American Tower Corporation ($AMT)

My name is Richard Choe. I cover Communications Infrastructure for JPMorgan. I'd like to thank Rod Smith, EVP and CFO of American Tower for being with us here today.

I'd like to just start off. It was nice to hear John Stankey earlier talk about there's not any easy way to build and invest in our networks. And in the end, I think long term, you need to have the best network to eventually win. And I think people forget kind of the critical part, American Tower and other tower companies play in this kind of evolving modern world. So in just kind of resetting expectations and kind of reminding people where you sit in the infrastructure, how do you feel your assets are in this ecosystem of kind of our increasing digital economy?

Yes. Good morning, Richard, it's great to be with you. And thanks, everyone, for joining. I think there's a couple of keys there that I heard you mention. One is the network quality. It's really network coverage, quality and capacity. That is critical for the networks, not only in the U.S. but around the globe to keep up with and perform well in an environment where mobile data consumption continues to grow rapidly. That really is the key to everything. The other word that you used, which I think is a key word is the critical infrastructure. Our infrastructure is amongst the most critical infrastructure within the wireless networks. Again, not just in the U.S., but also around the globe, the tower assets we have, coupled with the data center assets that we have in the U.S. And as we look at our portfolio today, we are perfectly positioned and probably strategically stronger now than we've ever been in terms of executing on and benefiting from that growth in mobile data consumption we see in the U.S., which is -- has been driven by technology advances, that will continue as 5G networks continue to kind of roll out there at tail end of rolling out 5G networks. 5G applications are still on the move. They will be coming in the future. That will put another wave of growth in mobile data consumption. 6G will be right around the corner. And AI workloads haven't even really hit the wireless networks yet, but they are coming as well. So there's a lot of evidence that there will just be a continuation of applications and services that require increases in mobile data consumption and our assets are perfectly positioned to help the networks perform the way they're intended to perform, but also help the networks to evolve to where they really need to get to, which is having those critical tower assets, having interconnection-rich, cloud on-ramp centric with heavy network density within the data centers that we have geographically spread around the world. And then potentially connecting those assets together, the data centers and the towers, to drive the edge to put more compute capacity closer to the base radios to put more -- to provide access to the cloud on-ramps closer to the base radios, there is going to be likely changes to the network that our assets are perfectly positioned to not only benefit our shareholders but also benefit the wireless carriers themselves and the subscribers to help the networks work well. So the U.S. backdrop is fantastic. We've got really what I would consider the best assets in the U.S. in terms of that infrastructure. Our tower assets are #1 in our position. No one has better assets in the U.S. than we have in terms of the portfolio. And again, our data center assets are very high quality, very unique and perfectly situated to drive the networks of the future. And then that is complemented by fantastic assets around the globe. The European assets we have are outperforming our business case. We expect them to continue to perform in line with the U.S. or faster in terms of the economic growth. Africa is growing a few hundred basis points faster than the U.S. now. That could continue for a number of years given where their network deployments are and the amount of infrastructure that the markets that we're in Africa need. Latin America is going through consolidation. It's been a fragmented market. This year, we'll mark the peak of churn in Latin America. Once that subsides, the markets will be in a more constructive position, particularly in Brazil, moving to a place where there are 3 wireless carriers today. And starting next year, we see churn decreasing rapidly and new business probably accelerating and that being very constructive to us in our Latin America portfolio returning to normalized growth when you get into moving towards that in '27 and '28 and beyond. So the backdrop looks really good. It's all centered with us with the critical assets, the highest quality assets in the U.S. within towers and data centers, complemented by the portfolios we have around the globe to take advantage of the different timing and cycle of network rollouts around the world.

Yes. I think it's great that you have this whole I guess, infrastructure that -- and we'll go back to data centers later, but you can kind of see what's coming down the pipeline. And right now, it seems like we have a lot of data creation, a lot of data analysis and generation at data centers, but eventually, people are going to use that on their devices. And the last mile is literally the tower. So I think people kind of get caught up a little bit in the kind of quarter-to-quarter, day-to-day, lease rates, churn. But over time, do you see any change coming in terms of your long-term domestic growth rate? Or do you see kind of the same outlook that we've seen for some time?

