Speaker 1Digital infrastructure has emerged as one of the most exciting sectors in the asset class. And over the past few years, particularly since the advent of the COVID pandemic, digital infrastructure's growth has accelerated rapidly. Artificial intelligence is playing a key part in this, necessitating the construction of more and more data centers. But digital infrastructure is bigger than that, and fiber and towers are both significant pieces of the puzzle. In this special episode, sponsored by Palastar Capital and Edmund Roschild, we will discuss the outlooks for each of these subsectors. We're joined by Josh Obler, an investment partner at Palastar, which is an investor laser-focused on the digital infrastructure space, and by Jean-Francis Duche, Global Head of Infrastructure and Structured Finance at EDR, which also runs an infrastructure debt franchise called Bridge. We will explore how AI is transforming the data center landscape, where to find the best opportunities in finance, and how to make the most out of the data center landscape. Thank you for watching. and why towers continue to make such a good investment. I'm James Lineker with PEI Group, and this is the Infrastructure Investor Podcast. It would not be possible to discuss digital infrastructure at the moment without touching on data centers. I put it to Duche and Obler that data centers are the hottest investment in infrastructure right now, and they agree. Here's Josh Obler. I think that's true,
Speaker 2and I think it's not just maybe infrastructure. I think it's probably all of Wall Street mobilizing for the big hyperscaler players, kind of $200 billion a year capex as an astronomical number that they're putting to work in the space. I think we're all building behind that power infrastructure, actual data center accumulating the land to support all of that. You've got a very, very tight current market, and I think that you've got a very, very long-term investable trend inside of the data center. I think that's true. I think that's true. I think that's true. So I think that everyone involved in the sector is now highly focused on how do you do this and how do you do it appropriately.
Speaker 1With tight supply chains throughout the market, equipment lead times lengthening, complexities in sourcing power, and a tight labor supply, there are challenges to overcome. There are also questions to be answered.
Speaker 2How do you make sure that you're not getting maybe taken out with the tide of huge investable quantums of dollars, making sure that you're investing behind the scenes? I think that's true. I think that's true. I think that's true. I think long-term stable projects, things that we can truly call infrastructure. So there's a lot to figure out today in this space. You know, it's really interesting. We brought on a new partner of ours, a gentleman, Gonzalo Bernardo, who ran the data center business at OMERS. But prior to that, he was an energy and infrastructure banker at Morgan Stanley. So you can start to see how a
Speaker 1number of skill sets are getting crossed here. Jean-Francis from EDR also sees the data center boom as a long-term opportunity, requiring a diverse skill set to fully understand. The temptation may be to chase every data center deal in sight, but it's important to be more disciplined than that. So how do you do that intelligently?
Speaker 3Well, hey, you hire the right people. And similar to what Josh said, I asked when we set up the Bridge platform, an ex-colleague of mine in banking, who was a veteran, who is a veteran of digital infrastructure project financing, to join and tell me if there was a case for data centers, because as always, the question of the definition of infrastructure and is it real estate? Is it infrastructure? And that was the first point to hire talent. Second was to, yes, structure it in a way where you could argue that you're meeting from the debt standpoint, the features that investors expect from an infrastructure play. It took us four years to make a first investment in data center. And that was, you know, I won't name the offtaker, but A-rated, mega large company, corporate. We were in Schiphol, so a good location. And, you know, it was like a 20-year contract. So it was almost creating a corporate, public-private partnership to a certain degree. And we got our risk comfortable with that. But as you know, we invested more and more. We could take a bit more risk, have more diversification in the subscriber revenue base. And that's how we started to really invest maybe more and across perhaps the whole range of opportunities that data center offer. But for us, and perhaps if we focus on Europe, we need also to demonstrate a few things, you know, European investors and asset managers as a consequence are very focused on ESG. So of course, we're going to try to make sure it's a good location. We're going to see if we are powered by renewable energy. That's a nice way to invest in the space.
Speaker 1Obler is looking for investments to tick many of those same boxes.
