I think we're in infill for the next two to three years. We move to densification. I was really clear at Metro Connect last week. I think densification is 2026 to 2029. We've got about a million small cells today. We're going to two million by 2032.
Remember, not all those small cells are just for 5G. You have to think bigger. You have to start looking at connected devices and IoT, which in the world of generative AI, we go from 29 billion connected devices to almost 60 billion connected devices by 2033. The real winner in terms of where the mobile data traffic activity is gonna be is in machine-to-machine learning and device-to-device connectivity. There is a whole phase of investment coming in 5G for near-edge and mobile edge, which is why I think you saw some of the early green shoots at Vertical Bridge in January when they had that massive—we had our best leasing month in the history of the company. That never happens. My experience is in 31 years of tower leasing, my best month has never been in January.
It's always been a November, December where all the CapEx is back unloaded or it's pre-summertime where carriers are really deploying a lot of CapEx in Q2. This was a real surprise. We called it the January surprise for towers. I've been saying it. I said it at Citibank. I said it at Raymond James, I'll say here. Don't sleep on towers. I think towers is one of the real opportunities that investors should be looking at right now as we set up into the next phase of generative AI.
That's it. There's a lot to unpack. While we're on towers.
Sorry. I went quick.
No, no, no. As I think about towers, is that one or two carriers? Is that broad-based? And is it, if it is broad-based, is it all sort of this move from coverage to more infill? If we could just get a little bit more color into why you're seeing the.
Yes. It is infill, 100%. These are new search rings. These are rings that we did not foresee coming. By the way, it is all three of our customers. It was not just one customer in particular. It was actually a steady diet from AT&T, Verizon, and T-Mobile. I think also, you know, at DigitalBridge, we have got Vertical Bridge. It is a great company. It is now the number three tower operator. We just integrated the Verizon portfolio. We also have build-to-suit programs in place with all three carriers now.
As the program with Tillman wound down, we stepped in, created a national program with AT&T that we are excited about. We already have a joint venture with Verizon, and we have had a long-standing relationship with T-Mobile where we have been the national BTS provider. We're in a really good spot because we're getting not only de novo co-locations, but we're also getting a ton of backlog into new builds. That combination was why January ripped. Unexpected, but it was, happy to have had it.
We touched on AI. I do not wanna jump off AI just yet. One obviously question in the marketplace is, are we, and you mentioned, you referred to us being sort of in the bottom of the second, top of the third. How do you think about or how do you answer the question around, are we headed for digestion? Has maybe DeepSeek concerns around Microsoft lease cancellations maybe changed the calculus at some of the larger hyperscalers around AI and what they're willing to spend at least in the near term?
I think this has been the topic on the circuit for the last couple of weeks. My observations are really there's four things I think you need to think about. One, we did not see DeepSeek as negative. We saw DeepSeek as a logical progression in the fact that costs are coming down in building generative AI applications. Remember, we opened that door. Meta's open source really was the foundation for DeepSeek. That was relatively easily free and gave them a very big head start. To get there, it implied that Meta had to make a huge investment in their open source LLM, which they have been investing in for a long time. A subtext winner of that was Meta to a certain degree. It is also something that we have seen before. This is not like a new movie.
You know, ultimately, as we've from the advent of the PC to the internet to mobile phones to cloud, innovation has happened all along those sort of four curves. The cost to deploy that infrastructure and the cost to deliver that service comes down over time. What we've seen in AI is that those costs, that cost curve has come down about 40x in a period of 36 months. The cost to build AI and to build these sort of applications like a DeepSeek is gonna continue to get cheaper. That's following a logical progression that we've seen in other subsectors. We weren't terribly alarmed by DeepSeek. We actually think it's part of where this goes over time. As costs come down, something else happens at the same time. Adaptation moves rapidly faster.
If you look at the J curve of adaptation to the internet, to mobile devices, and public cloud, you'll see it's pretty simple. Like internet sloped like this, mobile data sloped like this, cloud sloped like this, AI is sloping like this. It's the fastest adaptation we've ever seen in a technology shift. This is kind of the sort of what people are talking about when they say Jevons paradox. You have a faster adaptation, and you have more data being consumed and stored at the same time. Remember, data has two forms. You've got storage, and you've got compute, active compute, and storage. What we've seen is, as adaptation has accelerated and costs have come down, investment has accelerated. We saw that on display in the last two weeks.