We've been focused on driving a higher level of quality in our earnings across the globe, including in the U.S. Certainly, in the last few years in the U.S. specifically, we've had a couple of event-driven noteworthy churn events, being Sprint and DISH. With the absence of those, the other parts of our U.S. cash flow and revenue streams have been very consistent. And we think we return to that consistency next year when our cash and flow in our revenue no longer has Sprint, which it doesn't today, and it will no longer have DISH in it. And what is remaining is revenues from the 3 big wireless carriers, and we all know where they are in their network development and their focus on wireless networks. So they're -- I guess, I would say, they're renewed focus and prioritization of their wireless networks. So we're heading into an environment really beginning in '27, where consistency is the word particularly in the U.S., stable mid-single-digit organic tenant billings growth, a steadfast focus on efficiency, cost control and margin expansion and then really smart, disciplined capital allocation, all supported by a really strong balance sheet, highest credit rating in our environment or in our segment and the lowest amount of debt. We've reduced our reliance on floating rate debt. So we've got much more certainty in our interest charge line within AFFO. So we're in a really good place. And when the U.S. has that type of stability, which we are getting to and we'll be there for next year. The rest of the globe complements that. The Europe, Africa and LatAm markets complement that. And all of those markets are heading to a place where they are likely to perform maybe consistently 200 to 300 basis points faster growth. You could have disruptions in FX here and there. This year, FX is a tailwind. FX is not always going to be wildly negative. It could be mildly negative on a normal basis and sometimes it will be positive. We're coming through a period where emerging market currencies have devalued relative to the U.S. over the last several years. It's probably going to be a little bit more stable in the next few years. So you have that stability in the currencies combined with the amount of development that those markets need and the high growth rates that we're seeing, as an example, in Africa, that's going to be really constructive and complement that mid-single-digit AFFO growth that we see in the U.S.

Yes. And you were talking about having your offering last night in our I think it gives you a lot of flexibility in how you finance the business. But I guess for people that don't know as much about international markets, can you talk a little bit on how the organic growth is because you talked about FX being up and down. And then there's also CPI escalators that could move around. But it seems like there's real significant organic growth coming from these markets, maybe at different levels and different speeds. But in the end, I think the ability to communicate the need for good networks is kind of universal and maybe separating out the discussion a little bit between your European assets versus your more emerging market assets. How do you view those opportunities as you sit here today?

Yes, I'll start by making a couple of comments on AFFO and AFFO per share. Our view is we will have industry-leading AFFO per share growth. That's going to be underpinned by mid-single-digit revenue growth in the U.S. complemented by faster revenue growth in emerging markets and double-digit revenue growth in our data center platform, a position that's perfectly positioned to deliver mid- to upper single-digit AFFO per share growth reliably. What that means from the non-U.S. markets is Europe is a set of very high-quality economies, growing the revenue stream there at mid-single digit or better, the same or a little faster than the U.S. is very constructive, and that's what we expect. And there's consistency there. We're insulated from significant churn events in Europe because much of our revenue is tied to telephonic on long-term contracts. So there really isn't the opportunity for these one-time non-recurring event-driven material churn like we saw with Sprint and DISH. So Europe is very stable and will be additive to the U.S. growth. Africa, we're seeing upper single-digit organic tenant billings growth, and that includes offsetting it with a few hundred basis points of churn, which is what we're seeing. That's probably a pretty normal environment for Africa. We think upper single-digit AFFO or organic tenant billings growth is what that market should deliver not every single year, sometimes it will be double digits. And sometimes it may be lower upper single digits, but it's going to be strong growth. That's what we expect there. The offset to that is FX and escalators. The escalators really are a natural hedge to offset the FX risk. So what drives the FX risk or the devaluation of currency is when their inflation rate is higher than the host country. So in our example, when inflation is higher in Nigeria than it is in the U.S., you can expect the Nigerian currency to devalue over time at a rate that equals the differential in the interest in the inflation rates. So there's a natural hedge built in there. You may have some ups and downs over time, but long term, we should be able to cover off the FX risk, and then we are left with the real growth, the activity base growth, offset by any churn, and in Africa, we see that being upper single digits. So I think it's going to be really constructive over the long term and additive to growth. It helps us bring up that mid-single-digit revenue growth in the U.S. By the time you get down to AFFO per share, you're at upper single digit. It's because you've got these emerging markets that can grow faster. It's -- we add the expense focus in addition to that and the margin expansion. And with all that said, there are other elements here that could drive upside, most notably, not the only one, but most notably, AI workloads tipping out of the large language model facilities and the hyperscale facilities into facilities like CoreSite at an increasing pace. And eventually, I agree with you that those workloads will be getting out on to mobile devices. And that means that it needs equipment in towers in order for the networks to function properly. So this -- my expectation is there's a never-ending cycle of capital into wireless network to keep up with the insatiable demand of mobile data consumption. And that's just going to continue. And there will be applications and services available to all of Waza mobile devices that we can't even imagine today.

