SBA Communications Corporation (SBAC) Earnings Call Transcript & Summary

Okay. Well, good morning, everyone, and welcome to the 8th Annual MoffettNathanson Media & Communications Summit. I'm Nick Del Deo. And joining me is Brendan Cavanagh, the longtime CFO of SBA Communications. Brendan, you've been a regular participant at our conferences over the last several years, and I really appreciate you taking the time out of your schedule to join us once again today. So thank you.

Yes. No problem. Happy to do it, Nick.

Yes. Let's dive right in. I thought we would start with the domestic leasing outlook as that's the topic investors are most focused on and the most important potential driver of your business over the coming years. So 2022 looks like it's shaping up to be a good year for growth, hopefully with some follow through into 2023, after softer performances in 2020 and '21. You're starting to see a bump in site development revenue in support of future carrier deployments. And on your earnings call a couple of weeks ago, you talked about how your backlogs are expanding. So now as you think about the growth trajectory over the past couple of years, in the next couple of years and how expectations got reset after the T-Mobile-Sprint merger-related pause, is the totality of growth or the totality of revenue that you'll think you'll be adding consistent with your initial expectations or has there been an underlying adjustment to it?

Well, I think that the total revenue we'll sign up will ultimately be consistent with our initial expectations. It's starting a little later than what we originally thought, but the carrier needs for the network spending in order to -- if they're going to deliver true nationwide 5G coverage, that really hasn't changed. And so I think the recent messaging when you think about what each of the carriers, the major customers of ours, have been saying, it kind of supports that. There have been several public commitments by the carriers to significant wireless network CapEx over the next several years. And I'm sure we'll get to it, but we've signed global agreements with some of our customers as well. So all of these factors, along with what we're seeing as growing application backlogs, give us pretty good confidence in an increased level of growth coming up soon.

Okay. Now when you talk about growing backlogs, how do you define the backlog? Is that just the number of lease applications? Or is it an estimate of the revenue that the leases might encompass?

Well, for our leasing business, we're talking about the number of applications for new leases and for new amendments. It's not really based on the dollar value. However, this year, increasing the number of new equipment installations or equipment modifications that are being requested will drive increased revenue dollars, obviously. The reason we don't base it on dollars is because there's a price negotiation that is oftentimes needed that's not locked in until the lease or amendment is executed. So that -- you don't necessarily know what it's going to be. But the changes in the backlogs, nonetheless, are indicators of shifts and trends of carrier activity levels. So that's why we count it that way. And just for clarity, because sometimes there's confusion on this, the timing of an application coming in to when we are booking revenue is typically about 6 to 9 months. Although depending on the specific circumstances, it can obviously be shorter or longer than that for a particular lease or amendment, but that's about the average.

Okay. Okay. That's helpful. I think Verizon is widely expected to be the most aggressive at deploying C-band in the coming years. As you noted a moment ago, you signed an MLA with them to facilitate that deployment. Now at a high level, it kind of sounded like there were a few primary components to the deal, like a term extension, establishing some pricing ahead of time, streamlining the application process, some volume discounts if they hit certain thresholds. So I guess, first, a big picture question related to that MLA, would you expect that the revenue you collect from Verizon in the fullness of time under that deal is going to be much different than what it would have been under an la carte model?

Yes. I don't expect it to be materially different. Because of some of the volume incentives, they have the opportunity to pay a little less than they otherwise might have. But at the same time, they'll be incented to move quicker, which I think should drive earlier executions and [ rent ] commencements than we might have otherwise seen. And in the limited circumstances where they may have had alternative choices for deployment, they'll now have an incentive to choose one of our sites, perhaps providing some incremental revenue opportunities. But in any case, we're using market pricing levels, so I think the end result will be similar.

Okay. Since you mentioned it, can you expand upon the volume discount component at all? And the reason I ask is because the way that I've historically thought about the model is that most towers just don't have competition within a relevant radius. So if a carrier needs your location, they kind of need it and a discount wouldn't necessarily be that important to their decision-making process. So maybe expand upon the thought process behind the volume discounts and how that's likely to play out.

Yes. I mean, in most cases, your understanding is correct over the long term. However, volume incentive structure can influence prioritization of new sites, and I think it facilitates kind of a long-term working partnership between the 2 companies. And so when you think about those other components of this agreement as well, that also kind of go towards that. I think when you think about it holistically, it sort of makes it easier for us to kind of do business and greases things into happening quicker. And so ultimately, if you're getting it earlier, in many ways, you're actually getting revenue that had you not gotten it at that point in time, it's just kind of found money in a way, right? So that's really the -- I think, the biggest difference.

