Crown Castle Inc. (CCI) Earnings Call Transcript & Summary

I'm Phil Cusick. I follow the communications services infrastructure and media space here at JPMorgan. Thanks for joining us. I want to welcome Dan Schlanger. Dan joined Crown Castle in April of 2016 and is currently the Vice President and CFO. Dan, thanks for taking some time with us today.

Thanks for having me.

Yes. So I have questions. I am more than happy to have people stand up and shot questions out or there's a microphone that's going to go around, if you need to shout it, I'll repeat the question. But it's always better if we get them from the room. Dan, we've heard a lot about the industry going into a multiyear build-out cycle for 5G. But it seems difficult to reconcile that with lower carrier CapEx plans in the future. Can you help us bridge that gap?

Sure. Again, thanks for having me. We haven't been able to make it to this conference before, so I really appreciate the time. In every upgrade cycle that we've been through as part of the wireless industry, there's always going to be ups and downs even within that cycle of when our customers spend capital on certain things. Fortunately, for us, our customers very consistently spend capital on what is the most important thing for their network, which is deploying spectrum to get out to consumers so they can provide the best level of service and best quality of network they possibly can because that's how they compete with each other. So even in cases where we see overall CapEx spends coming down or there are different periods of up and down between the customers, we generally see a pretty steady level of spending. And in the 5G cycle, we believe that it's going to be a long cycle where there's new network that has to be deployed, but also new infrastructure that needs to be deployed in the form of small cells. So as CapEx moves between customers, I think it was somewhat of a canceling sign wave. Some may spend more while other spend less, and that kind of evened out for us as a generally upward trajectory. And then even for Crown Castle specifically beyond just towers, because our customers start making choices between spending on towers and spending on small cells to get the network deployed where they need it, we have the other benefit of even if spending on towers comes down, the spending on small cells might come up. And again, it kind of evens out over time to where we have a pretty consistent level of growth. And I think that's one of the best characteristics of the tower business overall is just the consistency of the growth. We are not looking to grow faster and faster every year. We're looking to grow consistently over a multiyear period to drive significant returns for our shareholders. And that, I think, is a little different than what you see in the fluctuations that you're talking about from some of our customers saying they may come down off a peak, which is fine. It may not be the peak for us. But as long as we maintain a really solid level of growth, we think that's what drives the most value. And in the 5G case, because we are, as an industry, going to see a significant densification of the network, as John Stankey said earlier in the conversation with you, densification is going to happen. We believe there will be a longer cycle of more CapEx and growth in the network, which should really benefit Crown Castle.

So maybe a follow-up there. We saw the 4G cycle start in sort of '08-'09 and went for a long time, how is the 5G cycle going to be different and maybe densification is one of the angles there?

