Cool Vector Video-Podcast TRANSCRIPT: "Building Scale in Wireless Towers, Fiber and Data Centers"

Cool Vector Video-Podcast: Charting the rise of data centers and the digital infrastructure asset class.

Episode Title: "Building Scale in Wireless Towers, Fiber and Data Centers"

Episode Duration: 34:26

Originally Posted: October 29, 2024

Listen to the Full Episode here.

Speakers: Omar Jaffrey, Palistar Capital⁠ MODERATOR: David Snow, Cool Vector

Episode Overview:

A conversation with Omar Jaffrey, founder of digital infrastructure private capital firm ⁠Palistar Capital,⁠ which is currently investing from a $1.9 billion fund.

Jaffrey gives a grand tour through the digital infrastructure investment opportunity, in which he says builders of scale will have advantages in winning customers. He cautions that some investors confuse the full stack of technology, services and human capital involved in the telecommunications industry with core infrastructure, the later of which has a history of more consistent performance.

Among other topics, Jaffrey also makes the case for the long-term durability of infrastructure hard assets, comments on the pressures faced by potential sellers of wireless rooftop and tower assets, and explains the many options for revenue streams that owners of many towers can realize.

---

TRANSCRIPT

David Snow: Hello and welcome to Cool Vector. I'm David Snow, your host. And today we're joined by Omar Jaffrey of Palistar Capital.

Omar Jaffrey: Thank you, David. Thank you for having me. Really excited to be here.

David Snow: So Omar, your firm, Palistar Capital, is a specialist in digital infrastructure.

Omar Jaffrey: So I think the data consumption is still growing at 30 percent a year, and it's growing globally at 30 percent a year. That pull-through is in wireless, that pull-through is in broadband, and that pull-through is in data centers, because you have to manipulate, use, curate that data and bring it back. And there are these leapfrogging elements that make us use more data.

So doing these video chats on your phone, now with AI, we're doing all this stuff, Apple and Samsung are building AI into the phone. Well, that actually has to go and get things. So data consumption and usage, I think, is the biggest tell on where digital infrastructure is going.

And that is at least another decade-long opportunity set, especially with the AI evolution, cloud evolution that we're experiencing live, 5G evolution, and we're now scratching at 6G, just so we can do more things as we move around, and we work and school and consume entertainment in different locations.

David Snow: A couple of years ago, you raised a nearly 2 billion fund. You invested in all kinds of digital infrastructure, including fiber optic networks, towers, and data centers. And we're going to learn about all that. But first of all, why don't we just start out with a bit of a thousand-foot view on what your investment activities look like now?

We all know that the world needs a lot more digital infrastructure. Digital infrastructure is going to require a ton of capital. You have capital. What's it like out there for you sourcing opportunities? Are you seeing the kinds of opportunities that you expected?

Omar Jaffrey: So I've been in this ecosystem for almost 40 years going back to the late 1980s. And I would say presently is probably the biggest delta in demand for these kinds of assets. And on the other side, supply and financing for these kinds of assets.

Data consumption and the overlay of AI has literally stripped away any particular corporate's ability to fund. The trends in the three main areas of wireless, broadband, and data centers are just tremendous.

And as an investor in it, we are looking for opportunities to partner. So we partner with corporates, we partner with other investors, GPs, and then we create on our own.

But the need is in the hundreds and hundreds of billions of dollars in every one of those verticals. There's just not enough capital, so enormous opportunity to partner. The capital requirements are outstripping the availability of capital of corporates, and that's where financial partners like myself are very valuable actors in this ecosystem.

David Snow: We know that the rise of digital infrastructure is seen as a huge opportunity to connect institutional capital with these digital infrastructure projects. And so there's a demand, or at least, as you would know best, a huge interest in deploying capital into these kinds of assets. In 2022, Palistar raised a 1.9 billion dollar digital infrastructure fund. As you were connecting with investors for that fund, what kinds of questions did they have for you? How would you characterize their interest and their understanding of this asset class?

Omar Jaffrey: We actually were in a really interesting time. We started raising our fund probably six months before COVID.

Prior to COVID, the questions I used to get asked were: why should I invest in digital infrastructure? I've got all these great airports and ports and toll roads around the world. They are concessions by government, very stable, lots of opportunities. This feels very techie. Why should I do this?