Speaker 2If you step back and we look at, I would say, the entire breadth of the sectors in digital infrastructure. So whether it's micro wireless, fiber data centers, we think that there's probably a huge range of risk that you can take on as a digital infrastructure investor. And we think that you can invest in this space ranging from really core, super core assets, really long duration, really kind of stable assets. And then you can move toward almost private equity type risks. And maybe on kind of the far end of the nexus, you can almost invest in basically growth equity in and around some of these types of subsectors. And I think that if you look at, you know, the opportunities in the data center space, then the question is, how do you make it kind of the most infrastructure-like? And we've got two modules that we think that we can turn, if you want to call them boxes. The one is business model, right? So how much risk am I taking on development? How much risk am I taking on my future need to land new customers? Am I investing in an asset that's already cash-flowing, already something that is in the ground? So one, what is the business plan and what is the operational risk that we are looking to support? And then two, how do we modulate that with capital structure? You know, broadly speaking, I think that we've got long-term sustainable trends behind the digital infrastructure space. This is a multi-trillion dollar global investable theme for the next decade plus, right? I think that we are all using huge sums of data. I think that there's new use cases coming out, what would you say, kind of on a daily basis, right? That people are doing more with their devices. AI has obviously been a huge boon, driving mobility, driving use cases like this. So I think that the asset base, the infrastructure required necessary and hugely long-term, but you always kind of need to keep an eye on it. And I think if you can focus on true fundamentals and cashflow, but the market loses sight of these things all the time, right? Valuations have a way of running ahead of themselves. People can invest way out, kind of forward on the curve. So you're always watching and I think trying to be mindful of all these types of developments.
Speaker 1AI has been a boon, as Obler says. And the world we live in today is markedly different, even from the world of five years ago, and is expected to change drastically in the next five years too. But are we all getting ahead of ourselves when we talk about AI? While it is fueling growth in digital infrastructure, are we in danger of overestimating its potential? Dusch thinks AI is fundamental to our future.
Speaker 3I think we're broadening to the whole digital infraplay. And actually you could argue that AI goes even beyond because I'm a big fan of the intermodality, interconnectivity of infrastructure. And I always say, you know, if you think of smart cities and whether you're managing what I call the iCloud of heavy charging, you're trying to optimize the consumptions up to, you know, the meters that, or smart meters that are being used, how you try to create energy efficiencies in the social space. I think all that points at a need for data and yes, AI is there. I even realized that some peers use AI to write credit memos. So, you know, it's everywhere, it's happening. It hurts a bit my old school values and thinking our business is about people, but actually you could argue it's complimentary. But now that I've said that, yeah, I think AI is there. The need for data is definitely there. And that's what makes the digital infraspace fascinating because it's sort of, you know, back to Europe, fiber optic, data center, towers, and all you can do. All that offers a lot of A, opportunities, B, diversification. And you can see that from countries to countries, they have different ways, different technologies, they try to push forward. And it's our job as asset managers to understand technologies. And interestingly, it's funny because Josh mentioned internet and I think we're old enough to have been in this business in the late nineties, early 2000. And it went from fantastic to overvalued, to really wrong. But if you look at the digital play now within the whole infrastructure universe, in many ways, I think that the debt structures are much, you know, before it was always a quasi LBO, leverage finance, it's a bit more project finance type. And I don't think the valuation of the underlying business are crazy. Whereas there are other sectors, and I'm not going to kill the other sectors we invested, but we are obviously more cautious. And I will say that, you know, I think it's, and I will always say that the digital infrastructure, as much as there is the AI challenge, but in terms of its positioning, et cetera, it's probably quite sound. And in many ways, the underpinning infrastructure are quite switching to traditional infrastructure. And I love digital. I think it's fascinating, but I think it's also a safe place.
Speaker 1Data centers may be a safe place for investors, and there's certainly no shortage of excitement about that part of the market. Yeah. But how about fiber? That's a sector which has faced its fair share of headwinds, but Palastar and EDR are both active in it. In fact, I put it to Dush that some 60% of EDR's digital infrastructure portfolio appears to be in fiber. He says that sounds about right.