Okay. Okay. As it relates to the term extension, do you feel like you gave anything up in exchange for that? And again, the reason I ask is because tower churn tends to be so low that I've always wondered if term extensions are that valuable, at least if you don't expect the customer to be acquired at some point.

Yes. I think that -- well, the primary value, in my opinion, of the term extensions is that they sort of demonstrate that long-term committed relationship between the parties and kind of the interdependence that we have on each other for our collective benefit, in some way similar to what I just mentioned with regard to the volume discount structure. So while it might be apparent that this is the case anyway, contractually kind of agreeing to it institutionalizes that commitment for both parties. And it's obviously just -- again, it's one part of a much broader agreement that has many components. So it's a factor in that solidifying of that commitment to each other. But there's really nothing specific that I would say we gave up in exchange for the extensions. We both get a heightened degree of certainty from the MLA, which is a good thing, I think.

Okay. Okay. That makes sense. And then one of the other components was the streamlined leasing process. In years past, you guys have kind of noted that your back office has always been quick at turning things around and kind of suggested that those provisions were of limited value. Was there anything different with this MLA on that front? Or was it kind of a nice add-in for the customer?

Yes. I don't think it really changed much. We're still very confident that our processes and turnaround times are -- really have been very efficient for a long time. However, some of the agreed-upon provisions in the agreement around equipment pricing and standardized forms should make the existing processes, I think, even better, which really benefits Verizon. It provides Verizon even greater confidence in their ability to meet their aggressive build-out goals. That's really the point that we work on certain things that gives them greater confidence that they're going to be able to move as quickly as they want to. And ultimately, their ability to move quickly is good again for us.

Okay. Okay. That makes sense. AT&T also picked up a good chunk of C-band spectrum. They've noted they're not going to be quite as aggressive as Verizon at putting it to work, at least their initial plans. As I think back over the past couple of years, they've done quite a bit of work as it relates to FirstNet and then AWS and WCS spectrum. That's all starting to wind down a bit. Is it your sense that AT&T's C-band related activity is likely to just backfill for the FirstNet activity that's starting to fade? Or in your view, does it seem like it's likely to drive an incremental improvement in a holistic sense?

Well, I don't think it's really FirstNet-specific. AT&T has been active with all the deployments that you mentioned, and some of those are definitely reaching their latter stages. But their C-band activity is likely to keep their overall activity level fairly stable from what it's been the last couple of years. There's still FirstNet work to be done and other needs that AT&T has, but I think C-band is really the main driver of the next cycle of activity with them. So I would think relatively stable in terms of the overall level, but that's not bad because they've been very busy.

Okay. Okay. So maybe not a benefit, but it's holding them at what was a very appealing level for you guys.

Yes, I think that's right.

Okay. Okay. I guess more generally about C-band, there are kind of some conflicting opinions regarding the propagation attributes and how that may or may not be workable in some less densely populated areas in particular. What have your engineering folks, the propeller heads as Mark affectionally calls them, kind of concluded regarding how pervasive those C-band deployments are likely to be over time?

Well, we -- based on what our folks are saying here and looking at it, we expect it to be very pervasive. We basically -- the way I'd summarize it is we agree with what the carriers have said. AT&T and Verizon in their recent Analyst Day presentations, they laid out a plan to get to over 250 million people population coverage for C-band by 2025. And when you think about that aligning with T-Mobile's goals of reaching that 250 million pop coverage with 2.5 5G by 2022, I think you've got a situation where we're going to see a very extensive build-out in that. Now C-band is the primary band for achieving high-speed 5G services nationwide. I think it's pretty clear at this point. And the operators have put a ton of money into that, right? They've spent over $81 billion to secure those nationwide licenses. So we expect to see it be deployed widely.

Okay. And any sense as -- obviously, initially, it's going to be primarily colocations on existing sites. Do you have a sense as to when you might start to see some densification or new colocations as a result of this? Or is it too early to tell?

I think that we will start to see some relatively soon, but the primary focus, like you just mentioned, is really on upgrades. So I don't really -- I think it's a little early for me to give you specific commentary on that. But based on how the applications are coming in, which include not just amendment applications, but new lease applications, I think it will be in the near term that we'll start to see that pick up a little bit as well. But I mean -- but I don't want to get too carried away on that because I think the primary focus is still going to be on upgrading the existing network. I think the densification won't be quite as extensive as maybe you're thinking.

