Good morning, everyone. I'm Ric Prentiss, Head of Telecom Services Research here at Raymond James. Welcome to day two of the 44th annual Raymond James Institutional Investor Conference, my 27th of these institutional investor conferences. Those of you that know, I wrote my first tower report in January of 1999. We've been around the tower space literally from the birth, the boom, the bust, and the rebirth, and the continued movement of the tower space. Full schedule here with tower companies, digital infrastructure companies, telcos, satellite companies, the whole gamut. I wanna welcome Adam Smith, Senior VP of IR. He's gonna give us a few minutes of prepared remarks, we'll go into a fireside chat, discussion, save some time for your Q&A, we'll pop downstairs for the breakout. Adam, welcome.
Good seeing you again.
Yeah. No, it's great to be here. Thanks, Rick. Thanks everyone for joining. Yeah, I'll just spend a couple of minutes, Rick, I think maybe five minutes, 10 minutes, just kinda giving an overview. The beauty of the tower model is if you don't know the model, you can all become experts in two minutes, and that's part of the beauty of it all. We are a owner, operator, and developer of communications infrastructure, largely tower assets. We are a REIT. We have made some modest investments in fiber.
More recently, at the end of 2021, and I'm sure we'll spend a little time on it a little later on, made an investment in the CoreSite data center business, to really kind of advance what we anticipate to be a pretty meaningful opportunity on the mobile edge back to our tower sites over time. The model is really quite simple. We own and operate the tower site. We own the galvanized vertical steel. We manage the site, the security, the monitoring. In the U.S., we typically have long-term control of the land. You know, we own probably 35% of the land under our towers, and the other 65% are pursuant to long-term leases. You know, probably in and around 30 years.
Very good, very good ground control. Really, our customers are, you know, the MNOs globally. The Verizons, the AT&Ts, the T-Mobiles, and they're really responsible for the active equipment. Think about it as the antennas, the cable, the base stations. We typically sign up here in the U.S. long-term, non-cancellable lease agreements. We actually have about $60 billion-$62 billion of non-cancellable backlog right now globally. The leasing model is really pretty simple. You know, there's probably three big components to organic growth. We have a contractual escalator in the U.S. that typically averages on a fixed basis around 3%. Internationally, with the exception of India, it's largely linked to CPI, so we get good inflationary protection on that front as well.
The other big part to growth is the co-location and amendment contribution to growth. Co-locations meaning being able to add a new tenant to an existing site with a new antenna. Amendment-driven revenue is really augmentation of our site, upgrading for 4G, upgrading for 5G, adding new equipment to your envelope, all of which we monetize, and all of which really converts at the bottom line at a very high degree. The partial offset to growth could be churn when a customer either consolidates or exits a market, in rare cases, decides they don't need the footprint or potential renegotiation. This has been a really strong, sustainable growth model. Over the last decade, we've averaged organic growth globally, kinda mid-single digit ballpark.
We've complemented that with the construction and the development of a very accretive and active built-to-suit pipeline as well, largely international. This is a function of building scale positions across key geographies across the globe and demonstrating capabilities with our customers, and we've developed about 40,000 towers, you know, really with a day one tenant NOI yield of, you know, low double digits, call it like 12%-14%. As we kinda get to the next slide, you'll really kinda see how the economics. This is an extremely simple slide, but it's illustrative. When we add a second tenant, when we add a third tenant, really the cost structure, the beauty of the model is the costs are largely fixed.
Incremental tenancies really convert down to the bottom line, kind of in and around 85%-90%. Our margins as a company are over 60% when you take out various costs that we actually pass through to the customer and recognize revenue for. It's kind of like a triple net lease type structure that we have internationally predominantly, where we actually pass through in and around 70%-80% of the direct expenses to the customer. When you take that revenue out, we're probably over 70%. It's a very high margin business. When you add in organic growth, that converts down to the bottom line at, call it an 85%-90% clip. You know, we really seek to further enhance those margins over time as well.