And I think what the investor community realized is everything in infrastructure was correlated except digital infrastructure.

We saw ports shut down, we saw nobody using roads, we saw airports basically go vacant. Vessels and transport. So what you normally thought of as a very stable sector of investments all choked up. The output of that was investors reflected on actually how correlated general infrastructure was to GDP and local economy, and how uncorrelated and growth-oriented digital offered.

So following 2021, generally speaking, investors wanted exposure to two incremental things. One was energy conversion, and the other one was digital infra. So we benefited from that.

It highlighted that there are parties around the world that just don't have connectivity and they're being disenfranchised. So it actually became an opportunity to even check the ESG box, which is if you invest in digital infrastructure, you're connecting the unconnected.

So I would say there was a fundamental shift due to COVID of understanding this ecosystem and how valuable it was, and it's never stopped since.

The question stopped being "why digital infrastructure" and became "how are you investing in it and how are you going to deliver for me better returns than the alternative?"

David Snow: It's interesting you say that pre-COVID people were a little hesitant because of the technology profile of digital infrastructure. But is there any trace element of validity to that?

I mean, you're dealing in assets that are technology-based. And I think people are probably worried about obsolescence. Maybe they invest in something, and then 10 years later, people aren't even using that kind of technology.

How do you get your arms around the long-term viability of these investments and the predictable revenue?

Omar Jaffrey: 100%. What makes digital infrastructure a little more complicated than buying a toll road, per se?

This is partly why generalist infrastructure funds struggle as they sort of lean into this digital infrastructure ecosystem.

This is where I've lived for 40 years. I've seen all the cycles. We understand it. If I took apart wireless as an example, is Verizon an infrastructure company?

I would submit that Verizon is not an infrastructure company. It uses infrastructure assets. What assets does it use? It uses wireless towers. It uses spectrum. So those are actually really powerful, hard assets that underpin how wireless companies provide connectivity.

If you can strip out that layer and invest in it—uber durable. It is the highest priority cash flow that is paid by the carriers. Longevity of the assets are 90 years with almost no drama, no churn, super long cash flow, escalators, inflation hedge. That to me—the wireless infrastructure part, which is the hard assets that enable wireless connectivity—are what's the infrastructure part and that's super stable.

But people didn't fully appreciate that. Is Comcast an infrastructure company, or is the broadband that it's laying down the infrastructure? I would submit it's the broadband that it's laying down as the infrastructure. And this is where some investors make the mistake of buying the entire vertical stack, where then you're buying the human capital, you're buying innovation, you're buying IP, you're buying customers, you're buying churn, versus the infra layer. And same thing in data centers. You can go all the way up into the AI element of the software stack, or you can provide the sort of picks and shovels.

So in digital infrastructure, it's a little more complicated because you could accidentally step into the actual service side of it versus the picks and shovels.

And I think that's where people tend to get confused about it. We focus very hard on the infrastructure picks and shovel layers, and it's just massive. And if you do that, then you're inoculated from a lot of the technology layers and other elements and evolutions, and you benefit from that dynamic.

David Snow: Can you give an example of perhaps someone making an investment in a digital infrastructure asset and getting more sort of operational risk than they bargained for? Is it like you invest in a data center expecting more of an infrastructure experience, and then you confront, as you put it, kind of the human capital complexities of running a data center?

Omar Jaffrey: 100%. There are companies out there, and they're actually a little more challenged. They would claim they're data center companies. If you look at their website, they'll say they have 40 data centers. But when you really double-click on it, where they make their money is providing services.

They provide connectivity, they provide security, they provide compute services. And they don't actually physically own the data center. They lease the data center. But who owns a data center? The guy who owns a data center, in my opinion, is the data center entity.

They're making long-term leases, stable 10–15-year contracts, and so that particular company chose to monetize their hard assets and lease them. So that's now purely a services company, not a hard assets company. There are companies like that. And their margins shrink because the margin on services is lower than the margin on hard assets.

And that's probably a good example of if you don't understand the layers of a data center, you can end up investing in the services element when you were really looking to invest in hard assets. And the same exactly is true in broadband.