So I think that is one of the differences. I think the 5G outcomes that we're going to drive will require a densification and distribution of the network that wasn't necessarily the case because in 4G, a lot of that densification happened on existing towers. There wasn't a huge infrastructure build out of actual physical assets. It was an infrastructure build-out of putting additional antennas on existing assets to drive what was a great wave of growth for all of the tower industry. It also drove a significant increase in the productivity that we all experience in our phones. And that was what ultimately is the driver of our business is the wireless data demand in the U.S. drives our business. And because our customers, the carriers, were able to provide a much better level of service so that we can all say connect to an Uber in seconds, if you went and try to do that on a 3G network, that would have been what a 45-minute ordeal and it never would have worked. So that kind of leap of benefit that the consumer gets is what allows then our carriers to spend money. They did that on towers, which is generally the case. What they do is they start with what is most familiar. And in the case of going from 3G to 4G, that was going on towers, they already had some sort of equipment on what we would call an amendment. So we amended an existing tower with an existing lease and add more equipment to it. And then they would move into building on towers, they weren't yet on, which we would call colocation, a new customer on an already existing asset, both of which are great for us. They're both additional revenue on an already existing asset that we don't need to pay a significant capital for. So the incremental returns are really high, which you can see as the yields of our tower had increased over time. But that move from what they know, the amendment into what they don't know as well, which is the colocation was the pattern of 4G build-out. What we're seeing now is the same pattern in 5G. Start with towers, start with amendments, move into colocations, but now there's a third leg of that, which is just an extension of that model, moving into small cells. So the customer knows -- the customers know their towers the most, where they already exist. They know the propagation characteristics and then the throughput and the data traffic. So they put more where they know they need it, and then they move into the colocation. And now they're moving into the small cell. And you can see that exactly with what happened with our business that in 2021, we saw all of our customers shift their focus into towers as they own mid-band spectrum or have gotten licensed to mid-band spectrum they wanted to deploy. That caused us to pull down some of the deployment expectations around our small cell business because they had allocated capital differently than we expected and what we had in place. That is continuing into 2022, high levels of tower activity, a lower level of small cell activity. But during the last 15 months or so, we've gotten more orders than we had in any other period. So we got -- in essence, in the last 15 months, we've gotten orders for 70% of the small cells we've ever gotten orders for, 50,000 of them from AT&T -- not from AT&T, from Verizon and T-Mobile. And what that speaks to is this pattern of focus on the towers and move focus on the thing you know most to the thing you know least, in this case, towers to small cells, and they knew that small cells took 18 to 36 months to build. So even in the midst of spending a whole bunch of money on towers, they knew they had to sign up a bunch of small cells. And that's just what we believe is the beginning of the densification in the U.S. and that over the course of 5G, which will take longer because we are building new infrastructure as opposed to adding assets to -- or adding equipment to existing assets that, that will be a longer cycle that will ultimately really benefit Crown Castle given our position.

I'm curious, you talked about from that they know best what they know least. As we've seen carriers deploy higher frequency spectrum on networks that were designed, so whether it's 2.5 or 3.5 gigahertz, on networks that were designed on a 2 gigahertz grid, what we heard a lot of going into this was, yes, but those goods are so dense that we won't have to worry about it. And yet Verizon, in particular, I think, is finding that the quality is not quite what it used to be as they deploy their 5G at 3.5. Have you seen, not any particular carrier, but carriers, in general, come back and sort of push hard on the densification in pockets that maybe you didn't expect a year ago?

I guess the answer to that question is yes. We are seeing carriers focus on densification, whether it's in pockets we didn't expect is a harder question to respond to. What we expected and what has happened is our carrier customers focus first on the markets where the most people are because that's where they get the biggest bang for their buck. So the top markets in the U.S. is where the most density is, the most data traffic is and they focus on those markets. What we thought was going to happen and what did happen was, first, focus on central business districts, places where most people gather and congregate for most of the day and then move out towards the suburbs as people went -- there are 2 places people generally are. They are either home or at work. And that's where you're going to need wireless connectivity and the densification happens and the wireless demand happens, and we need -- we, as an industry, need more densification in those areas. And so we expect it to move from central business out to suburbs. That has happened. What is now happening is our customers are also asking for outside some of the top markets because they're seeing data density in those markets start approach levels where the tower model is not sufficient to serve the density, and it's the physical attributes of towers that make that true. They can't get close enough together in some cases. So when the tower markets first started, when the network -- the wireless market first started, the best tower was a 300-foot tower where you could go on the top of it, put an antenna and send the signal out 5 miles because that entire 5-mile radius would be covered by the amount of spectrum. As more and more people use more and more data, the people on the last mile of it didn't have much spectrum left because the people within the first 4 miles took all of it. So it's like a water hose. It goes all the way to the end. If we poke enough holes in the beginning, it doesn't get to the end and that's what happened. So the -- what was cell splitting was moved to a tower, that was closer together. So 2.5 miles and then 1 mile and then 0.5 mile, and we're in the cities and with 0.25 mile or 0.5 mile radius of most towers. Even though the spectrum can still go 5 miles, there's no way you can make it there and still have capacity in it to carry data. So the densification happens because you can't get towers any closer together without interference and you can't get cloud more towers built in the areas because of regulation. So that's where small cells come in. Towers are still a great business because they are still the most efficient way to deploy network. They're just insufficient for the amount of density that is coming and that we see in most of the top markets. What we found interesting is that there are other markets that our customers are already seeing that density come and are responding with small cells, which is exactly what we expected to happen. It's just a matter of the timing of when it happens. Right now, the biggest question in our head still is, how quickly will the overall industry shift into more and more small cells because right now, we still feel like we're really early on. We're kind of just beginning this densification process. And how quickly that ramp happens is really the question that's the hardest for us to answer.