It's funny. I want to hit Latin America and Brazil at some point. But since you went to data centers, 1 have to talk about CoreSite. I mean, there are very few assets. I feel like that have the interconnection density and footprint that CoreSite has and Equinix in Digital Realty are there also. But CoreSite was a great acquisition. Kind of how are you looking at your investment in CoreSite? How much can you invest and grow that business? And what kind of perspective does it give you in terms of your tower assets as you see AI inference kind of grow?

Yes. We -- we've owned CoreSite for a number of years now, and it's performing exceptionally well, better than we originally underwrote. We've had a number of record-setting new business years in a row. And the demand in the pipeline continues to be strong. So some say our timing was lucky. We like to think that this was a really important asset that we identified and bought. The reason we bought it was strategic. It wasn't just financial. It wasn't just opportunistic. It was the idea of advancing the networks within the U.S. and around the globe to the point where they can handle the higher levels of service that come with 5G applications, 6G will be coming. And now with AI, it just accelerates all of that, where to have the services work properly and for the user to get the full extent of those types of services, the networks have to be reengineered. The ability for the wireless networks to have symmetry with uplink and downlink is going to be really important. That doesn't exist today, and that's got to be built in. When that happens, the latency within the networks also has to improve for 5G, 6G and AI applications to function properly. That means you want to have more compute power paired at the tower site where the base radios are with a network that has as much uplink capacity as it does down line capacity. That means more antennas and lines than the towers and more compute type equipment, maybe cloud on-ramps and interconnection ability at the tower site. So with our CoreSite facilities connecting those into tower assets, we are in a perfect position to continue to benefit from the development of the mobile data, the mobile networks and the mobile data consumption growth that we see. One of the interesting things between the data centers and towers. For us, they're not just 2 asset classes. They really are strategic and they represent a critical infrastructure in these networks of the future. And the demand drivers are very similar, if not outright the same. It's compute power. It's what the end user is going to be doing on their devices, whether they're mobile or stationary. We spend lots of time we have for decades understanding how people use devices and what that means to traffic on networks and what traffic on networks means to carrier investment and our ability to support that and monetize it. It's the same on both sides. The activity and the growth that we see in data centers gets out and drives growth on the towers. So we've got the same revenue drivers, the same demand drivers that we understand really well, and we can invest in that and not just in towers, but also in data centers very successful. And I think we've proven that. So we're going to continue to invest in data centers. We've got those investments up now from 200, it's gone to 400, then 600, 800 in it. That's because there is a record-setting pipeline. We're getting double-digit growth as a result of that. We're getting to stabilized yields faster, which means we then have to replace new available capacity. And eventually, we hope to and look to tie those assets into the tower sites.

Yes. I think people should realize that it's not just your availability of data centers and power, but it's the connectivity. And I think a lot of the wireless networks right now is kind of download focused. But with AI applications, we're kind of seeing more 2-way traffic, call it, and uploads. I guess as CoreSite has developed, like what should people know about the kind of connectivity aspect that is selling the product, not just having data center space in power. As you've seen things evolve, how has CoreSite kind of differentiate itself in a kind of overall data center market that seems to be going really well.

Yes, it's a great point that where differentiation is important. Not all data centers are the same. They don't all function the same. They're not at the same level of development. What we saw in CoreSite originally has proven out to be true, which is it is a high-quality platform with exceptional management. The high-quality platform really can be measured or identified in a number of ways. We've got cloud on-ramps and multiple cloud on-ramps represented in many, if not most of our facilities across the U.S. that in itself makes the centers much different than a data center without a cloud on ramps. So customers, the big name, enterprises want to be in there so they can tap right into the cloud on ramps sufficiently. The quality of the network, the number of networks that are represented in the are important to, we've got over 400 in our 9 campuses, lots of networks in there. And then interconnection represents the ability of enterprise customers to cross connect to 1 another and communicate directly with other enterprise customers. We are driving a level of interconnection is really only second behind Equinix. So for a relatively small portfolio on core site, it is interconnection rich, and it's a fantastic attribute of that network, certainly. The other pieces that people don't think about a lot is just the quality of the infrastructure within our data centers and how well they position us to be critical in the networks of the future. It's the power availability. It's the backup power availability. It's the cooling systems that allow the GPU density to increase in the amount of compute power in order for that to work, the facilities have to have the right setup in the right environment and CoreSite facilities have that across the board. So we have been increasing the GPU density. We use liquid cooling in a lot of our facilities that's central to the the assets we have. That's what is driving the customers that are in there to increase their interconnection, which is really them increasing their commitment to be within that facility over the long term. And then they want more space and more power, which is why we're building so much as to keep up with the demand not only from our existing customers, but new customers that want to come in and be part of that ecosystem. Tapping the networks, tapping in the cloud on rents, interconnecting with 1 another, being able to increase their GPU density within our buildings in a cost-effective way where we already have the liquid cooling, we're able to have -- provide that density for them. So we have a very -- we call it differentiated, but it is of a quality that is really unmatched across the industry. And again, the interconnection in the cloud on-ramp is only second to Equinix.