Okay. Okay. And I guess to the extent that C-band does ultimately drive some densification, do you see that opening opportunities for developers to build towers that are competitive with yours? Or does that strike you as unlikely? And to be clear, I'm not suggesting a sea change or anything like that on this front, but asking if there might be some incremental opportunities for developers.

Yes. I think it's unlikely. In order to derive the maximum benefit of C-band 5G, it really needs to be deployed along with 5G on the lower-frequency bands, 600, 700, 800 bands, and on the same RF grid. So when you think about it, a significant amount of investment has already gone into and is continuing to go into upgrading the low-band equipment to 5G on the existing towers. So as a result, we would expect to see C-band being deployed on existing tower sites first. And once those deployments are fairly advanced, then I think you're likely to see new sites in strategic locations where either additional coverage is needed, some new area; or where further densification is required to just address high user density. However, even those sites will be a fraction again of the number of existing sites that are used. So that's why I think you saw, for instance, Verizon sign their agreement with us, and it's also why we're so confident about the future opportunity for us. I think there will be some incremental densification, but it won't be as massive as what they'll have to do on their existing base of sites.

Okay. Okay. That all makes sense. I think deploying C-band is likely to be a multiyear long-term project for the carriers. It might even take a long period of time than other bands given some of the propagation attributes. In your experience, how long does it typically take a carrier to deploy a new band across its footprint? I mean, I guess, in some ways, it never really ends since they cell split and whatnot, but maybe the time it takes to reasonably pervasively deploy.

Well, it can certainly vary by spectrum band and by carrier. We've seen that over the years. But I would estimate that projects of this magnitude, what we're talking about here with C-band, typically would take 3 to 4 years for a nationwide deployment, and that seems to be about what we're on pace for with both FirstNet and T-Mobile 600 megahertz deployment. So I think that's probably a good estimate. And in the case of a C-band deployment, it's not just new spectrum, but it's part of a new technology generation. And the carriers have historically focused their new technology deployments in the top 50 markets initially and then extended it to other markets over time. And that's likely to be the case here as well, particularly given the clearing schedule for the spectrum. So we'll likely see rollouts for the top 50 markets in the next, I'd say, 18 to 24 months, with deployments in the other markets following that. When we think about the B and C blocks of this spectrum are not really expected to be available for use by the carriers until late 2023. So all this suggests that it will be -- certainly be a multiyear rollout.

Okay. And to the extent that your portfolio extends beyond the top 50 markets, would that suggest that the -- as we think about the leasing cadence for you guys, that it maybe wouldn't be, call it, back-half loaded in that sense, but perhaps a little more even across that time frame?

Yes. I mean, we -- well, we certainly have assets that are located in all different markets. And so we've seen historically in previous generations of network upgrades and rollouts that we've seen a steadier kind of longer period of activity, I think. But I don't think that will be -- it certainly won't be back-half loaded, that's for sure I mean [ it's ] [indiscernible] now. And we have -- I think there's kind of a misunderstanding of our portfolio. People sometimes say it's rural, which is really not the case. The majority of the sites are located in the top 100 BTAs. And so we don't see that ever kind of play itself out. We see activity when others see activity as well. But they ultimately do get to all those locations, and so I would expect that it kind of will extend it. Again, that multiyear process that we talked about for a build out I think will be something that we're participating in throughout that entire period.

Okay. Okay. Let's shift to DISH, which is one of the other exciting parts of the domestic outlook. A national greenfield network deployment is kind of the dream of tower owners. Obviously, you can't talk about the exact terms of the MLA that you signed with DISH, but the dollars that they've committed seem to be pretty significant. Can you share any thoughts on whether there's likely to be upside to that MLA if DISH is successful in the marketplace and ends up deploying a network with parallels to the majors in terms of sites and coverage? Like how forward-looking should we think of this deal as having been?

Yes, it is. When we think about our deal, DISH has committed to thousands of new leases with us. However, the more successful they are and the wider and deeper that their network deployment goes, there would certainly be upside if they need more sites or they have greater equipment loads that they need on existing sites. So in a lot of ways, it's really no different than any other carrier relationship in that sense. I mean the more that they have success and the more they need to do, the better it is for us because for the most part, it's all equipment-specific type of pricing.

Okay. Was there a meaningful volume component to the deal -- or should I say, volume discounts that might influence that? Or is it not -- again, if they get to that sort of scenario, it wouldn't be a relevant consideration?