We've had four of our customers come out and say, "You're increasing CapEx, not decreasing CapEx." You heard announcements from Meta, announcements from Apple. Alibaba had an announcement. We've seen also Oracle revise its CapEx while Microsoft reaffirmed its CapEx. The important thing was Microsoft confirming their $80 billion of CapEx. When Microsoft came out and said, "Okay. We're gonna we're gonna move some workloads around, and we're gonna shift some things," that was somewhat expected because of the relationship that they have with OpenAI, which to be direct in Sam's defense, he's shifting his providers. He's not gonna go all in on on Microsoft. Microsoft's one of his biggest shareholders. They have a long-term agreement to develop infrastructure with Microsoft. Also, OpenAI has the opportunity to go build their own infrastructure, which again, to Sam's defense, he's doing that.
He's building a team. They're out looking at locations that are gonna be OpenAI-only locations. That's a logical progression in Sam's arc in terms of where OpenAI is gonna go. It's going to be an independent for-profit company. Part of that is building the muscle memory to build your own data centers. That customer is an important customer to us. I look at data center demand as a series of buckets, right? Buckets of water. You gotta look at one bucket as how many leases we signed as an industry in 2024. That bucket filled up to six, effectively 7 GW was the amount of leasing we did last year. There's this other bucket, which is how much data center capacity did we turn on? Did we light up as an industry? We lit about 6 GW .
You have both of these buckets kind of at six, but the leasing bucket was 1 GW heavier. As it spills into the bucket of construction, that bucket could not hold it. We operate, we start in 2025 with a 1 GW deficit where the supply cannot match the demand. You have the forecast this year where the industry is gonna lease 8 GW. We are on a cadence to lease 8 GW. We are not on a cadence to deliver 8 GW of data center hall capacity. Part of a lease is you sign a lease, and you have the conditions precedent to deliver to the customer. That is, you gotta deliver the data hall. That data hall has to be complete with power, space, cooling, and connectivity. If you do not deliver those four elements, guess what?
Microsoft doesn't have to move into the data hall. At some point, you have penalties. If you're late, Microsoft can even potentially say, "I don't wanna be at this data center." You'd have to be really late in that instance. By the way, it's the same in the tower business. We all have to deliver a tower at a certain point, right? You have an NTP, notice to proceed. You gotta deliver the tower, and it's gotta be ready for the radios and antennas and coax or fiber connection. The industry in my 31 years of doing this, it's the same thing. It's about delivery. You gotta deliver. Our industry right now is operating at a deficit. It's going to continue to operate at a deficit in terms of what we can deliver for a while because we have challenges.
I like just being super transparent with investors about what the challenges are in data center land. Data centers are great. You know, it's a great business. By the way, it's still not as good as towers. Data centers are pretty good. And fiber's pretty good too. The challenge with data centers is it's a series of challenges that digital infrastructure's never confronted before, which is power. Putting a ton of power into a small piece of real hard. Power density is difficult on 10 acres, 12 acres. Think about a Vantage Data Centers facility in Santa Clara. You've got six floors, and you got 48 MW stacked in six floors. You gotta bring all that power into essentially like four acres. That's really hard. The industry has some challenges. We've already discussed what those challenges are. It's power, and it's supply chain.
To build a data center requires a series of specialized components, whether it's backup gensets, UPS units, production crews, backup batteries, all of these things you need to deliver a data center on time. You need fiber. You need a ton of connectivity. That is where we have a big advantage. Having Zayo as one of our investments allows us to take the connectivity piece off the playing field. We're able to deliver connectivity. Both of our big portfolio companies, Switch and Vantage, made deals in 2022 with Siemens. We had a ton of backup gensets that we planned in the middle of COVID to go take a lot of that backup capacity. By and large, we've had very few incidents where we haven't delivered. We got a 10-year head start. We started working on this 10 years ago.
I think to the newly indoctrinated or the tourists that are in the industry, they're gonna have challenges delivering data center capacity. For us that have been around a long time, I think we've been planning on this for a long time. It does have challenges. I think as long as that supply and trade imbalance exists, you know, blips in the road like DeepSeek are not negatives. They're positive. If you just listen to not the noise, you just listen to the CapEx, you listen to what's actually really happening, stuff like this is a positive catalyst for the industry. It's not a negative.
Foresee anything exceeding these constraints? Because it seems like a pretty, you know, seven gigs leased, six lit, eight gigs leased this year, another six it seems like gonna be lit. What maybe eases some of that friction? Is there anything in the new administration that can help there?
I think, look, the administration has been very constructive on both power and AI. I applaud the current regime for doing that because I think, as the president has always said, "We wanna stay competitive. We wanna be the first in AI." I agree with him. We do need to be first in AI. I think as much as the executive branch can be helpful, our challenge really sits at the states. It always has sat at the states, which is as fast as any executive order can take you on data center construction, you're still regulated by the local PUCs, which control the grids in each of those states. Our ability to light up power is a state issue, not a federal issue. That being said, there are things that we can do on a federal basis. Permitting, too. Will-serve letters.
