[Prentiss Head of Telecom Services Research] at Raymond James, and we've added the M as well for media to our coverage, so stay tuned for more on that one coming. But one of our long-standing companies that we've covered for a long time, 25 years probably, I think, American Tower. Adam Smith is here to update us on what's going on with American Tower. The format that I like to do is, since we have journalists, PMs, international investors here, my format is typically we'll do a few minutes of prepared remarks, and we're hoping the slide deck is showing up that was sent but is somewhere in the ether to show up. So we'll do a few minutes of prepared remarks, and then I'll run a fireside chat format, and we'll save some time for Q&A at the end.
And of course, we have the breakouts, which are always highly useful to people. With that, Adam, you want to just kind of talk about AMT, and if the slide deck—I see men in black showing up in the back—and unless you have the little light thing that makes us all go forget what we're thinking about, we're about to have.
What's good is it's a very easy business to visualize and talk through. It's not overly complicated, but yeah, the graphics are always kind of fun and helpful. But maybe just an introduction. My name's Adam Smith, Senior Vice President of Investor Relations with American Tower. Been with the company for the last 14 years. For those of you that don't know, American Tower is an independent owner and operator of multi-tenant digital infrastructure. Probably around 95% of our cash flows are derived through tower assets. Beautiful. Here we go. See? It's pretty basic, but it is colorful.
You got the clicker? Your clicker.
Excellent. Thank you. So here you go. So we are an independent owner-operator of digital infrastructure. Probably 95% of our revenues are derived from just your typical macro towers. We did recently expand our portfolio to the data centers over the last several years, which is really kind of through the lens of leveraging our macro tower assets to play a larger role in a 5G ecosystem, which we believe will seek further opportunities for the edge to proliferate and where we can also monetize our distributed points of presence longer term. The model is very simple. So we are a landlord. We own the galvanized steel, the vertical galvanized steel, but it's really up to the tenants to kind of manage the active elements of the network. We own the land in the US under 35% of our sites.
The other 65%, we have average remaining terms on our ground leases of around 30 years, so we have very good ground control. We're also a REIT. So what all this kind of supports is a really compelling, predictable growth path, and I'll get into some of the revenue drivers here in just a second. We're also complementing that with a pretty attractive yield of about 3.5%. So I think that combination of growth, sustained growth, recurring growth, high cash flow generation plus yield has been really attractive, and it's really, I would say, attracted a pretty diverse investor base consisting of REIT, yield, growth, value, shareholders. We continue to kind of grow that REIT profile, I'd say, over the last five years in general. If you're trying to kind of think about and visualize the recurring nature of our revenues, it's really pretty simple.
These are long-term contracts. Our average term that we sign up is usually 10 years in duration, maybe even more, non-cancelable. We do have fixed escalators in the US of about 3%, and when we go international, we typically try to seek out CPI-linked escalators to give us a degree of insulation to local market volatility. Separate from that, we drive our revenue through incremental lease-up and amendments. We call that organic new business. If you're kind of trying to visualize a tower site, an amendment would be when an existing tower or an existing tenant, say T-Mobile, is on a site, and in a 4G or 5G cycle, they'll upgrade or touch their sites, add new equipment, expand their envelope, and that's an area that we're able to monetize. The other way that we monetize is through what we call a colocation.
So if T-Mobile is on a site and Verizon wants to get on the site, that's incremental revenue. And the beauty of our business model, as you'll see in a subsequent slide, is really there's very little fixed costs associated with an incremental tenancy and very little capital intensity. So not only do we really kind of amplify the margins and benefit from that inherent operating leverage, we also are able to kind of maximize our return on invested capital because there's very little capital investment associated with that incremental tenant as well. And then offsetting that gross growth would be a level of churn. We're historically a very low-churn business in the US, and I refer to the US I mean, it makes up probably 65% of our operating profit. Historical churn is kind of in that 1%-2% range.
So when you kind of put all this together, a fixed escalator in the US, pretty robust colocation amendment growth, we've laid out a multi-year guide that kind of supports an expectation of 4-5% achieved through, call it, gross organic new business offset by a very low degree of churn. And I think you got a pretty compelling kind of mid-single-digit organic profile. And then internationally, we would look to complement that with something probably a couple hundred basis points above that in the US This is kind of a nice illustration to really kind of help you visualize kind of the beauty of the tower model. Again, usually with a day-one tower, it varies in terms of our day-one, call it NOI yields. In the US, that could be kind of low to mid-single digits.