You can end up investing in the services layer of broadband versus the hard assets element of broadband.

So if my asset is irrelevant and the CEO's execution strategy is what's critical, that's a very different business. JFK Airport is an example. It works because it's JFK Airport and there's no airport right next to it. And so it's hard to muck it up, because the airport is the asset, and the concession is the asset.

If you apply that theme to other areas, you can very quickly figure out where the asset is not as relevant as the management team is. That, to me, is more private equity and growth equity mindset investing than where the asset is critical.

So what we do is we invest in the assets. We invest in the assets of the data center, invest in the assets in broadband, invest in the assets in wireless. And management teams are really important, but the assets are the critical function.

David Snow: I want to ask you about a couple of themes that I know your firm prioritizes. Decentralization and densification. Can you give us a succinct, brief overview of what that means?

Omar Jaffrey: So decentralization, I think if you look at your wireless business as an example, it exists on 200,000 unique locations in the United States—all over the country, incredibly well decentralized. They're owned by various individuals out there.

Data centers were much more consolidated. They were much more consolidated in Virginia, as an example, when we started, because that's where the fiber was, that's where the compute was. It's like building a big shopping mall— all the other retail stores want to come to the same shopping mall.

And that's been getting decentralized. It's been moving to other cities, major metropolitans around the world, secondary cities. So that's been happening in data centers. It has happened in wireless and it has happened in broadband. Densification is critical in the sense that you need more in the same place because customers are using more.

If you look at New York City as an example, you could have a wireless hotspot over here and a wireless hotspot over there, but both of these are congested and customers are getting a poor experience. You're going to build one in the middle.

David Snow: So are you focused mainly on building new assets, or do you also acquire? Let's use towers as an example.

We try to find the best risk-adjusted return, David. We created our own internal platform that we call Symphony Wireless.

We built this from scratch. There's 60 people there. We know where all these 200,000 locations are, and many of them are owned by individuals. So part of our opportunity is to go knock on that door and see if we can purchase that from the individual and get some scale.

David Snow: What's the benefit of getting scale?

The benefit of getting scale is then we can offer those to our customers like AT&T, Verizon, T-Mobile, and it makes it easier for them than to go deal with thousands and thousands of individual locations. We can do that at a pretty meaningful arbitrage and strong returns. So we're the largest aggregator of individual wireless assets in the United States today. There's a huge opportunity—$50 billion plus—to go get one at a time, and we're doing that. So that's one strategy: aggregating, operationalizing, and scaling up.

We're also the largest builder of new 5G wireless locations in the United States. We've positioned ourselves as a partner of choice.

We are working on a build of a fiber platform in partnership with an investment-grade telecom operator that wants to upgrade their homes to fiber. In that case, we would be building it jointly with them, for them, for their use, and getting a good offtake. We also provide structured solutions, so if we don't think owning an asset and controlling it is the right answer, we're happy to work with partners who own the equity to help them fully fund their business.

We are a partnership-oriented investor. There has to be some connectivity with the counterparty where they choose to work with us. That's basically in our DNA.

David Snow: When it comes to the individual owners of towers in the U.S., what are the dynamics around convincing them to part with their asset? Are they getting a lot of people trying to convince them to sell? Are they surprised when you show up and knock on their door?

Omar Jaffrety: In New York City, as an example, you won't see any towers. The wireless infrastructure sits on buildings.

So we knock on those doors. A lot of folks like their cash flow and they don't want to monetize. Others have needs for capital. Last year, as an example, our team made 300,000 phone calls—just the sheer volume of knocking on doors.

And we close a couple hundred million dollars of deals a year.

David Snow: U.S. office buildings—many of them are under a lot of pressure now. Is that increasing their appetite to part with assets like rooftop digital infrastructure?

Omar Jaffrety: Yes, absolutely. I think the motivation to sell has to be triggered by something. Either they have capital investment needs in the building, or they're just trying to pay down some of their debt.

It has to be triggered by some capital need, especially when interest rates are higher.

Imagine you have floating rate debt—your cost of debt is higher, your cash flows are more squeezed. Also, life events— a lot of these buildings are owned by families and smaller businesses.