So let's go down that path a little bit. You have bought fiber or built fiber in a bunch of markets. And a couple of years ago, you put out a pretty detailed sort of 3 or 4 markets that were examples of how the fiber and small cells sort of came together. As you see carriers look at densifying, number one, are you doing more in those markets to sort of prepare? And number two, are you looking -- how are you looking further afield? Are there markets where we say, okay, we're close here or we're -- this is a really interesting place to go? What's the view today?

Our analysis basically starts when a customer asks us to go because we're not going to go speculatively into a market. That's not our business, and we don't believe that that's a good use of our capital. Because if we make that choice, even though some period of time in the future, there may be small cells, if you build assets that are not used for a period of time, that doesn't help your return. So we're always looking to make a return over and above our cost of capital. The way we believe we can do that is to wait for our customers to say, "Hey, we want to go into this market." As we do that, we look and see, do we believe there will be colocation, some additional customer to come in and put small cells in the same market over a reasonable time frame that we would make money because we don't make money when we build the first greenfield anchor tenant system. No shared infrastructure company should make money on that first one because then what's the point then the customers would do it themselves. So we don't make money in the first one. We need a second one to start making money, a second tenant, a second customer on the same asset base. We book and see what type of data demand we expect to happen, what the tower market looks like in that area and whether we think small cells are required and then we make a determination whether we think that's something we want to build. Going speculatively is not something that we anticipate doing because we don't think that it provides a good enough competitive advantage just to build fiber anywhere because you're not exactly sure where the small cells are going to go until somebody goes and maps out where the small cells are going to go. Once that has happened, though, the other customers tend to follow because if you think about what they're doing, they're densifying their network in order to provide a better network service to a consumer, think of yourself walking down the street and you have AT&T, your friend has Verizon, your other friend has T-Mobile on their phone. And one of them is better than the other, and you look and say, "Hey, your network is so much better than mine. I wish I could get that. Who are you with?" And you start switching. And that's what they're trying to avoid. So if one customer has built out a small cell system, say, in Boston, and you're walking down the street and the other one doesn't feel as good because you try to get a connection and you can't, then the other one is going to say, "Oh, I have a network problem there, what's happened? Oh, that other company has built small cells. I got to look at what they did." They may not act that day. We can't predict their prioritization of when they act and where they act. But we know over time, they're going to act because they can't let their network quality suffer in relation to their competitors or they will lose customers.

Maybe a different way to think about it. As we look at DISH, for example, building out a new national network, they've talked about you as not just a tower provider, but a fiber provider as well. Maybe talk about how that discussion went and the advantage of having both pieces rather than partnering with another fiber provider?

The majority of the fiber that DISH was looking for was to connect their towers, not to build small cells. So they were looking to build what they would call an open RAN structure, in essence, putting less equipment at the base of the tower and more equipment at a centralized hub and, therefore, you need to connect via fiber connection that is stable, dedicated, easy to use. In order to build that infrastructure, they wanted a kind of what they would call a partner to have both the towers and that fiber so they had 1 person to go to and say, "Okay, did my network work there? No. That's their problem. Crown Castle calls them and fix it as opposed to try to piece that together. And that was important to them because they actually wanted to announce all that at 1 time, so they could say, look, this is the easiest way to get all of these towers deployed and all of this network deployed. And that was where we came in, and nobody else had that combination of assets to offer, capabilities to offer to say we can do that and we can help you build your network, no matter what you want it to be. Now I think the side effect of that, that will happen over time is when DISH needs small cells, which they will need, they don't need them upfront because, again, towers are the best way to deploy spectrum over large areas, large population areas that when they need small cells, they're already using us and using our fiber and have confidence in us, and we will likely get those small cells as long as we're reasonably priced. Still can't overcharge. So we're always looking at how do we balance all of those things to get as much business as we can, and we think we're really, really well positioned for DISH and for any other customer at this point.