Inevitably, at the end of the day, you are beholden to the state PUC. You are beholden to that regulated utility company actually turning on the power. A will-serve only says, "I will serve that piece of real estate with that amount of power." It does not say when I am actually gonna turn on the power. There are two differences between having an active will-serve letter and having an actual commencement date of when your power is coming to that site.
Again, to the untrained, you know, who think, "Oh God, I have a piece of land and a will-serve letter. I am in the data center business." You are not. You really have to be in a lot of different things. You gotta be in the turning on the power business, building the data hall, getting the fiber there, and dealing with all of the supply chain issues, and the special components. This is not a place for tourists. It's a complicated industry.
I wanna bring back a lot of the success that you've had with the data centers back to the DigitalBridge story. There was an interesting discussion you had on the earnings call around MW and GW in your portfolio companies at least and how that comes back to the DigitalBridge shareholder. Maybe if we can revisit that just to sort of frame how the success that the portfolio companies have had translates to success for the common sort of DigitalBridge investors.
I think that's what makes our story quite unique. I think I started the conversation by saying, you know, "Where's this disconnect in our share price?" You know, we're sitting here today at $10-$12 a share. Some of the parts analysis would indicate, you know, we're worth somewhere between $18 and $23 a share. We've gotta take that argument to shareholders because I think just on a pure FRE basis, if we trade at, you know, 18-22 times FRE, the stock's gonna trade kinda where it is. That's not particularly exciting. What's really unique about DigitalBridge is we have a big balance sheet. We have really unique assets sitting on that balance sheet, which is DataBank and Vantage. Those are really good assets.
Those assets continue to appreciate every quarter, particularly DataBank, which is growing at 20+% CAGR growth right now. DataBank is the only real edge compute business at scale across the United States in over 70 markets and, the second largest interconnection provider behind Equinix. It is a really unique growth story. I think that's gonna continue to push our balance sheet up along with our GP stakes. Our balance sheet controls basically that Fulcrum capital that initially goes into all of our funds. While our assets under management increase, so does the value of the GP increases. So does the value of Vantage and DataBank increase. We've got a balance sheet that's worth $1.5 billion today. You know, my estimation is that balance sheet grows to $1.6 + billion this year. We've got $190+ million of cash.
We're in a very good cash position. Last but not least is the carried interest. I think what we tried to do there was bridge the gap between as we light data center capacity, what does that mean for the carry? I think we've, you know, getting close to $100 billion of assets under management, over $40 billion of equity at work. Investors, we've gotta help investors do the math on carried interest. We're sitting on an arsenal of carried interest. Whether you take the low end of our guide or you take the high end of our guide, there's hundreds of millions of dollars of embedded shareholder value sitting there.
Until, as I said on the call, we start turning that out, we start creating distributions, as fund one matures and fund two starts moving into maturity, and we start returning capital and we start returning profits, the stock is going to accelerate because people are gonna see that that's not episodic. They're gonna see that we have 51 companies as we start exiting investments. That creates a lot of profit, and it creates a lot of EPS for our shareholders. We gotta bridge that gap. We're gonna keep talking about it. We're gonna talk about the power of the balance sheet. We're gonna talk about carried interest. As we light gigawatts and we turn customers into data center halls, we gotta start converting that into carry, which we will.
One other sort of topic of conversation that comes up a lot with DigitalBridge is the competition. There's a lot of ulterior who are maybe dipping their toe or doing more with digital infrastructure. How do you maintain your differentiation to stand out in this environment?
I think we have a couple of really key differentiating points. One, we have an ecosystem of companies that is a force multiplier for our existing company. Just talked about the power of Zayo working with Switch, Vantage, and DataBank and how when we build a new data center, we're pushing our own fiber straight into that data center. The power of the portfolio and that flywheel creates a ton of opportunity. Other GPs just do not have that. You know, the likes of Blackstone, KKR, and Brookfield maybe own two or three digital infrastructure investors who are 51. We just have a little bit more connectivity, and we have a little more eyes in each. Second thing that differentiates us is we're developers. The DNA of the firm is a series of entrepreneurs that have worked together since 1994.
That 30+ years of operating experience is not what you see at a generalist GP. We can talk this business at a level that other GPs can't. When you're sitting with an LP and you start going deep on our pipeline, how we build stuff, how we create opportunity, they see that difference. We wake up every day. We've got 350 people that focus just on this. Usually at a Blackstone or a KKR or GIP, they have four or five people working on digital. We just have a lot more eyes and ears in terms of what we can do. I think the real difference for LPs today is track record.