You know, one of the things we really like to look at globally is what we call vintage tower analysis, and looking at the sites that we've owned for 20 years, 10 years, 5 year, 1 year, and what are the type of returns that we're generating on those assets. I think this is really illustrative of how we look to expand those returns and margins over time. You know, an international tower build probably has a day one yield, like I said, of 12%-14%. You add a second tenant, that probably goes to upper 20s. You add a third tenant, it probably goes to upper 30s.
It's really kind of you know, magnifying the conversion rates down not only to our EBITDA margins, but really to the type of outsized returns that we can generate on those assets over time as well. Over the last decade plus, we've really built scale to a high degree across the key geographies globally. I joined American Tower in 2010. At the time, we had 4 markets and you know, probably 25, 30 thousand sites globally. We've gone through a period of significant expansion over the last decade. We now operate in 26 markets. We have over 220,000 sites. The large majority of our cash flows do come out of the U.S. tower business, probably 60%.
Now we also have a U.S. data center business that probably contributes about 6%, and our European business probably in and around 5% to operating profits. Kind of think about developed markets as defined by Europe, U.S., and Canada as kind of making up, you know, probably over 70%, and then complementing that with different international markets that are at various phases of network rollouts, in a lot of cases, even bypassing your typical fixed line infrastructure to roll out wireless connectivity. We very much look at how we've deployed capital in our market selection to really be, you know, complementary in terms of driving different type of growth vectors, capitalizing on various rollouts really across the global footprint.
I think this is a good illustration of how we've kind of done some level of market expansion to really capitalize on that, on that growth in terms of data consumption globally. If you were to go back over the last five to 10 years, we've seen CAGRs of data consumption really grow, you know, 20, even close to 30%. We very much anticipate ongoing significant data consumption looking ahead over the next five years and capitalizing it on a variety of different markets that are, again, at various stages of data consumption, but also growth potential.
If we look at what's really driven this exponential level of data consumption and demand, I mean, like I said, you kinda go back to the start of 4G, and we were playing Snake, and we were sending emails and texts, and that very much evolved into, you know, exponential data demand consumption when you start introducing mobile video streaming. You start thinking about, you know, as we look ahead, autonomous vehicles, mixed reality. These are all the industrial trends that are really driving a significant level of not only upgrades to sites, but really a high degree of densification has driven a very meaningful level of co-location amendment activity, but also internationally, a high degree of building out infrastructure through built-to-suits as well.
I think what gets us really excited as a company is I think there's still very much a lot of work to be done here on the 5G initial upgrade cycle. When you look at video streaming, I mean, you're talking for standard definition, high definition, probably single digit megabits per second in terms of consumption. When you get to, call it 4K, that's probably 25 Mbps . When you get to 8K, you might be looking at 50 to 100 Mbps . What gets us really excited is when you start thinking about the max specs of mixed reality, augmented reality for the networks of the future, you start thinking the Metaverse and, you know, thinking about the resolution, the enhanced resolution, the rapid refresh rates that are ultimately gonna be required with those type of use cases.
You're talking gigabits per second in terms of data consumption, which very much we believe is gonna require not just overlays to the sites for 5G here that we're seeing, you know, over the next couple years, but a significant level of densification cycles that ultimately will drive a high degree of co-location on our sites over the long term as well. How is this translating into growth for American Tower? We've got a very strong track record of delivering double-digit annual growth really across our three key financial metrics, including property revenue, Adjusted EBITDA, and how that's flowing through down to attributable AFFO per share. We do have an investment-grade balance sheet that has afforded us a diversified sources of capital, being able to deploy capital in an accretive manner over the last decade.
We have grown our portfolio. I'd say since the start of 2012, we've added about 160,000 sites to our portfolio through M&A. Our scale positions and capabilities have afforded us the opportunity to construct around 40,000 towers over the last decade. That scale, and that presence, and that credibility with our customers has really helped drive the real attractive growth rates on the organic side that I mentioned. We're complementing our appreciation and our growth in AFFO per share through a dividend. Again, we are a REIT. We probably distribute this year in and around 60% of our AFFO in the form of a dividend. We've grown our dividend annually on average by about 20%.