Speaker 3It's interesting. We had an investment committee where we discussed whether we had too many investments in fiber optic. And our position as investment managers is no. And if you look at our overall platform, probably 25% is digital, quite important from where we started. And fiber optic is probably 15%. So you could argue that almost from a mathematical standpoint, we're in that 60% zone, probably also because of the scale of some investments. So yes, it is that. And we've been an early mover, but it's true as well that we're trying to keep it within the assets we want to invest in. A lot has been done already in terms of deployment of fiber, but there's still a lot of opportunities. And I think the key word is how selective you can be.
Speaker 1Being selective. This is a theme that also came up when we discussed data centers. So how important is it to be selective when looking at fiber?
Speaker 3It is crucial for us. And like I said, each year we have up to 15 opportunities for fiber optic. And we go through, you know, quite a comprehensive grid of criteria to see what could make sense for our portfolio if we tick the boxes which we think are right. And by the way, right doesn't mean we're risk averse, right? Is it a new country? Is it diversification? Or, you know, is it complete? It's elementary for what we've done. Is there competition from, I don't know, other technologies, satellite? And so we go through and we select, you know, the five, six investments we're going to focus on for fiber optic. Okay. So where are those opportunities? I think there's a need that more countries, again, in Europe, we invested initially a lot in France, not that long, actually, because we're not totally satisfied with that quasi PPP scheme. We thought it was good, but not. The traditional PPP, but we invested. We invested in Germany, we invested in other countries, but Scandinavia has a few good opportunities now as well. Look, across Europe, still a lot of opportunities in core Europe or other countries. But, and I mentioned it, you know, we are beginning to look at investment opportunities in the US. There's a lot going on in the US. And I think Josh is probably the best suited because, you know, debt, as ambitious as we are, we're not going to be able to do that. So we're going to have to look at investment opportunities as we are. Comes after the equity holders or whatever have made investments. And you're in New York, so you may observe quite some activity there.
Speaker 2We see a lot of activity, obviously, in the US market. We see it, you know, internationally as well. Europe, the UK, a lot of kind of investable opportunities in the fiber sector. You know, in the market here in the US, there's a ton of investment going on. And there's a lot of opportunity to build both connecting people's homes with fiber, fiberizing the US market. Traditionally, the US market was connected all with copper. Lines, coaxial fiber. We had a big build out, you know, decades ago of fiber in the US, which didn't necessarily happen in other countries around the world. And cable played a pretty good role in getting the US connected to the internet and supplying broadband to huge chunks of the country. But I think as time moves on and people have big demand for symmetrical speeds, you know, big capacity up and down fibers, you know, one of the few solutions that can really provide that in a big way. And so I think you're going to see much, much higher penetration of fiber across the US market. And that's, you know, providing both the people's homes, which we see a lot of opportunity to support those build plans, as well as enterprise fiber, as well as, you know, more long haul type fiber, connecting data centers and things like that. You know, fiber businesses, if you ask us, probably are a little bit more competitive than some of the other project financing type opportunities you might see around digital infrastructure compared to, let's say, certain types of data centers or compared to macro wireless towers. So when we look at fiber, I think, you know, we're very cognizant of are they free cash flow positive? Or are they in their development cycle? How much money needs to be invested for them to get to their homes past or to support the amount of penetration that they need to get to a steady state place in their business? What are their churn rates?
Speaker 1Obler suggests the key to success in the fiber business is structuring. At the heart of this is how capital can be used best to invest in and develop broadband businesses. I refer to
Speaker 2structure in two ways. The one is we've got to be able to do this. We've got, we call them dark fiber or wholesale partnerships with private network operators that can be, you know, basically an offtake. So we become more of a physical asset network provider and someone else will then manage the customer relationship. And that's a pretty good slicing of the risk. And I think that we, you know, like to own physical network asset, high margin business, and then someone else can be in more of the customer facing role. So that's one example. And the second is using capital. Capital structure. So, you know, trying to maybe get a little bit more senior in using preferred instruments or sort of debt-like instruments where we can use capital to maybe catalyze completion of some of the fiber network build-outs that are going on. So then how can capital play an important role either to drive consolidation, rationalization, and basically get the network to where it needs to be kind of completed and finished. So there's just a ton of opportunity, a big need for capital.
Speaker 1So we have heard that there is plenty of opportunity in data centers and in fiber. But how about in that third key pillar of digital infrastructure? Towers. In some ways, towers feel like a very traditional sector of the asset class, but that's certainly not a criticism.