Just going on the small cell angle a little bit, you talked about the 50,000 that are coming. Remind me how many of those are sort of overlaying a 5G radio or antenna on the same small cell or a fairly similar location versus new densifying locations from here?

Not sure at this point. We need to go through a conversation with our customers because what they did is they've signed up to buy those or contract with us. They've already contracted for those 50,000. We need to cite each 1 of them and build them. Of the 35,000 we signed with T-Mobile, what we said the majority of which is co-located on existing assets. What they've said publicly is most of those are upgrades, which is what you're talking about, going -- adding 5G on top of 4G or something like that's what I believe they mean when they say upgrade. So I would say our comment of the majority is co-located in their comment that a lot of them are upgrades would speak to, there's probably a fair amount that they believe will be 5G upgrades on existing assets. To us, that's unimportant because we make the same return on it. We still -- that's still a second tenant on an existing asset, that's adding revenue that is not yet present on that asset to drive our returns over and above our cost of capital. What is important over time is to get both that and the densification happening because we need -- as a country, we know we need more of each. We need more densification and we, as Crown Castle, need more asset colocation to drive value. So as long as both of those things are happening in concert, which is exactly what's going on, we feel really good about the business model. It's pretty rare that a customer will come to us and say, "Hey, this is a 5G deployment on top of a 4G deployment and write that out in a contract." Like that's not usually how our business works. You ask me, is Verizon putting 5G up on some of our towers? The answer is maybe. We know that they're putting their C-band radio up there. How they're connecting it back into the network and whether it's going to be 5G is actually a direct question for them than it is for us.

Okay. So on that not particular carrier, for example, but it's all about return, not a rate card. It's not, hey, this is $1,000. It's about every individual location is different. Is that right?

Yes. We do not price to a $1 of revenue per month rate card on small cells. We price to try to drive a return because the build cost in every city is totally different. Build cost within a city is totally different, which is fairly different than how towers were grew up. Towers grew up where we bought large portions or portfolios of assets and they were negotiated on a national basis because we bought them all at the same time. So even though the build cost may have been different, we basically said, look, well, even though it's the ones that are sitting in 1 area may be more expensive than the 1 sitting in another area, overall, we'll make money because we can look at that and see the portfolio. We're not doing that with small cells. We're building them individually. And to try to charge a specific price for those individually would not work out very well for us or our customers. So we're much more targeted in how we think about the pricing.

So how does that -- I hate to be predictable. How does that return calculation change as we go from a 0 inflation environment to something higher for however long?

When we put our models together, we know we are going to be -- we plan to be the last owner of the asset, which means our planning horizon is really long term. These assets are 40-, 50-, 70-year assets. As we have made those investments over the last 10 years, we've always anticipated and put into an estimated in our models, a more normalized cost of debt. So we never put in 0 cost debt or 1% cost debt as part of the planning. We always went back to what historically had been the 10-year average over a very long period of time. So we had building into this an increase in interest expense every time we made an investment decision. So now we don't have to change our return thresholds because it was already there. However, if interest expense goes way up above historical norms, we might have to think about that. But I think our thought would be, just like we've built in that increase, we would build in the decrease at some point because that also won't last. So I think what we'll do is likely leave it somewhat -- our return threshold is somewhat similar to where they have been unless we see a longer-term reason to increase the debt side. But given the returns that we look at and the cost of our capital, the vast majority of our capital is equity based because you look at our -- the way we trade, call it, 25 times EBITDA, we have 5 turns of leverage. That means 20 turns of equity in there. So the equity cost is the more important cost to us. Then the question becomes, what do you bake in for your equity cost the same thing. We had an increase in interest expense, and therefore, because we're a dividend player, but we had anticipated an increasing equity cost. I'm not saying we're going to go sell equity for this reason. I'm just saying that that's how we think about returns. Again, if those interest expense -- if that interest expense goes way above what it has historically been, we'll probably have to reconsider. But at this point, we're not there.

So nothing that's happened in the last 6 months or year is changing the way you think about that whether it's a tower or a small cell initially pricing and return characteristics?