Like I said, we've laid out a 10% expectation here in our guide subject to board approval for 2023. We've also distributed about $1 billion, a little over $1 billion through a share buyback as well. You know, operating in 26 markets, we do have a lot of different options to deploy capital in an accretive manner. Our investment-grade balance sheet gives us a lot of flexibility to do so. Our diversified sources of capital, whether it's private investors in Europe, with the data center business in the US, access to various pools of capital, both here in the US and in Europe, in a high liquidity position has certainly afforded us, you know, a lot of really attractive opportunity to drive shareholder value. We've been guided, kind of thinking about the strategy.
We introduced a Stand and Deliver framework about sorry, 6 years ago. It's really centered around 6 key pillars, or sorry, 4 key pillars. The first being driving operational efficiency. Again, trying to seek opportunities to become more of a global business, franchising that tower model across our global footprint, you know, streamlining back-office functions, looking at drone monitoring, looking at our cycle times and our leasing processes, really finding opportunities to not just sustain those really attractive flow-through, or conversion rates of revenue down to the bottom line, but even enhance it. The next key pillar is growing our portfolio and our capabilities. This very much centers around continuing to evaluate strategic and attractive macro tower assets.
At the same time demonstrating our capability to our MNOs and our global partners across national partners, our capabilities in terms of providing a value proposition back to them for their network. That means built-to-suits, it means our power as a service. I think that continued demonstration of our capabilities has very much accelerated our opportunities to deploy capital over the last five years. Like I said, we've built around 40,000 sites lifetime to date. Probably 26,000 of those have actually come in just the last five years alone. That's a function of building scale. It's a function of demonstrating, you know, our capabilities of a company.
I think it's extremely important at a time where, and I think we've been rather vocal about this, we're having a tough time making the math work on M&A that's currently out in the market. We are focused on deleveraging, so that's, you know, that's fine. Being able to continue to deploy at a time where we might not find M&A to make financial sense, the ability to build scale through an accretive internal organic CapEx pipeline, building, you know, call it 4,000 sites a year. We built close to 7,000 last year with day one NOI yields of 12%. Being able to execute on a really attractive development pipeline through CoreSite. These are all really attractive opportunities that we have that might not be afforded to other companies, just given our scale and our capabilities globally.
Our focus on platform expansion. What this gets centered around is how can we play a larger role in 5G. We are very bullish on the leasing that we see over the next five to 10 years just through normal course of leasing and the propagation characteristics of 5G and what that ultimately will necessitate in terms of densification and co-location. Also trying to look at our core competencies and capabilities as a company. The fact that we've got 43,000 distributed points of presence here in the U.S., 180,000 sites internationally, good ground control, fiber and power already at the site, M&O relationships. How can we play a larger role?
That really gets to the genesis of CoreSite and an interconnection-rich cloud on-ramp rich ecosystem that ultimately we think over time could be distributed out to our points of presence and address the low latency use cases of the future. It's certainly, you know, the mobile edge is, and I know we'll talk about it here shortly, Ric, so I won't get too much into it. It's very much evolving, but we very much believe we're positioned to play an even larger role in an accretive manner over the long term. Finally, enhancing industry leadership. This centers around being very active with NGOs, government bodies, our key stakeholders, the MNOs, to really advance mobile broadband globally.
Also doing our, you know, executing our growth in a sustainable manner, being a good corporate citizen and executing on various pillars of our ESG strategy as well. You know, this has not just been a framework, Ric, that we've used to kind of guide us over the last, call it 5-10 years. It's a framework we'll certainly implement going forward. You know, I think, to summarize the secular demand trends that we're seeing in wireless remain as strong as ever. We've rolled out a guide for 2023 with really strong fundamentals that we think are sustainable.
I think the way we've positioned our portfolio globally, not just with a diversified footprint, but bringing in an interconnection-rich asset like CoreSite very much positions us to not just execute on 5G from the traditional sense for a tower company, but a much larger role as well.