Speaker 3There's that almost feeling that it might be old school at the same time. It is probably also what investors would see as a tangible infrastructure, obviously. And we've all known from the moment, you know, mobile operators, MVNOs, you needed these towers. Yeah, we feel it's a sector where we can have either, you know, opportunities in the senior investment grade space. And I would say for us, it's a question of finding the right balance between a certain risk profile and how it's priced, because it can be priced aggressively in many ways, which is good for the equity holder. For us, it's a question. There's a lot of whole core situations. But what's interesting also linked to towers is made us think of other sectors, like what we call, you know, land infrastructure aggregators, you know, where, you know, you would find land where, you know, you can grant a right or the ownership of the land so that people can manage, for example, towers.
Speaker 1Macro wireless is an interesting angle to this. Palastar has been active in metro markets for over a decade. It can be tricky for wireless carriers to provide coverage in urban areas where densely packed buildings pose a clear challenge. So rooftop locations have some clear benefits, as well as some that may be less obvious.
Speaker 2We started investing in, you know, rooftop assets, as well as the ground lease underneath the cell phone tower. So pretty niche portions of the wireless infrastructure space, but highly defensive, very, very high margin, triple net businesses that we would basically run at 95% free cash flow margin on long term off takes with mission critical assets, investment grade kind of, you know, triple B plus. A-rated type counterparties. And the amazing thing about these assets is that they also grow, right? So in the U.S. market, you've got a lot of growth from co-location, which means, you know, if you've got an existing lease with Verizon or AT&T that your second customer will come on and co-locate in that same location. You can amend the lease. There's contracted escalators that are part of the economics there. So I agree with you, you know, kind of feels like old school infrastructure, you know, very, very long term assets in place for a very long period of time. We see, you know, little to no churn on those assets. So, you know, highly, highly valuable.
Speaker 1With data centers, demand grew rapidly with COVID and growth is accelerating again as AI comes online. So could a technological shift from 5G to 6G have a similar effect for towers? Well, maybe not. And Dush suggests we shouldn't get ahead of ourselves. There could, however, be a boost from AI in this space.
Speaker 3It's always exciting. But, you know, the thing is, my take is, have we invested in 5G? No, I think the mobile operators have sufficient cash flow to self-finance it or through corporate debt. So the days where we were fighting, you know, to win the right to be an underwriter of, you know, UMTS are probably behind. Also, perhaps because the licenses, you know, prices have gone down as well. So I would say 5G, 6G. We monitor it for us. The key thing is, and it's almost back to towers, is how you can deploy it in some areas efficiently. You know, because we always talk of competition, fiber optics, satellite. And I would love to finance a satellite because I've been trying for the last 25 years. So at some stage that will come. But, you know, what is interesting with all these technologies is how it points at the needs and perhaps to use the various technologies available in the digital space. To create, you know, the kind of almost full coverage we're trying to achieve. And that for me is a challenge which we've already seen with 5G to a certain degree.
Speaker 2I think AI is going to create more demand for physical assets across all three subsectors. This is one giant connected value chain ecosystem, right? Having data centers connected by a robust fiber, robust fiber connecting people's homes, connecting people's businesses and connecting cell phones. And then mobility obviously plays a key part in data consumption. Just, you know, how do we use our phones? How do we manage our lives? I would say that, you know, it's probably 30, 40 years, right, of mobile playing an increasingly large role in our lives. So it would be hard to see us not seeing a lot of mobility. about participation from all of the investment going to AI, whether that is using your device to review Instagram, Facebook, and seeing, you know, new content generated for advertising or who is receiving which ad, whether it's interacting with chat GPT to ask for something advice on the move, on the go, or inside of your home, you're always kind of interacting through your personal device, you know, whether that's, you know, more robust physical infrastructure for connect devices, cars, you know, so forth, industrial applications, all of that is going to create more demand and more need for mobile infrastructure, right? And for towers.
Speaker 1Clearly, there are exciting developments on the horizon and digital infrastructure is riding the wave of technological advancement. Predicting what will come next is never easy, but it is also quite fun. So that's exactly what I asked both of our guests to do. Here's what they had to say.