Ideally, we've had conversations with them multiple times before then, and they know us, versus just-in-time M&A. Most of these things have a lot of years of work going into them.

David Snow: And with regard to building new towers in partnership with another group where they say, "Hey, could you build a tower in this place?" What would make you say, no, this is not the right kind of a place for a tower?

Omar Jaffrety: Part of the challenge is zoning and permitting. This is a real estate business at the end of the day. So do we think we can get the zoning rights and permitting to build in that location?

There's a much better understanding today of how important wireless infrastructure is, but there's definitely a NIMBY element—Not In My Backyard.

How difficult is it? Some jurisdictions, believe it or not, are still challenging on zoning and permitting.

So that’s the first cut. I think the second cut is, for that topography, for that area, do we have easy access? Is this on top of a mountain? So I’ve got to build a two-mile road to get up there. That’s a problem.

Third would be environmental issues. So if we can check those three boxes—zoning and permitting, logistical and topographical access, and no environmental issues—then it's a good place to build, especially if the need is coming from a strategic player who says this is a priority for me.

If a carrier is saying, “Hey, I’ve got a problem here,” that means others must also have a problem. So for us, that’s a good filter to say if we build it here for customer A, over time we believe customer B will join and customer C will join, and we’ll see improvements in our cash flow.

David Snow: We'll have a question about mobile backhaul. What is mobile backhaul and what do the trends within mobile backhaul indicate about the opportunity for digital infrastructure?

Omar Jaffrety: Mobile backhaul generally refers to your phone connecting via spectrum to a location, and then at that point dropping it to the ground and using some version of physical connectivity.

Years and years ago, you had an antenna facing your phone, and then a microwave dish facing the other direction to connect to the main hub.

More and more, that backhaul is becoming fiberized. It used to be copper cables, it used to be cable, and more and more of it is now fiber to get multi-gigabits, 10-gigabit type higher speeds. Because the more you're putting spectrum on, the more customers you're serving, the more backhaul you need for connectivity and lower latency.

And so a lot of investments are going on now in making that backhaul infrastructure more resilient, redundant, broader, and lower latency.

David Snow: Omar, you mentioned NIMBY—Not In My Backyard. Is it safe to say that municipal or regulatory resistance to building is more pronounced on the data center side than on the wireless tower side?

Omar Jaffrety: I think it's evolving, honestly, David.

I think today there's a very large recognition even on the wireless side that wireless connectivity, broadband connectivity—it's a social good. The federal government spent almost $20 billion on something called FirstNet to harden the network, because when hurricanes hit, most of the infrastructure is gone.

You’ve got to get it back up really quickly, and the best way is wirelessly. So there's a lot of recognition at the municipality level that these types of things are actually a social good, are important for economic growth and welfare.

I think on the data center side, it's not so much that zoning and permitting is challenging, it's availability of power, availability of real estate, and availability of broadband—in scale—that is limited.

Today the data centers are getting bigger and bigger for AI. Where are you going to locate them? How are you going to get them powered up? How do you ration power? How do you forecast power ten years out in which jurisdiction? Those are becoming really critical issues.

The volume of what you need for zoning and permitting is quite different in data centers versus wireless.

David Snow: A question for you about deployment of capital. One risk within institutional investment is that you can allocate, let's call it a billion dollars to a strategy, but if the managers in charge of putting that capital to work are unable to deploy it in the agreed-upon timeframe, then it just doesn't perform as expected.

Is that a risk within digital infrastructure—that you raise some capital and you want to put it in data centers and fiber and wireless towers, but for all kinds of reasons—topographical, regulatory—you’re unable to put it to work in the timeframe that you expected?

Omar Jaffrety: I've not seen that problem, honestly, David.

We fully invested our first fund—we raised $700 million in 18 months. We had originally four or five years to do it. We did it in 18 months.

We raised our second fund in the middle of COVID. We ultimately raised a lot more than we had originally anticipated. But we've now almost fully deployed it in less than three years.

Normally we’ve got about five years to invest. And we have looked at a ton of stuff and said no. We right now have probably a very actionable $10 billion pipeline that we think can generate the best risk-adjusted returns versus market beta.

If we were much, much, much bigger than what we are… We’re fully conscious we can deploy well inside three years, and not just write checks but asset-manage and have a good exit strategy in the back end and deliver strong returns.