Great. Appreciate the prep, Adam. Obviously a long history of growth. As we think about that guidance you've given for 2023, and then looking even beyond that, how should we think about, start with the US, kind of leasing revenue growth rates? You've got gross, you've got some churn, 5G. You've also talked about in the past, maybe less than half the sites or maybe half the sites have been touched. Elaborate a little bit more on kind of what the medium-term growth rate might look like.
Yeah. We've laid out a pretty specific longer-term guide, Ric, for what we think organic growth is gonna look like in the U.S. You know, I think if you were to go back to the 2015 through 2020 timeframe, we put up very strong numbers in terms of organic growth, kind of in the mid approaching upper single-digit ballpark. What we've laid out here in 2023, which is supportive of that next five-year guide that we laid out, which calls for at least 5% organic growth, we are absorbing some Sprint churn, but we've lapped the largest part of contracted Sprint churn over the last couple of years. We do have a pretty good run rate going forward, but it's at least 5% organic growth on average over the next five years.
That goes to at least 6% when you strip out the Sprint churn that I think is pretty well known by the market at this point. We very much have lapped that big peak. I think when you look at the components of our growth, you know we have a 3% escalator, right? Maybe let's think about it in the context of the at least 6% organic growth absent the Sprint churn. You know we have a 3% escalator. Our historical churn rate in the U.S. absent Sprint is typically kind of in that 1%-2% range. Those two things net to, you know, call it 1.5%, right?
To get to 6, that calls for that colo and amendment growth to kind of be in that 4-4.5% range. 2022, we were at 3.4%. The year before that, we were probably in around 3%. We've got a pretty bullish expectation for a sustained and accelerative level of co-location and amendment leasing. I think what's a little more unique to American Tower is a lot of that is backstopped by these comprehensive MOAs that we actually have in place with the carriers. For those that don't know, I apologize, I'll probably use holistic and comprehensive MOA interchangeably. It's the same thing. We're under comprehensive MOAs with Verizon, AT&T, T-Mobile, and Dish. What this framework really represents is they'll get a set of real estate rights.
Through those real estate rights, absent, separate from the escalator, their monthly run rate will increase by a specified amount at a specified date, and it's really a baseline. You know, it's a contracted minimum. When we look to 2023, we've laid out a guide of approximately 5% for the U.S. Again, it's greater than 6% absent the Sprint churn. We're setting a record level of new business contributions of about $220 million. It's around 5%. You know, the 5G cycle is very much in swing. We're very much monetizing and great pathway to growth, but probably over 90% of the growth is locked in.
You know, I'm certainly not looking for a scenario where everyone puts the shovels down and stops what they're doing. The fact is that could happen, and I think our growth over the next several years is largely intact. Now granted, you know, some crazy event could alter that, but contractually, that's what we'll do. We look ahead to 2024, maybe the piece parts are a little bit different between those different categories, but we see that largely supportive as well. If we look over the next, call it, 5 years and what's kind of supported that guide that we've rolled out, probably 75% of the growth is locked in.
Things that could be upside is certainly densification initiatives over and above what we've kind of underwritten, which we very much believe will be there. The potential for new entrants is certainly something we haven't underwritten, you know, that's obviously, I think as the ecosystem in 5G unlocks new use cases, I think a requirement or maybe a need for distributed points of presence could grow, and we could monetize on that as well. Long and short, we're really bullish on the U.S. growth rate going forward. Again, a lot of it's really locked in through a contracted minimum.
Great. If we go to the bottom line, attributable AFFO per share growth rate, good history there too. Obviously, feeling some pressure on interest rates. You guys did some refinancing recently in kind of the mid-fives. Where are you seeing the marketplace on interest rates? Also that kind of long-term view of attributable AFFO per share growth rate.
Yeah.
Give and feel?