Internationally, usually with a new build, we target kind of a low-teens to mid-teens day-one NOI yield. But the beauty of the tower model is when you add that second tenant and the third tenant. And we get pretty good disclosure in terms of all the assets that we've put into place over, we call it vintages. So over the last five years, 10 years, 15 years, how do you kind of see that revenue growth profile, that tenancy profile? And ultimately, what it should do is drive a really compelling level of return over a multi-year period. So when you go from one tenant to two tenants, we add very little costs, right? All the incremental equipment, the active equipment, is the responsibility of the customer. All the incremental equipment that you're largely seeing on the ground is the responsibility of the customer.
And we add incremental revenue, but again, add a very little amount of capital to our denominator for ROIC and really high conversion rates, usually kind of in the 85%-95% range down to our profitability. We're already a very high-margin business with probably low 60% EBITDA margins. And that also takes into consideration the fact that we've got about $1.6 billion of what we call pass-through revenue in our international businesses. So internationally, we do pass through to our customers kind of like a triple net lease, various components of our OPEX. So in Latin America, we pass through land rent, which is about 80% of our operating costs. In Africa and in India and parts of Europe, we pass through power and fuel, where we're actually the we provision and procure and manage the power on the site.
A place like Africa, the grids either are unreliable or unavailable. So we actually invest pretty substantially in primary power sources, which is a diesel-run generator, but we've actually been investing pretty heavily into more green solutions, including lithium-ion batteries and solar. But you kind of take that out, and our margins will probably be closer to 70%. But the fact that we're converting organic growth kind of in that 85%-95% range, we still think there's a pretty good runway to further amplify margins. We already have a pretty low overhead. We are a $130 billion enterprise value company, but it's supported by about 5,500 employees. So our SG&A as a percent of revenue is very low, probably around 7%. But with all that said, we're also further globalizing the business.
Bud Noel, who's here, he's the head of our US business, always uses the phrase "rationalize, standardize, and centralize." I think we've grown very rapidly over a 15-year period globally, and I think there's come a level of decentralization that comes with that, which I think was certainly necessary to stay nimble and support growth over the last 15 years. But I think there's still a lot of opportunity to kind of streamline some of our operations and overhead. And here in 2024, despite an elevated inflationary environment, we were pretty flat last year, quite honestly, on overhead costs, and we're actually taking $30 million out of the business this year. So kind of taken all together, high-cash flow business, beautiful operating model.
We've been pretty strategic in terms of building assets and acquiring low-tenancy assets over the last 5-10 years to really kind of give us a really attractive runway to further grow and do so in a highly profitable, high-return manner, as you're kind of seeing up here on the screen. This kind of gives you an idea of our global footprint. We operate in 25 countries globally. Like I said, we have a little over 5,500 employees, and we have 224,000 assets across the globe. The whole thesis for American Tower is we have an exceptionally strong anchor US business. Like I said, that probably makes up 65% of our operating profits. The large majority of our cash flows and revenue is generated from the big three carriers, so T-Mobile, Verizon, AT&T. It's a very high-credit, high-quality anchor tenants that really drive the value of our company.
But the goal is to franchise this model, leverage our contractual expertise, our landlord expertise, construction expertise, global procurement, and really kind of find an incremental value proposition by expanding to other parts of the globe and leveraging those experiences. I think what's a little unique today, Rick, is a little different than 15 years ago. 15 years ago was about giving investors access to these different parts of the globe and kind of augment that US growth that investors otherwise didn't have access to in parts of the globe where they're bypassing kind of your typical fixed wireline type of business and in some cases going to more of a wireless type of dependency. I think what you have today is investors can actually do a lot of their own diversification, right?
I mean, you could buy IHS, you could buy Cellnex, you could buy Indus and SBA and Equinix and Telesites. And I think what's extremely critical for American Tower and something we're always keen on is not only do we need to manage and actively manage the most attractive portfolio with the best terms and conditions, the best growth, the best operating capabilities that afford us accretive development opportunities, and that means power as a service. It means built-to-suit. So we need to have the highest quality portfolio, but we also need to operate it in a manner so efficiently that it can't be replicated by the aggregation of some of our regional peers. And so that means top-line growth optimized, but also converting it down to profitability in an optimized manner as well.