A $20 billion scale—if you had a fund like that—I think then you'd feel a little more pressure because you've got to find these unique, interesting opportunities. We purposely have smaller funds so that we can invest wisely and quickly and have it deployed and start trying to drive the value.

David Snow: Digital infrastructure as an asset class has rapidly left its former status as being a niche asset class. It's becoming a major standalone asset class, and so capital is coming in.

Omar Jaffrety: Yes.

David Snow: How do you predict the increased competition will show itself as more capital comes in?

Omar Jaffrety: I'll give you a perfect example. In our Fund One, when we created it in 2014, we might have accidentally been the first digital infrastructure fund. That word wasn’t even used—it wasn’t called digital infrastructure.

We said, look, these are such high-quality assets that if you're a secured fixed income player, you'll understand this and you'll want this because it's durable and long-lasting over time.

We think insurance money that needs to match long-duration liabilities with long-duration assets that are inflation-hedged will like this. We think these are REITs and real estate-centric assets, or real estate people might want to go into digital real estate.

This ecosystem checks all those boxes. It checks the insurance box, it checks the pension box, it checks the infrastructure box, it checks real estate boxes. Lots of folks want to buy assets that have durable returns, in scale. So what are we doing? We're trying to arbitrage that by creating that scale—buying things one at a time. I put 1,000 of them together, I’ve got some scale. I put 2,000 together, I’ve got more scale. I build them one at a time, I get some scale.

We're trying to create that bouquet for the larger players, and there's probably a hundred potential buyers interested in this type of asset class who don't have the ground game and access we do as a specialist.

It’s just an incredibly exciting environment, even so much so that I would say other GPs who are not in our ecosystem are considering expanding into the ecosystem by investing in our area.

You may have seen Blue Owl buying IPI as an example, which is a very specific data center company. Blue Owl is a large aggregator of assets. That element of our business is evolving as well—larger players are figuring out how to fill in the gaps on digital infrastructure.

And those are some really important conversations to have—how do we all partner together inside this ecosystem of alternative asset investment? There are folks that are leading and investing, and then others with a lot of capital and LP relationships trying to figure it out.

It's a very vibrant environment to be discussing this ecosystem.

David Snow: And does it remain relatively friendly? I mean, has it reached the cutthroat point yet, or is everyone still trying to work together?

Omar Jaffrety: It's pretty friendly. There's a lot to go do, and a lot of opportunities to partner.

I think folks have realized it’s a big ecosystem. Whether we're partnering with corporates—we just did a transaction partnering with another GP—we’re working on another one, we're working on some corporate partnerships. You have to take that approach.

There's just a lot of vibrant dialogue within the space.

David Snow: Under what circumstances would Palistar Capital build a new data center? Is that something you would consider, or is that for someone else?

Omar Jaffrety: We've considered it quite a bit. To me, finding the right location and power is critical. Finding an anchor customer that's willing to backstop that contract.

What we're not comfortable doing is wildcatting—just finding a piece of land, hoping I get power, hoping I get broadband, hoping I get customers. There are people who do that. It's sort of a classically real estate thing: I’ve bought a piece of land, I'm going to build a building, then see if I can sell the apartments or lease the office space.

We’re reluctant to do that. It’s just not in our DNA.

Right now we own about 5,000 wireless locations around the U.S. You may have heard of edge data centers. It hasn't really picked up full deployment speed, but you can imagine, as cars are more connected, it's easier for a small data center to be at that tower location to speak quickly to that vehicle—versus going back to New Jersey and then returning to that location.

Over time, data centers will move more toward the edge. And we ultimately own the edgiest of edge assets—because we’re the ones where it goes wireless. So we're actively watching that and seeing how we can enable the land we own on these tower locations.

David Snow: So you're saying that if you own the land under a tower, you have upside optionality to perhaps add assets to that footprint? Can you talk more about that and how you think about the possible upside?

Omar Jaffrety: Traditionally, a tower sits on a block of land, and the tower leases that land from someone. It could be a municipality. Maybe they signed the contract 30 years ago. The muni has a contract on 3rd and Main, and on that sits a tower. The tower company’s paying rent and all is good.