The number one thing you hear today, if you sit with any sovereign wealth fund or you sit with any pension system, they say, "Have you returned capital to me?" This notion of that $9 billion of DPI that we've created in the last 24-36 months, investors have taken notice of that. As our fundraising has accelerated and we're clearly punching above our weight class, $9 billion against $96 billion of assets under management, you look at what Antin put out last week. Antin is a similar-sized GP to us. They raised EUR 1.8 billion last year. EQT raised, like $11 billion last year, and they're four times bigger than us. We're clearly punching above our weight class in terms of capital formation, but that's because we've returned capital.
The last thing I would say about the fundraising environment that's very interesting, and this is kind of a trend in the last, you know, sort of I call 12 to 18 months. Most big LPs, most big allocators around the world, whether you're talking to a sovereign wealth fund or you're talking to a GIC and ADIA or a Korean pension system, they'll tell you, "Look, where we wanna be in the next five years is we want less GP." So they're actually paring back the number of GPs they work with. The common theme comes back. They say two things to me. One, we wanna be with the best of the best. We wanna be with the big the big five. Someone joked to me once, "Nobody ever got fired by putting money in a Blackstone fund." Probably true.
You're not gonna get fired by hiring Blackstone to do credit and private equity. The other trend that we see as they go to the big guys is they want the industry specialist. The specialization is key. They're pairing a generalist strategy with a specialist strategy. The three areas that they talk the most about are power, digital, and private credit. Those are the three hottest topics right now in fundraising. We look at it and say, "Great. We're the specialist. We've got almost $100 billion of assets under management. The nearest specialist to us was IPI at $10 billion, and they just got bought by Blue Owl." There really isn't a comp to DigitalBridge. When I'm out fundraising, I'm actually not competing for Blackstone and KKR and Brookfield dollars.
I was competing against other specialists which really haven't manifested themselves. So we do operate in the space where we really have a differentiated approach. And we have a brand. I mean, there's not, you know, right now today, 18 of the top 20 LPs in the world allocate with DigitalBridge. So my focus is how do I grab the next two? I gotta get to 20 out of 20. And, so our brand is well recognized. I think in our sector, in infrastructure and in alternatives, when it comes to digital investing, LPs are allocating with us. That's the choice they've made. Now we gotta go out and perform. We gotta keep returning capital. We've gotta keep putting up the good returns. Our third fund is performing really well, performing better than our first and second fund so far.
As long as we keep putting up returns and we keep putting money back into LPs' pockets, we're gonna continue to raise. I always said the cadence is you gotta kinda take one step back to go two steps forward. As you know, if I return $9 billion in three years, my goal is to go raise $18 billion in new fresh equity. That's what the algorithm has worked out for us. Worked out really well.
Maybe just to tie it all together with the algorithm, how should investors think about the progression of fee revenue and FRE growth at DigitalBridge? Both in I mean, you've laid out 2025 targets, but maybe if you could just leave us with what you're looking at in terms of multi-year growth profile through 2028 .
I think I've said it pretty clearly. I've sort of put out a marker that I think within three to five years we should double our assets under management. I think the initiatives that we have going in power and real estate and our flagship fund really affords a steady cadence of going from $100 billion of assets to $200 billion in the next five years. We've generally been growing assets under management by $20 billion each year. That's kinda been the step function we've been on for the last three years. With these new products coming into sharp focus, just the sheer volume of power that's needed to power the AI economy, that intersection of pushing digital and power together is a place that we sit in a good spot.
We now have experienced a few of these case studies where we've been able to build it off-grid or behind the meter. I think that's a really big opportunity. I think also, as we've talked about before, that cleanup trade where there's $120 billion of stranded hyperscale data centers that other GPs own, we view our partnership with Goldman Sachs as the chance to go do a cleanup trade, go buy the best investment-grade stabilized data center. That's a huge opportunity that no one's really fishing at right now. We think we've got two new swim lanes that allow us to go really fast. The flagship product is working. Our core fund is working. Our liquid strategies are working. Credit is going incredibly well.
Our credit strategy has been by far our top-performing fund, believe it or not, is Credit Fund I in terms of IRRs and plus that coupon that investors get. I said earlier at another investor conference last week, you know, we're at about $3 billion of assets under management in credit. I'd like to double that this year. I think there's a really good path for us to double our AUM in credit. I think we can go raise more capital in credit.
I think it's a great place to end it. Marc, thank you.
Thank you.
Great. Appreciate it.
Good to see you. Thanks, Matt.