Yes. And it seems like your business is growing not just with hyperscalers but also enterprises and I guess, some of the leading-edge new tech or AI companies that -- what are you seeing from more traditional kind of enterprise businesses? And what are your conversations like in terms of the capacity or the level of connectivity that they're looking to buy from CoreSite?

Yes. I mean we have a whole host of enterprise customers that provide some of their workload requirements in our CoreSite facilities. Over the last few decades, we saw a shift from enterprise customers building out their own computer rooms within their facilities, managing all their IT on their own, keeping all their data and everything on-premise. That over time shifted towards cloud. and everything was moving out of the enterprise location and moving remote off into the cloud. And now we see a hybrid where part of the enterprise compute and data they want access to a closer or in their facility and much of it can still be up in the cloud. CoreSite is the piece there that stays closer to the enterprise. So we're seeing a lot of enterprise users pull things out of the cloud and put it in CoreSite and then they want that cloud on-ramp, so they can connect back into the cloud on-ramp. But they have close proximity of their compute to their enterprise location, which is key. We see that the demand of compute power, just the demand just continues to go up. The other piece that we're seeing, and this is maybe specific to AI, but it's also kind of a technology evolution where the amount of data processing is just going up and up and up on the networks. We are seeing now enterprise customers continue to do all the normal things that they would do and maybe even at an increasing pace. But now we're also seeing them look for space, power and cooling and GPUs and the density to build their own smaller language model so they can figure out how AI is going to be applied to their enterprise and applied specifically to their data, where they'll use the large language models, but they also build their own to process their own data. That is something that we are seeing taking shape within our CoreSite facilities think we're at the very beginning of that. It's not cannibalizing the existing activity. It's additive to that. And when I say all companies, even American Tower, we're all doing it, we're all looking at ways to use AI within our businesses, and it's going to require more compute power and more facility space and dedicated modules and smaller language models. We're seeing that hit the data centers early on here, but I do think we're at early stage of that. And that's why we see compared with the traditional business continuing to grow double-digit growth within CoreSite and AI just coming into the facilities. We just see a very constructive pathway for very solid growth and the ability to continue to deploy capital into CoreSite.

And it's interesting because you have so much go over on your earnings call. I don't think we always get to that level of specificity. I think the private AI deployments that you're talking about, we weren't expecting until 2028 and the fact that you're seeing some of that now is impressive. But I think more so, I guess, your ability to handle that level of density is something that you haven't talked a whole lot about. Can you talk a little bit more about your ability to service these higher levels of density that people might not realize that CoreSite is able to?

Yes, I think we're in very good shape for a couple of reasons. One is many of our facilities are anchored with liquid golden chillers that allow us to cool very deep GPU dense cabinets and areas. So we've got that in place without a huge capital upgrade. And we've been doing that very successfully, having much more density with the GPUs and it works perfectly fine, and we can continue to do that. The other thing I would say is, years ago, when we first bought CoreSite, we expected a ramp-up in activity with a larger parent, more financial support, being able to invest into the tailwind that, that sector had, unlike CoreSite on its own as a public company, struggled investing a little bit. That was a pain point for them. And so we knew that, that was going to be coming. And so we were ahead of the curve in terms of procuring power in our facilities across the U.S. we've been at a pretty good clip expanding and buying land adjacent to or within the campus as we have to build brand-new shells. And we've been building brand new shelves in a number of facilities, and we have the power available and the land available to continue to do that to handle the pipeline and the backlog that we have today. So if we were waking up now and saying, AI is coming in we would be a few years behind in terms of trying to catch up. But luckily, early on, we began to ramp up capital investments, procure the land, procuring the power, and that has proven to be really wise. And we're continuing to accelerate that. That's why you see the CapEx that we're investing in CoreSite has gone up noticeably. But the key that I would also say there is it's within a very disciplined capital allocation approach. So we've sharply increased investments in CoreSite. We haven't sharply increased our capital investments overall. We've reallocated investment capital appetite away from emerging markets and towards the U.S. and in CoreSite, and we're doing build-to-suits in Europe.