Yeah. We in terms of predicting rates, I'll get out of that business. I think, you know, obviously, we're feeling a pretty high degree of pressure in our guide for 2023. We'd like to think a lot of that is really one time in nature because we're probably absorbing about 8% headwind from financing costs in our guide for 2023. Probably 6% of that is from floating rate debt. Again, I'm not gonna predict rates, but absent a similar acceleration in the LIBOR that we saw over the course of 2022 and started the year sub 1% and rose over 4.5%, we would expect to see, you know, some level of stabilization probably in that headwind going forward.
With that said, we do have probably roughly 20% of our debt is floating. You know, we do like a degree of floating rate debt. Typically, it has afforded us access to a shorter curve, but, you know, typically cheaper debt. It's obviously pretty flat right now across the curves. I think the other thing for American Tower is we're very much in a deleveraging mode. We did acquire CoreSite and brought our leverage up to probably around 6.8x I think everyone in the room can probably assume whenever we do a deal like that, we're in lockstep with the rating agencies. We've come to a very agreed-upon deleveraging path that's supportive of our investment-grade balance sheet and our investment-grade rating.
We'll certainly target and we're committed to getting back down to the 5x, sub 5x level over the next couple of years. I think when you have a deleveraging commitment, you know, certainly having access to a pool of debt that's pre-payable, penalty-free, and opportunistically terming that out is certainly attractive to us. We did that with the Stonepeak proceeds last year, even that incremental $570 that we closed on in the back half. Quickly, you know, being able to prepay and pay down debt is very attractive. We're also a very high cash flow business that will opportunistically look to term out debt as well.
I think in terms of the long-term growth for American Tower, at the AFFO level, I think looking at how the algorithm's coming together in 2023 is pretty illustrative of what we think to be pretty sustainable core fundamentals going forward. You've got the 5% organic growth in the U.S. We would anticipate something higher for the international business. We are above that modestly with the international business in 2023, albeit we are absorbing what we think to be peak churn in Latin America. Given the operating leverage of our business, that converts at a very high degree down to the Adjusted EBITDA level. If you were to take a look at our guide, you know, I laid out the organic growth pieces. That's converting 85%-90% down to EBITDA.
Our core EBITDA growth is over 7%. When you keep maintenance CapEx flat like we're doing in 2023, what you end up getting is, you know, combined with accretive development opportunities, I think over time, as we continue to get more financial flexibility as we get back down to our targeted net leverage range, having additional opportunities to deploy accretive capital maybe in the form of a buyback. When you add those things together, you know, we do see a very sustainable level of core growth and kind of that upper single digit, you know, unlevered AFFO ballpark. I think relative to probably how we underwrote the longer-term target a couple of years ago, admittedly, the curve is significantly higher. We'll certainly see pressure there.
I do think we're probably taking the brunt of that here in 2023.
Yeah. You talked about the edge, the CoreSite acquisition. People are confused by that. The acquisition, but also just the edge, right? I mean, it's just when is it? Where is it? How is it?
Yeah.
Help people understand what should they be looking for to figure out when it's coming.
Yeah. I mean, the whole concept of our perspective of the edge is and if you were to go back even 5 years ago when Jim Taiclet was our CEO, and it's consistent with Tom Bartlett today, we've always thought our distributive points of presence can play a larger role. We have the M&O relationships, we have ground control, we have power, we have fiber. I think what we very much found through trials that we've done over the last, call it, 5 years, in some of the smaller, you know, trial edge data center investments is, if you can't solve for the ecosystem, bringing interconnection to the site, then your value proposition is kinda limited.
I think I mentioned it last night, it's kinda like owning a piece of land along the highway and just hoping a tower ends up there someday. You know, we very much wanted to be more proactive in understanding this opportunity. Again, it's leveraging existing assets. It's leveraging our ability as a company to redevelop and maintain distributive points of presence. It is very much a core competency of ours. I think what we certainly found is we had to solve for the interconnection. We had to have a more strategic value proposition to a partnership or, you know, our relevance with the CSPs and the MNOs as this opportunity unfolds. If you go read the tender offer documents that are out there, our engagement with CoreSite actually started talking about a partnership.