Yes. It's hard for me to answer that specifically because in their situation, it is a little bit different. They're obviously starting with nothing, right? And they are agreeing to a certain minimum volume, and so yes, that gives them some certainty on pricing. So I guess inherent in there, in theory, there's some discount, but what that number should be is not an absolute anyway. Their equipment load is not the same as the other carriers at this stage. And that's why there's opportunity that depending on their success and their need to increase that equipment load, which has been our historical experience across all of our carriers throughout history, there's opportunity for us to monetize it. We're not capped in any way.

Right. Right. On the flip side, you might argue that there's more credit risk with DISH. It's -- they're not a player today. It's possible they don't gain traction in the marketplace and so on. How do you take that risk into account when assessing the deal or considering the opportunity cost of space on your towers?

Well, based on the steps that DISH has been taking, we're confident in their long-term viability as a customer of ours. It's not that commonplace where we have multiple carriers requesting the same location on a tower site at the exact same time. And once it's contracted for, it's no longer available. So we're not holding space for anyone, unless it's already been committed to under a lease agreement. And given the significant structural capacity of our sites, we've had very limited opportunity costs or I think we have very limited opportunity costs if we face any credit issues with any customer. So it's really not an issue, in my view.

Okay. Okay. Based on your interactions with that team, do you get the sense that they're ready to go and are unlikely to face meaningful deployment hiccups that could kind of throw off the timing of revenue recognition?

Yes. I mean we've been impressed with DISH's progress so far. We've worked with Dave Mayo for many years at previous companies. And we're enthusiastic, I would say, about the team he's put together at DISH. So we also think that the experience and reach of our services team has been very valuable to DISH in supporting them in their network planning efforts. So we feel very good about all of that. Now having said all that, the build-out of a brand-new nationwide wireless network is obviously a major undertaking, but they seem very committed and are making good early progress. And I think I would just say, I wouldn't bet against them.

Okay. Okay. Well, that's good to hear. Sprint, you're going to be facing some headwinds from Sprint churn in the coming years. You've been pretty transparent about what you think that's going to look like over time. When you dimension the churn that you share, are you just including the revenue from overlap sites specifically? Or is there an assumption for sites that are likely to be commissioned in addition to overlap sites? And if the latter, how do you kind of estimate what that might be?

Yes. We have included both overlap sites and our best guess around other sites that might be impacted, taking into account the individual lease term end dates. The potential loss from nonoverlap sites is typically based on a proximity analysis of the Sprint legacy sites in relation to the T-Mobile sites. This is, of course, not perfect and subject to change as T-Mobile updates and modifies their plans, but we think it's a reasonable estimate at this point.

Okay. And then I guess, more generally speaking, we spoke about Verizon and AT&T. How are you feeling about the activity you're seeing from T-Mobile?

Very good. We're feeling very good. T-Mobile has been active deploying both 2.5 gigahertz and 600 megahertz. And really since the second half of last year, their activity has been extremely strong. In fact, in 2021, thus far, they've been our largest -- both our largest services and leasing customer so far this year. And we would expect, based on what we're seeing, that they'll remain very active for not only the rest of this year, but the next couple of years.

Okay. Okay. That's good. I think it was a call or 2 ago that you and Jeff argued that it would be reasonable to think you could get to kind of gross domestic growth in the high single digits and net growth of 5% or better for some period of time, subject to the vagaries of the Sprint-related churn. Over what sort of time frame do you think those growth rates can be sustained? And at a high level, what do you see those results being predicated on? What needs to transpire in the marketplace to get there and stay there?

Yes. Because -- well, first of all, because the reported growth rates are really trailing indicators and we're seeing now increased activity levels now, we're confident that the gross growth rate will increase later this year and particularly into next year. And that rate range, that higher single digits rate range should be sustainable, I would think, for a multiyear window. The bigger question is really about the net growth rate because it's obviously influenced by the amount and timing of churn, specifically the legacy Sprint churn. So as we've discussed, we think, and we've given a lot of information about that out, but we believe there will, as a result, be some lumpiness to that churn, which will cause the net rates probably to fluctuate around a little bit. But excluding the Sprint churn, we're comfortable that, that mid-single-digits growth rate, that area of growth rate is very achievable.

Okay. And again, so we should kind of think of that as being a reasonable target for the duration of, say, the C-band deployment cycle, is that a way to think about it?

Yes, again, subject to Sprint, the level of Sprint churn, which will vary in certain years, for sure.

Okay. Okay. Makes sense. Aside from Sprint churn causing it to bounce around, what would you think of as the biggest sources of risks for that outlook?

Well, just generally speaking, the biggest risk, I think at this point, is the magnitude and the timing of churn. For the gross growth rate, I really see limited risks, but it is based -- having said that, it is based on all 4 carriers generally being active at the same time for an extended period of time. If any would materially change their plans or timing, I guess that could create some risk at the absolute growth rate level. But right now, we don't see that. But it's fairly simple, right? I mean, if any of them pulls back for any reason, that's what would affect it. But right now, I don't see that being the case.