No. I think what I would say is that the increase in interest expense happened faster than we could have anticipated. I think that's true every person that has thought about this over the last 9 months. But it hasn't changed the outcome very much. It hasn't changed the end where we think it's going to end up.

Okay. Fortunately, my entire tower career, I've been watching carriers say that they're going to go get lower tower prices, right? Verizon in 2008, AT&T and Sprint over the last sort of 5, 6 years, none of them have been successful. And so as we go into a position where the escalators that were set up 20 years ago at 3% and your lower small cell escalators start looking like they're in line to or maybe lower than inflation that doesn't change the way you look at the business, I'm surprised.

No, it doesn't. So let's start with the tower side. The 3% escalator was born out of the underlying cost structure of our business predominantly rests in the ground rent under our towers. And that ground rent had an escalator in around 3% because real estate generally accretes in value over time and the landlords wanted an increasing rent payment to them. We then asked for -- that from our customers and got it because they would have to pay it if they own the tower and got the ground rent even. So it was basically a thought of as that's a parity that you can't get away from if you're the deployer of a network like our customers are. And that 3% has held steady through ups and downs in inflation precisely for that reason is that our underlying cost structure has held steady at 3% regardless of inflation. So we don't get squeezed, and we don't get outsized returns. We believe the stability of that growth that is dictated by a constant escalator is more important than trying to eke out incremental returns in good times and getting less returns in bad times and having volatility in the growth, that ultimately ends up at about the same spot over time. So we believe for us and for our investors that understanding that long-term stability is really important, particularly given when you think about the terminal value of this business because at some point, you have to assume some terminal value. And because like I said before, these are extremely long-term assets, it's important to think what that is, and we believe it's really important to have that 3% escalator in there to show real value over a really long time. Because with a 3% escalator in what has historically been 1% to 2% churn, that's a really unique business to be able to show terminal value growth with no significant investment. So we believe that, that has worked really well for us. We also believe the exact flip of that for our customers work. They know over time how much they're going to pay -- and it's not going to be more when things are inflationary. It's not going to be less when things are less inflationary. They know how much they're going to pay, to pay for towers over the next however many years, our contracts run and likely longer because that's kind of the norm in the business. So we have not tried to renegotiate those nor would we because we really like that stability. The same is true on small cells. It's just that the fiber portion of the small cell doesn't have escalators in them and the small cell does. Now we don't bifurcate the contract, but that's how the negotiation ultimately started and went. So our small cell escalators are closer to 1.5% because you still get it on the small cell, but not on the fiber portion of the business. So we believe all of that makes sense in an inflationary environment and that stability of growth is really important. And our ability to withstand inflation is really high because we don't have many costs tied to inflation. Like I said, the vast majority of our costs are the rent under our towers and the cost on small cells is the deployment cost, which has labor and, therefore, inflation in it. But we can make up for that as we price. That's part of what we do. We look at it and say this is build cost, here is how much we think is going to cost, here's obviously we want to make out of that from a return perspective, and we can change that over time.

Okay. Let's switch gears a little bit. Edge data centers is something we've talked about a lot over the years. You have an investment in Vapor IO. I thought Charlie Ergen at his Analyst Day, a couple of weeks ago was interesting, he sort of threw out there that they have an edge data center at every cell site is basically 1 rack, but it's a start. And so is there any view on how you price going forward that shifts with that 1 rack there and the potential that, that grows over time?