I think you're seeing that in what we've done over the course of last year and into this year, both through bringing down SG&A, benefiting from that inherent operating leverage benefit that's within the tower model, and really maximizing our margin conversion and expansion over time. When we look at mobile data growth, I mean, ultimately, this is what drives incremental demand for our towers. Over the last decade, we've seen in the US probably 30+% annual growth in mobile data consumption. What you see up here on the screen is our view on average monthly smartphone data usage, and it's kind of forward-looking, so 2023 through 2028. I think this is probably one of the few, I'd say, projections that always proves to be relatively conservative historically looking. The base that we're growing off of today is massive.
Even though projections might suggest a level of deceleration in terms of CAGR growth, the growth is still massive, and the base is even larger. So we've kind of gone from a lower double-digit average monthly gigabyte data consumption in the US to about 26, 27 per month. We're bringing down, or at least it's projected, to see a lower growth over the next five years. But the base is so massive that the incremental data intensity that we're actually putting on these networks is substantial. What this ultimately means is the carriers want to put their spectrum assets to work. They want to utilize the efficiency of mid-band spectrum to bring down their cost per gigabyte.
Quite honestly, for a tower, it'd probably be even more beneficial if they didn't have the spectrum because they would have to significantly densify the networks to accommodate this level of traffic on a 4G network. So we always kind of push back on this narrative that the carriers aren't monetizing on 5G. The fact is they need this mid-band spectrum to accommodate the massive data that's running through the networks today and kind of maintain an appropriate margin profile that kind of matches revenue and cost per gigabyte. But when we look at the next several years, we've always looked at the 5G cycles being one that's probably a decade long in duration. Speaking to the US, with every incremental generation, we've typically seen another $5 billion of annual investment on average. So in 3G, we saw $25 billion per year. 4G, we saw around $30 billion per year.
In 5G, we anticipate $35 billion or, I'm sorry, $30 billion per year for 4G, and we anticipate $35 billion per year with 5G. What we've seen so far in 2021, 2022, 2022, we're around $45 billion among primarily the big three plus-ish and then some of the national carriers. In 2023, the narrative has been carrier pullback, but the fact is they're still investing about $35 billion. I think estimates for 2024 are still kind of in that $33-$36 billion range, which we believe very nicely fits into this expectation for sustained elevated investment against some of the more historical generational comps. We also look at 5G, just like we've seen with past generations, as being one of kind of like a sine wave, so one of two peaks of investment.
That initial peak, which we're kind of coming off of, off of that peak of $45 billion, is coverage-focused. It's when the carriers want to take these spectrum assets efficiently but in an accelerated manner, roll it out across their footprint. That's how they kind of are able to narrate to coverage pops and kind of fill out the map when you show what you're covering for 5G. You'll see a level of grooming, and we believe we're kind of in that grooming period right now, albeit still driving really attractive growth and sustained growth. And then if you kind of look at a chart like this, as you continue to see more strain on the networks, there's going to be an incremental need for more capacity. So peak number 1 is coverage, peak number 2 is capacity. And I think we're already seeing tea leaves, right?
So we're seeing green shoots, we're seeing a level of activity pick up. Our pipeline is picking up, albeit off of relatively low volumes kind of during that trough in the back half of 2023. There's still a tremendous amount of work to be done on our tower assets. Probably roughly a little over half of our assets have been upgraded to 5G at this point in time. So we still think there's a tremendous runway ahead of us. And I think you're already seeing the tea leaves in terms of people jockeying for position for capacity-focused investments. You see that with CTIA's white paper. They're kind of the group that advocates for the carriers. They're suggesting we need another 400 MHz of spectrum. There's probably 600 MHz of mid-band spectrum in the market today, a little bit more of more higher frequency, higher band spectrum.
But CTIA is suggesting we need another 400 megahertz by 2027 and 1,400 megahertz by 2032. You're seeing T-Mobile acquire Comcast spectrum assets for the 2028 timeframe. You're seeing the Biden-Harris administration run a feasibility study on incremental spectrum that ultimately will be required to accommodate the applications and high data intensity applications of the future. And I think from where American Tower sits, we can monetize in one of two ways or both. To the extent the carriers and enterprises get the spectrum assets, it's a good monetization event for us, primarily maybe more from an amendment perspective. If they don't get the spectrum assets, it's going to require significant cell splitting and densification, which is collocations. So we see a really sustained, attractive runway of growth.