Okay. Okay. You're also forecasting a nice increase in services revenue and gross margin this year based on those higher activity levels. How should we think about the durability of services revenue into 2022 if leasing continues on the trajectory that you're talking about? Is that work kind of so front-end loaded versus leasing that it might recede? Or do you think it's going to persist for a while?

Well, we are looking at a record or at least close to a record year for us in services volumes. And I don't want to be too committal to the long term in this area of our business, but given how busy the carriers are, I think strong services volumes should continue to some level into next year. One thing that we're watching is that there will likely be some shift by our customers from site acquisition and planning work towards more true construction work. So depending on the profitability of this work, our desire for it may vary between projects, and it might impact those future volumes a little bit. We might pass on certain things we have in the past. But right now, it's extremely busy in this business, and we still have plenty of opportunities ahead. So we feel really good about it right now. But it one of those businesses that depending on what's happening, what competition is available, all that kind of stuff, you have to kind of just make sure that you're disciplined in taking the work that is value and additive.

Okay. So we should think that as you get towards installations as opposed to some of the planning work, it's probably less differentiated, more competitive, so that sort of shifted influence how your revenue and margin factor into...

Well, it shifts in -- I mean, typically, that work is actually higher margin usually.

Oh, okay.

And it's something that we have a little more control over securing as it relates to our tower sites. As you move more towards construction work, it's often not quite as higher margin or there's greater risk to the margin maybe is the way to say it and carriers' relationships with others might affect that a little bit. But really beyond -- it's -- separate from that, it's really about whether we would be interested in some of that work because it can be very -- they put a lot of pressure on their vendors and those costs to drive them down, as you would imagine, where they have the opportunity to do so. So some cases, it makes sense for us; in other cases, it doesn't. So that can affect volume. I think the point I'm trying to make is that the opportunity set is large enough with all the activity that's going on that we can continue to have, I think, very strong results there, but we may not push it to its max in order to be smart about what the returns are on that work.

Okay. Okay. That's helpful. That all makes sense. Let's turn to edge computing, a popular topic. You've acquired a couple of traditional data centers. You've got a few kind of edge data center science experiments up and running. I think Jeff has been clear that it's not something that's going to contribute to results now or maybe even the medium term, but if applications come about that require proximity to the wireless network edge, then you'd be well positioned to participate. I guess in very broad strokes, how should we be thinking about the potential upside from edge computing? Do you see it as something that could measurably impact growth if the stars aligned properly?

Well, there's -- first of all, there's still a lot of ifs, for sure, and it may or may not ever produce meaningful growth. But the potential, if it all goes the right way, could be significant. We've got the potential to gain upside in leasing revenue from customers wanting colocation in a powered and cooled environment at tower locations that we already own. So we're well positioned from that perspective. I think if applications come about that require proximity, we see ourselves as location with the lowest latency potential from the application to the mobile user or device. Today, we've done some analysis around our existing domestic sites, and we've got over 8,000 sites that could accommodate an edge data center today, and we probably have anywhere between 6 and 20 racks at each location. So that's a sizable amount of organic leasing potential, obviously, if we can realize on that. Today, we have initial customers that have colocation leases at tower edge data centers that we've deployed, and we're working certainly on identifying more and using that basically to gain experience in the space. But it's still -- I think at the end of the day, it's still really early to know for sure how it's going to play out. But given our significant portfolio, we think we need to take the opportunity seriously, which we are.

When you talk about the 8,000 sites that could potentially host an edge data center, I'd imagine that's size of the ground compound, power availability, a certain number of fiber providers. Is there anything else that goes into dimensioning that?

Those are the main points. I think the space is probably the #1 point, but all the things that you mentioned are the primary factors.

Okay. Okay. Have you reached any early conclusions about how the mechanics of the model might work? For example, would you be inclined to rent out land to somebody who built the facilities? Or do you think you'd build them yourself or something else? Any conclusions on that front?

Well, we're still evaluating all those options. It's still -- again, it's still really early, and we haven't drawn any final conclusions at this point. And in fact, I think it's best at this early stage that we should be keeping our options open because we might find something that changes the way that we feel about it. In the instance where we own the tower edge data center, we've seen that a hub-and-spoke approach probably provides for the best customer experience. Customers who collocate at the local tower edge data center, which would be bespoke basically; and have the ability to interconnect to the Internet through our metro data center, which is the hub; and where we don't own a metro data center, we can work with others, I think, to do that. But right now, that's the structure that we're kind of focusing on and spending time on.