As technology has advanced, there are several things you can do. You can bring backup power and lease it to the carrier, and now you've got an income stream. You can put in green power—there’s miniaturized solar and even miniaturized wind. You can bring batteries to begin redundancy, so now you've got another revenue stream from that land location. You can actually put some edge compute there—a little box—and folks have been testing the little box.

The question is: what content and smarts are you going to put on that box versus the big box you’ve already built with half a billion dollars? What’s going to go in the little box? That’s still being worked on.

Now, just having one piece of land isn't going to help you. There are 200,000 locations. How are you going to get attention with just one? So having a meaningful network and critical mass that can actually help in deployment is important.

That’s partly what we’re doing—aggregating so that we have access to meaningful scale to have those larger deployment conversations.

If you've got granular scale across a few thousand sites, it all adds up. That's where we are. We're spending time and doing R&D with our partners on what's important to them.

We're trying to map what we think is the art of the possible with what is practical for our customers. That’s the gap between a return on investment that’s sooner versus one that’s longer-duration.

David Snow: As you think about the supply chain of all the things necessary to build out digital infrastructure, are there any materials or natural resources you’re worried about—things where a shortage could slow everything down?

Omar Jaffrety: That’s a great question. Each sub-vertical probably has a slightly different answer.

In wireless, the choke point is spectrum. The world ran out of true wireless spectrum a few years ago. So governments went to satellite companies and said, “Give me some of your spectrum back.” That was called the C-band spectrum. Companies in the U.S. spent almost $100 billion on that type of spectrum.

The derivative of the choke point is: the more choke point you have, the more you need to divvy up what you’ve got. So as asset owners, we benefit. If you have more spectrum, you're going to use it more on my location, so I’ll see the benefit. But the less you have, the more you need to diversify and densify your network.

That’s the front-end choke point.

On the other end now—really bizarrely—the availability of power is a choke point. AI came so fast, and the need is massive. It’s multiple times what’s available in a community. In certain communities, it’s beginning to squeeze out homes and businesses.

So legislators and local folks have to think hard: do I permit this data center and give it power from the grid? And how do I satisfy the rest of the businesses that need it?

Solar, for example, works during the day—doesn’t work at night. Wind works when it’s windy. These are episodic solutions—they’re not evergreen.

So you have to rethink how to supply power. Then you get into elements of the supply chain that connect to that—transformers, for example. There’s a four-year waitlist for transformers today. You never had that before. Transformers connect the grid to cleaning and making power available across the data center.

Folks are getting creative—going to secondary and tertiary suppliers for transformers. And you're seeing this in massive scale with NVIDIA—chips. No more chips. There’s a waiting list. Things you didn’t imagine as choke points are now choke points.

That’s just because demand has so dramatically outstripped supply.

That’s very bullish for folks who actually have access to capital, customers, and locations.

The final choke point is the availability of capital. There’s not an infinite amount. LPs who fund firms like us haven’t seen a lot of M&A activity, so they haven’t seen much return of capital.

So: wireless—spectrum; data centers—power and some critical components; and across all of this—capital.

David Snow: When an LP allocates to your fund, where does that allocation come from? In institutional investing, people like to talk about buckets—infra, fixed income, real estate. If a large institution didn’t previously have a digital infrastructure allocation, where are they pulling it from?

Omar Jaffrety: That’s a great question. I’ve been raising capital for almost a decade, and in earlier conversations it was lost—folks didn’t know where to put us.

It used to be private equity and credit. Those were the two original things. Infrastructure sort of squeezed in as a derivative of credit. It’s secured, it’s credit-like, it’s long-duration. So it came from that window.

Private equity used to own data centers. It used to own fiber. Now, as infrastructure has grown, certain LPs have developed a real assets area.

Real assets means real estate. If I own fiber plants—real estate. I own wireless locations—real estate. I own data centers—real estate.

The easiest place for it to come from is the real assets bucket. If they don’t truly have a real assets bucket, it’s generally called real estate.

David Snow: Well, Omar Jaffrety from Palistar Capital, thank you so much for being on Cool Vector today. I hope I can invite you back to a future episode.

Omar Jaffrety: Thank you, David. That was great. Lots of fun.

END