We've engaged with many data centers to evaluate, is there a partnership to be had? I think with any partnership, you know, you certainly kinda run into capital allocation priority differences. I think data centers probably want, you know, for as long as they can, to kinda keep the compute capability within their core data centers and maybe, you know, not relinquish quite as much upfront towards the opportunity for aggregation edge or the mobile edge. What we saw in CoreSite, and I would very much emphasize, not every data center would have made sense for us.
It needs to be kind of the unique characteristics of CoreSite, which we see as a reasonable cost to provide the scale, kinda that ecosystem, interconnection-rich, cloud on-ramp-rich platform that ultimately we can extend or evaluate to extend out to these distributive points of presence over time. That's why we execute on it. I think for us, it represents a reasonable cost option to meaningfully enhance our opportunity at the edge. We're already, you know, Tom Bartlett, our CEO, Steve Vondran, our president of the U.S. business, they now have direct commercial relationships with Amazon, Microsoft, Google. I think you're already starting to see a collective recognition across the footprint or across the industry players with MNOs partnering with CSPs on their MEC strategies.
I think you're seeing a collective recognition that ultimately this is gonna be an infrastructure that's gonna be needed because it's going to have to solve for the ultra-low latency sensitive use cases of the future. I think, you know, maybe the core data centers today can facilitate at the wireless edge 100-millisecond round trip times, but the use cases of the future might require 5, 10, 15. Continuing to push out these distributive points of presence closer to the end user, I think will be critical. You know, for us, again, this enhances our seat at the table. We could get to the point where it might not make economic sense. We might decide we need additional strategic or financial partners. We have a better seat now to really evaluate that.
That's why you do this deal. We underwrote the deal on a standalone basis. We didn't bring in the mobile edge. We saw CoreSite as a unique opportunity because, you know, I think it had been largely underinvested. We're helping, we're giving support. I think under the American Tower umbrella, customers are embracing it as well because I think there's a better runway for investment. You know, CoreSite represents a 6% operating profit today for us. There isn't an expectation to bring that to 20%, 30%, 40%. It is really a unique opportunity for us to grow the CoreSite business, really kind of enhance that option at the edge over the long term.
Let's see if there's some questions out there. Leslie.
Good morning. I was wondering, do you have any updates on the thin solar panel films that we've seen in India a few times over the last few months?
The ESG question about thin solar panels and...
Yeah. Not really at scale. I'm sorry, what was the name?
Leslie.
Hi, Leslie. Not really at scale. I mean, we very much are evaluating a lot of different models like the solar film that you're talking about in India. Whether or not that can actually be a scalable solution that we can obviously run towards our 75,000 sites that we have in India. I think as you probably know, in markets like Africa and India, generators are kind of the primary source of power because either the grid is largely inaccessible or unreliable. We're very much evaluating both in Africa and in India what these solutions could ultimately look like. There's nothing really at scale in India with regards to what you're referring to.
However, we have deployed cumulatively, probably $350 million between largely Africa and India in terms of what we kind of consider to be, you know, eco-friendly investments. Largely solar panels and also lithium-ion batteries. That's, you know, that's probably the larger driver to our Scope 1 committed reduction, greenhouse gas reductions. I think in a place like Africa, we've deployed solar panels to around 15,000 of our sites. We've deployed lithium-ion batteries to about 18,000 of our sites. If you were to go back to, call it like, 2017, when we really started to first roll out these initiatives, we've probably reduced the average fuel consumption per site by about 35%. It's a fantastic investment for us.
It probably, to be quite honest, Leslie, probably exceeds even a BTS return, but it's the single largest driver towards achieving those Scope 1 emission reductions as well. You know, between what we're doing in Africa and then doing various pilots like what you're talking about in India, it's a, it's a critical key focus. I would say probably what we're doing in Africa is more at scale right now, obviously, doing different pockets of trials like you're talking about in India.
All right. We'll wrap it there. We've got the changeover. We'll be going down to the breakout rooms.