To the extent that you've had any [ interested ] customers or customer discussions, is there a customer type or application where people seem most enthusiastic about the prospects as we think about who might actually be a customer?

I would say we're open to any number of customers now. That's one of the things we're still working on. The market for data centers, in general, is very diverse. You've got varying demands from all different segments. You've got large players that -- as we think about it in terms of edge data centers, they might be looking at -- for multiple sites for applications like video caching and something. But then you've got smaller companies that may just simply lack the in-house infrastructure to meet their IT needs and would be happy to have access to a nearby data center. So I think that the tower edge locations are suited very well for both, and they're also suited to carriers, I think, as they move over time to more of an open RAN and virtualized infrastructure. So I think there's a lot of potential and many different potential customers that we could pull from, but time will tell.

Okay. Maybe not the same but kind of related to the open RAN comment you just made, I thought it was interesting to see that Cellnex did a deal in Poland, sale-leaseback that actually involved active network components rather than just the tower. So that was a new type of deal, at least new to me. I'm not quite sure how to wrap my head around it. I'm sure it's something you guys have contemplated, those sorts of arrangements. Would a deal that included active network components, like antennas and radios, potentially be of interest? Or is that kind of a bridge too far in terms of where you think your circle of competence lies?

Yes. I mean it would depend on the situation, but it's not likely, I would say. The beauty of our business is that the shared infrastructure does not become obsolete with each new generation of wireless technology or spectrum band. The constant change and, as a result, the investment that would be required to deal with that and the challenges, frankly, to handle each carrier's specific needs through that type of shared antenna model I think make that type of arrangement a totally different business than what we do today. In addition, entry into active infrastructure requires investments in much more extensive back-office, field operations, and it carries, frankly, increased risk for maintaining service levels. And that -- when you step back and think about it in comparison to what we do, that type of arrangement really just doesn't fit our traditional business model.

Okay. So you want to say, at the -- on the surface, it might sound similar, but if you dig down, it's quite a different model.

There's a lot of differences. I mean the general concept of shared infrastructure, I guess, is the same. But when you get into the specifics, it's very different. And I'm not even sure, to tell you the truth, Nick, that how many carriers really would want that. Certainly, domestically, I can't imagine that they'd be interested in sharing active network components. I don't know what they have left at that point. But...

Yes. It would be quite a day to see Verizon say they want to share their gear with someone else.

Yes.

All right. Let's shift to overseas a bit. Most of your overseas business is in Latin America. Historically, that was a region that embraced the independent tower ownership model before most other markets. I think the last real carrier hold out there has been América Móvil, at least of size. But they were forced to hive off their Mexican towers a few years ago, and they've announced that they're going to separate their other Latin American towers probably later this year into a new public entity. Do you have any initial thoughts regarding whether that separation is going to matter for your business or not, for example, the proximity of those sites to yours, their propensity to lease them to third parties, their suitability for colocation, anything like that?

Yes. Well, it will certainly matter to some degree because Claro has a large portfolio of sites, some of which would be competitive with ours, for sure. However, Claro's impact on the colocation market, I think, will probably be muted for a couple of reasons. The first one is that when you think about it, their towers in many countries have already been kind of open for colocation. And so we've already kind of been in competition with those sites if they're near our sites already. Spinning them into a new entity, I think, changes very little from the existing competitive market. And basically, if a customer didn't want to go on towers that were owned by Claro because they were owned by Claro, I think simply putting them into a tower company owned and controlled by the same shareholders isn't really going to change their minds much. And then second thing and perhaps most important is that Claro has traditionally built one tenant sites to only support their own equipment. And they've often utilized a broader vertical footprint on the towers. They take up a lot more space. So that really, I think, makes the number of sites that are ready for colocation much less than the total number of sites that they own. So we'll see how it plays out. We're obviously going to, of course, keep a close eye on it. But we think in most cases, we offer something better for our customers.

Okay. Okay. Those are great insights. I guess Latin America and especially Brazil have also been hit pretty hard by COVID. It seems to have had more of an impact on carrier activity down there than it has here. How would you characterize the current state of the leasing environment? And what are you anticipating as the region starts to emerge from the pandemic, hopefully sooner than later?