Speaker 3When we launched Bridge 10 years ago, it was, yeah, fiber optic is emerging. Is there really a case in Europe to project finance 5G? Maybe not. Satellite, we discussed it. Towers, yes, believed in it. Data center, we had to make a case. Now, I wonder if there's a technological shift, per se, a new technology that might arise, or if it is all about, for example, being able to provide that huge capacity that we need. You know, if you have a. If you ask me about even energy, you know, five years ago, I would have said, you remember, it was like, ah, Gen 2, energy transition in the field of energy, green mobility for energy, battery storage, cogeneration, maybe hydrogen. Me, I don't see a new technology emerging because even 4G, since 4G, I think 5G, 6G is an evolution. And I think it's more about finding that extra capacity when we look at data centers. Are they still good? Are they still good sites where to develop these data centers? So that's what I would say. You know, maybe, you know, we improve the efficiency of the infrastructure. We find space to do a bit more. And I think, and maybe I'm wrong, but I was listening to you, Josh, and I think it's more about how you connect the different technologies and use them in the most efficient way to provide, you know, ultimately the data and satisfy the demand. Also linked to AI.
Speaker 1Obler agrees that technology is key in a space such as digital infrastructure, where advancements are made so regularly. A lot of this advancement is incremental. AI could be more of a step change.
Speaker 2I think that looking forward a couple of years, if you said in the tower industry, I think we will see kind of in the U.S. market, let's say, modest incremental growth in tower count. And then you're going to see, as we approach the end of the decade, you know, further accelerating of CapEx investment from the wireless carriers, as they prepare for, you know, 6G, as they continue to densify their networks for 5G. I think if you look in the fiber space, the U.S. market is going to keep pushing to approach, I don't know if it's going to be 100% penetration, but, you know, getting fiber past as many U.S. homes as possible. I think that there's a good long-term investable theme behind that, you know, kind of beyond the next two, three years. And then I think, you know, data centers, data centers are going to be talking about AI for a long time. I think that there's, you know, tremendous movement inside of the U.S. tech companies to keep investing in data centers. I think that Wall Street views it as growth. I think that there's, honestly, national security implications around data centers and, you know, kind of global questions of who's driving the next kind of generation of compute and technology. I think one of the more interesting questions that we're all going to see over the next couple of years is where are we going to find enough power generation to support all of the new data center cases? And is it going to come from a cleaner source? Is it going to buy us toward nuclear? Nuclear has very long kind of lead times behind it. It's never going to be able to be, you know, kind of, you know, wind or solar powered, perhaps not.
Speaker 3Totally agree with you. That's the question mark. And it's interesting because you nailed it in terms, again, of the interconnectivity to a certain degree and intermodality of infrastructure. And it always comes back to the question of are we going to manage enough power generation? And it's interesting because for us as an asset manager so far, we have had an exclusion for nuclear for our investor in Europe. But we've seen, for example. Some countries, France, and it's not because I'm French, we're London based, but France has been, you know, an early mover in the nuclear space. And they even, you know, consider that it was part of the taxonomy in definition in France. But taxonomy has different interpretations across Europe. But you can see other countries are starting to think about it, especially if you look at Japan being able to produce nuclear energy, perhaps without all the theme of a uranium waste, et cetera, the fission of particles. So it's interesting because you're right, the questions around many topics linked to data of cleaner, greener infrastructure is about the energy production.
Speaker 1That energy question is a big one. Data centers are expected to account for a significant slice of power generation in the near future. They already account for around 2% of US power generation and could hit 10%, albeit varying quite a lot by the year. That variability is true in Europe, too. In Ireland, for example, data centers have already been using more than 20% of a country's metered electricity. But the energy conversation could be a whole separate podcast. And unfortunately, that's all the time we have for now. For lots more on the growing opportunity set in digital infrastructure, or for that matter, the energy transition, or the rest of the infrastructure investing universe, go to www. you might also want to listen to more of our podcasts, and those are available in all the usual places or on PEI Group's various titles online. Thanks again to our guests, Jean-Francis Duch and Josh Obler from EDR and Palastar Capital. I'm James Lineker. Thanks for listening.
PodcastTranscript AI