Yes. It's funny that you can spend hundreds of millions of dollars and people don't [indiscernible] that level of capital. You mentioned it earlier, like you have this kind of AFFO per share growth algorithm. And you have this great CoreSite business, but sometimes it might get lost in the larger portfolio of assets. But -- it seems like you are committed to kind of keeping these assets together to help invest and leverage each of the assets off of each other and kind of under this financial portfolio that you can kind of allocate across the different types of businesses. What are investors missing? Because right now, there's kind of this disconnect of the valuation that the public markets are seeing versus what we're hearing, how the businesses will do or are doing. How do you look at these different parts? And what do you think people are missing in these assets kind of being together?

I think the first thing I would say is that these are long-term assets and creating value for shareholders over the long term is really the lens to look at this infrastructure in this real estate, both towers and data centers. That goes for our U.S. assets as well as our assets around the globe. They're long term in nature. And the value creation is over the long term. And when you look at it through that lens, consistent earnings growth and even small improvements over the way compounded over decades, creates a lot of value. So if you have long-term investors, that really look at these assets and they worry less about, oh, there's 1 year of DISH churn or we had this happen [indiscernible] India had a little dilution like those things come and go, and we would rather have them not come at all, which is why we've been focused on quality of earnings, which I think we've made great improvements there. But now we are sitting in a place where we have really the essential infrastructure for networks around the globe to work properly. And we have our very best assets in the U.S., the most constructive, highest quality economy where we see the demand is just going to continue as we go forward and there could be synergies between towers and data centers and driving the edge and tying things together. But even if there isn't these -- both of these assets are really well positioned and they're both critical to the networks, and we understand the revenue drivers. They're not misplaced by having them together, in our view. What I think investors might be missing is not looking at that longer-term view. Being too worried about what happened this quarter, last quarter, as new business is going to be at this level this year, or is it a little bit lower than that. Think about this over the long term, think about the growth in data consumption over the networks, think about the technology upgrades, the new spectrum, the services that will be coming, try to envision the services that you can't even envision today, but you know they'll be here. The world will be different 10 years from now and our assets will be essential to make the networks work well, and we and the shareholders will benefit from that. That's one of the things I think the private investors do a little bit differently is they don't have quarterly reporting requirements. They're not overly concerned about onetime events and growth on a quarterly basis, their internal rate of return driven. And there's a time horizon that they invest in, and that's getting longer with these assets, not shorter. Even private equity invest in these assets for 7 years, 10 years or beyond, and they really look at the revenue drivers, their economic ability and power to drive a certain internal rate of return over that time and they're very comfortable with what they see and they price the assets accordingly. I think public investors look at the shorter-term volatility a little bit too much. You're not getting full credit for the data center. So maybe you should separate those. We may not be getting full credit for the data centers. We may not be getting full credit for the towers. Over time, I think it will become much more clear how important these assets are to the networks, how anchored our company is with these fabulous U.S. assets and the nice complement what we have with assets around the globe and how they will contribute to growth over the long term.

Yes. And I mean, I guess it feels very short term to think you need to try to realize value for certain assets at a certain time. But the way you're looking at it is these are probably critical assets that would be hard to replicate in another way if you're starting from scratch or even a private enterprise. Is that how you're viewing kind of your whole portfolio in view?

Yes, I think that's exactly right. I mean when you think of the changes that will happen to the networks, the requirement for more compute power and cloud access closer to the base radios, you'd have network networking companies needing to talk to tower companies as well as data center companies and maybe others or for a tower company to participate in the Edge that need to be a partnership with a data center company that gets very complicated. We have the greatest assets in the U.S. We have some of the greatest -- we have the greatest tower assets in the U.S. We have some of the greatest data center assets. We have the cloud on-ramps, the compute power, the interconnection, high-quality names, all the hyperscale players are in our facilities. They are our customers now. The folks that will make the Edge work are our customers. The folks that need the Edge to work are our customers as well. We are perfectly positioned to tie things together. We don't need to do economic sharing to kind of partnership to bring the right assets. We own the right assets, and we're in a really good position. I think that will become more evident over time. And you'll see just how critical these assets are and how stable the revenue and earnings growth can be notwithstanding the fact that we've had a couple of years, a few years where we've had these onetime non-recurring event-driven headwinds. The Sprint churn, the DISH churn and things like that, but that really is behind us.

Great. I think we'll leave it out there, and it will be nice next year to not have to talk about those onetime items. Have a good day.

Yes. Thank you, Richard. Thanks, everyone.