Yes. I mean, it's -- yes, it's definitely been hit hard. But leasing thus far this year has been very strong basically in all of our LatAm markets and Brazil. I think if anything the pandemic has exposed that there's a big need for increased network capacity and coverage in many of these markets. The carriers' results have actually been better than most people expected and the carriers have been very active on our [ towers ]. We think that the need or the increased need, I guess, for network investments will continue to strengthen the case for shared passive infrastructure. And the good news is we're obviously well prepared to support that upgrade, I think, that needs to happen to [ make ] really 4G and 5G by our customers as well as them having to expand coverage into underserved areas. So even though they've had their struggles, I think the wireless side of it, the network capacity needs have actually improved. So we feel pretty good about our position.

Okay. That's great to hear. On the new build front, you typically build, call it, 300 or 400 sites overseas each year, again, mostly in Latin America. One of your peers, American Tower, has been increasing its pace of new builds globally and plans to significantly ramp in the coming years to help drive growth. Do you see opportunities to push harder in this area? Or do you feel like the current pace is kind of consistent with what the market can absorb while meeting your return thresholds?

Yes. I mean we have always -- it's part of our history and culture here, but we've always worked at growing through new builds, and we'll continue to do so. Consistent with that history, we continue to be focused, though, on building high-quality, kind of high-value sites that are in good locations that we think will deliver value for our shareholders. We definitely don't chase volume, but we're rather focused on hitting our financial criteria when we're looking at these opportunities. The competitive environment in some markets has driven down the returns on new builds. And in those markets, we either don't build or we build very selectively where we think we can get superior returns. So subject to further geographic expansion, the number of sites we build each year will likely be in the same range, I think, that we've been in the last few years.

Okay. Can you expand a little bit upon the point you made about yields and how they've changed in certain markets over time?

Yes. Well, I mean, the yields on the sites we've done today -- to date, I should say, have been very good, north of 10%, certainly. And as you would expect, the more mature sites are obviously producing the best results. I think if we remain disciplined, we would expect to continue to produce similar returns going forward.

Are there markets that you're seeing as more attractive?

Not -- I mean it depends on the specific opportunity within the market more than it is the market in and of itself because there are times at which that -- the market that's best for us in terms of new builds can shift and change. We've had great success the last 2 years in South Africa. It's been tremendous, and that's basically been almost entirely new build driven. And we've got over 2 tenants per tower already across all those sites, and they're all brand-new builds in the last few years. So that's probably stood out, I think, in terms of performance. But depending on the wave of activity at a given point in time, other markets can shift. Central America in our early days of international expansion was certainly the top performer by far. So if markets become more mature or there's different cycles, then that can affect it. But we try to be specific on location and opportunity for multiple carriers.

Okay. Okay. Let's talk about the balance sheet for a little bit. As I look back over the past few years, I feel like every time you've issued debt, I've kind of thought to myself, they can't possibly do any better. And yet every time, your borrowing costs keep declining. And it's not just the overall rate environment, but the spreads narrow, too. I think you just issued secured debt at 1.6%, if I'm not mistaken. So if you put kind of broader interest rate moves aside for a moment, do you feel like there are, call it, more fundamental opportunities to -- for your spreads to improve further? Or do you feel like you're kind of at the limit of where you can practically go?

Well, the deal that we just completed that you referenced represented spreads that were at some of the lowest levels we've ever seen. Of course, when you combine that with low benchmark rates, you get record low all-in interest rates. It is hard to imagine that there's a lot of room to move the spreads tighter. But I think the positive thing for us is that we can still reduce our interest costs further through our opportunistic refinancings. The other thing just to throw in there about our recent deal that was really great is that we were able to achieve record leverage from the rating agencies at this price. So we basically borrowed at 7x leverage, 7x cash flow leverage for that 1.6%. So I mean, if you can do that, obviously, you're in a good spot. So we're pretty pleased with the result.

I wish I could get that, personally. You've suggested in the past that as opportunities to grow the portfolio moderate and your dividend is more meaningful as a share of AFFO, you'd probably be inclined to lower your target leverage ratio a bit. Would that remain the case if rates were to stay at these sorts of levels, such that your carrying costs were very low?

Yes. Certainly, the current low-cost debt environment that we're in makes us less inclined to think about lowering leverage anytime soon. Our comments around the impact of the dividend were really focused on a time when the dividend became a significant percentage of our AFFO, which would increase the perceived risk around the availability of liquidity to address that service, particularly in a rising rate environment. But given that our current dividend represents only 22.5%, which is what it was in the first quarter, of our AFFO, I think we're a long way off from that being a consideration.