I don't think what DISH is doing has allowed us any insight that would give us real confidence in how we're going to price edge data centers. We always charge for rent on the ground under our towers. And how much we charge is based on a lot of concepts. And as DISH would expand, we would make more money because they would take up more ground rent. Or if other people were to expand, other customers were to expand, that data center or that the edge presence, we would make more money. The question is how? Do we want to own the data center itself and deploy it and make a return on all of that capital and just have the racks being rented out? We're not going to get into running like owning the equipment inside, that's not our business. We're passive infrastructure only. Do we just want to lease the land to a data center company? Do we want to put out land to any customer who wants to come? Those are all questions that we don't know yet enough about how the market is going to develop that to have a real solid answer. That is part of the reason that we invested so early on in Vapor was try to figure a lot of that out and it's helped a tremendous amount to have Vapor. They are very good at what they do. They understand the market extremely well. We've learned a lot from them, and they have a very good form factor that we think will be successful in the market. And when we figure out how all of that plays together, we'll figure out the pricing mechanisms. What we believe is happening and what was good about what DISH said is it's true that there is reason to have data processing and storage capacity at a tower. That's validation of how we think the market is evolving because DISH is doing what we think other companies will do. We already aggregate some data there anyway. Why wouldn't we put a data center there to try to parse through all of the interconnections that would happen. And if we, as an industry, can start consolidating that to where traffic already is existing, which is at a tower site, we think that's really good for tower businesses. We think we're extremely well positioned for that, not only because we have 40,000 towers and, therefore, associated land. We have the fiber that's going to connect all of those data centers, which is going to be important because instead of building each of those data centers to be redundant in and of itself, there's no way that 1 rack, if it gets hit with lightning is going to be able to stay running or if there's a flood there that it may or may not, depending on whether it's elevated. But if 1 goes down, you can just shift the traffic to another. And the only way you do that is using the fiber connectivity to run these as a system as opposed to individual islands of data centers. So this is a different architecture, we believe, that is going to be developed for edge data centers and develop for large data centers where they have to be in and of themselves redundant. So the redundancy is going to be a system-wide redundancy, which requires fiber, which puts us in a better position than any other company because we're the only company that has tower sites and fiber. And then you add to that small cells where we know more about where the traffic is, and I think we're best positioned for whatever edge data centers are going to be. But having said all of that, the only way any of that's really important is if there's a person willing to pay for it, there's revenue associated with having an edge data center. And we believe the only way that revenue is going to be associated with edge data centers is that there's a low latency use case that requires data to be stored closest to the customer because otherwise, why wouldn't use the existing data centers. If 30 or 70 milliseconds is good enough, those are really efficient. The only way you move towards the edge is getting really low latency, which means there has to be a use case that is going to be a wireless mobile use case where data has to be processed and stored that close to a consumer. The only way you do that is right back to small cells. You can't get there without having way more small cells, orders of magnitude more small cells than what we have now because that low latency has to go wireless at some point because elsewise, what are you building? There's no use case that you're coming up with, that's going to be all wired and needs that low latency because then you can just use the existing data centers again. It's all wired. So we think that this is a great setup for Crown Castle, not only because the edge data centers were best set up, but you've got to go through way more small cells to get there.

I want to follow up on 1 thing you said there, and then I think we're going to be out of time in a minute. But you said Crown Castle is a passive infrastructure company. And you have some active sort of fiber side. But for the most part, you've stuck with that. I'm curious as you see American and Crown buying data centers and other sort of infrastructure companies coming into the tower space, do you think that 5 years from now, Crown is still a mostly passive infrastructure company? Or does it make sense to -- and focus on towers and small cells? Or does it make sense that these are sort of broadening industries over time into the digital space?

We believe that passive infrastructure is our competitive advantage and that owning the active infrastructure electronics doesn't add a whole bunch. Now we have a fiber business where we need to do that, and it's a good business that we're really excited about the potential for. But it is our view right now that, that passive infrastructure is where we can drive a return over and above our cost of capital, better than any other company. So relatively best for Crown Castle. We are always looking for what is going to come next or next after that. And as we look at the market now and the landscape that we see in front of us, there could be other areas that might require more active participation. We do not see that as the next step for us. But as we evaluate and see how the market evolves, I'm not cutting off the opportunity that we may do so. But it is definitely not our current path, nor is it one that we're strategic that we see a lot of value in because what I talked about is how -- basically, how do you densify and distribute the network more and more. We believe Crown Castle is well positioned for that and has a lot of growth available to just continue doing what we've done so well over the last 25 years, which has put money into assets that will be leased up by our wireless customers and by other customers who will ultimately need connectivity to a consumer. And we don't see any reason to deviate from that as of right now.

Okay. I think that's a good place to stop. Thanks, Dan. Thanks, everybody.