Okay. Okay. Maybe a macro question. I think back over the most -- most of the tower industry's existence, inflation in the U.S. has run below the typical 3% escalator that you have, which has been a nice tailwind to real growth. Obviously, there are concerns that inflation might tick up in coming years, whether temporary or to more permanently higher levels. Obviously, you have some protection from that in the form of your ground leases mostly being fixed as well, but are there other puts and takes we should think about with respect to the effect of inflation on your business?

Yes. Not really. Our margins are obviously quite high. We have over 84% of our cash flow margins in the U.S. And our largest direct expense by far is the ground rent, which, as you mentioned, is typically capped by fixed escalators as well. So we've benefited from our escalators for the last 2 decades, and we'll continue to see growth through those escalators, which will continue to grow our margins. But we have very limited opportunities, frankly, to modify the escalators.

Yes. But what about -- as I think ahead, let's say, we got into sort of a new paradigm for inflation. Do you feel like your negotiating leverage with respect to new business, not covered amendments or new colocations or whatnot, do you feel like your leverage would be such that you could get better pricing or new escalators on those components of the new business to help mitigate any inflation risk?

Well, the escalators are typically not modified, but for new lease and amendment pricing, we can always consider external factors like inflation. Negotiating leverage is comprised of a lot of different things, though, which they're really mostly, I would say, about the individual site location. So it's really not a shift from the way things have worked for the last 20 years. I think we have leverage based on the quality of the locations and the exclusivity of the locations that we have, and inflation is a factor to consider when setting pricing. But our positioning from a leverage standpoint is not that different than it always is, which is pretty strong.

Okay. Okay. You've consistently acquired land and extended ground leases over time. Obviously, the amount that you put to work against land bounces around from period to period, but it's come down some as of late. Is that mostly a function of competition? Or is it a function of the number of viable targets shrinking as you've done more and more work over time? Or other factors?

Yes. I mean, in the U.S., in particular, it's mostly a function of there being less and less opportunities. We've been through most of our domestic portfolio many times over at this point. Internationally, there are still opportunities, and I think we'll continue to make a lot of progress there. Competition, frankly, is much more limited now than it was a decade ago. And I think, frankly, we're better positioned with regards to our landlords and our sites than we were at that time.

Okay. Where do you currently stand in terms of the share of domestic sites that are owned or held under an easement or some other arrangement that's akin to ownership?

Today, we're -- around 40% of our domestic sites would fall into that category of owned or easements, but even the sites that are on-the-ground leases are under our long-term control. So I think when you -- the statistic that we usually give every quarter, which includes our international sites, obviously, is that we own or control for at least 20 years the land under our towers in 71% of our total sites. Domestically, the number is obviously much, much higher.

Okay. Okay. As I think about other potential asset classes, I feel like historically, you've kind of mentioned fiber overseas as potentially interesting, something you'd at least consider. I don't recall you doing anything notable on that front in the recent past. Does that remain sort of an accurate characterization of how you're thinking about any extension beyond the core tower business overseas?

Yes. I mean we -- I don't think we've changed much our view in any material way. We're still focused on kind of that focus on exclusive infrastructure assets or things that are logical adjacencies to those assets. We've generally not had an interest in fiber here in the U.S. because of the lack of exclusivity, but we have considered it -- the possibility of it in select international markets. However, as of today, we haven't really seen any particular opportunity we thought we had to have that we didn't pursue. We're only looking really at these nontower opportunities that we think that are logical extensions of our existing business and what we're good at, our skill sets, or they will benefit our tower business in some way to help support that business further.

Okay. Okay. I think we're just about out of time. I thought we might close with sort of a broad kind of qualitative question. Obviously, you've been at the company for over 2 decades, if I'm not mistaken. Maybe talk a little bit about how you feel about the position of the business, the enthusiasm you feel in the organization and kind of the opportunities ahead relative to years past.

Yes. First of all, I appreciate you pointing out that I'm old. Yes, no, I mean, I have been here a long time, I've been here over 23 years, actually. And it's amazing to me that, at this point, I can still feel as good about things as I do. Now I feel very good about where we are today. I think the organization as a whole is very enthusiastic about what lies ahead. We're entering into a window here where all of our large domestic customers will be busy with their network investment at the same time and for multiple years. Our international markets are expected to see solid growth going forward. Our balance sheet is in as good of a position as it's maybe ever been. And we have the opportunity to keep deploying capital to enhance value for the shareholders. So I'm feeling very optimistic about the next few years for sure.

Okay. Well, listen, that's a great note on which to close. Brendan, again, thank you so much for taking the time to speak with us today. I really appreciate it.

Absolutely. Absolutely. Thank you, Nick. Appreciate it.

Okay. Take care.

Bye-bye.