You kind of look across the board in where we operate, you're seeing approaching 20%, over 20% annual CAGR anticipated across these markets in which we operate. So we think there's a really attractive runway to support sustained, attractive organic growth that, again, converts in a really attractive manner down to our bottom line. To help visualize the data intensity, what's really kind of helped ramp up this incremental intensity that we're seeing across the networks, I mean, it's easy to just kind of visualize what's really putting strain on this network. We're always asked, "What's the new killer app?" Quite honestly, I think we're sitting with the killer app right now. I think what we have at our fingertips in terms of mobile data streaming is really adequate to further push or facilitate that incremental step from 26 GB to 62, 63 GB.
You're already starting to see a conversion of higher resolution video among 5G users versus 4G users, the amount of time that they're spending in enhanced video, the amount of data that a 5G user is using versus a true 4G user. So again, we're really bullish. I mean, even just kind of thinking about how do you get from 26 GB to 62 GB in a month, it really only requires a bit of transition from, call it 480-pixel video to 1080-pixel video to 4K pixel or 4K HD video, 8K HD video. So I think there's a very attractive runway, and I think a lot of it's even at our fingertips, even without taking into consideration us all doing this from the metaverse with Apple Vision Pro headsets a year from now, which would be an incremental case.
On this slide, we're just kind of illustrating all these data demands that we've seen over the last 10 years has really translated to attractive growth. We've grown our revenue in EBITDA to CAGR of 13%, and that's converted at a pretty attractive level of profitability at the AFFO level, which is adjusted funds from operations, which is really kind of our critical cash flow metric. And we've done this through three key ways at the revenue line. I think one is organically. In the US, we've averaged, even while taking on Sprint churn, which is the result of the decommissioning of Sprint when it was acquired by T-Mobile, we're still growing in the US around 5% on average over the last, call it, decade plus. Internationally, we're above that. We're a couple hundred basis points above that in the US
And then we've complemented that through very select and accretive M&A and built-to-suit opportunities. So we've probably acquired around 160,000 sites over the last 10 years during this period. We've complemented that with the built-to-suits or new builds of around 40,000-45,000 sites. So we've really kind of executed on building up a really sustained, attractive growth profile that we believe can go on well into the future with the leading mobile network operators across the globe. And then obviously supporting this strategy in a level of predictability and sustainability is our investment-grade balance sheet. And that affords us really diversified sources of capital, attractively priced sources of capital, which really allows us to kind of maximize the accretion of the investments that we're making. Just the last slide, I'll just touch on this quickly.
I mean, everything you're seeing in terms of what we've executed over the last 10 years and what we're doing over the next 10 years is really kind of centered on what we call our stand and deliver strategy. There's four key pillars: enhancing our operational efficiency. This is getting into the globalization of our business, finding opportunities to, as I put it earlier, rationalize, standardize, and centralize, finding opportunities to further drive margins and profitability, grow our assets and capabilities. So again, one of our key differentiators across our footprint is leveraging shared knowledge across our operations and being the best partner that we can be for our customers. And that should come in the form of maximizing organic growth, but also our development opportunities. Invest in platform extensions.
We do believe there's an opportunity to leverage our distributed points of presence to play a larger role in a 5G and beyond ecosystem. That's why we do CoreSite. We believe that our 43,000 plots of land in the US and the fact that we largely have significant ground control, we largely have adequate power, adequate fiber, and appropriate parcels of land distributed across the key metro and suburban areas across the US could ultimately solve for that improved compute and round-trip time to facilitate the latency-sensitive use cases of the future. So that's one of the key areas we're focused on today while we manage CoreSite. And then augment industry leadership, so being a trusted partner for our customers, playing a significant role in the communities that we serve, and advocating for the neutral host model across regulatory and stakeholders across our footprint.
I tried to get through it all, 25 years of history, in a couple minutes. Everybody knows we wrote that first tower report 25 years ago.
I know. I think you cut the ribbon at one of our first slides, so I appreciate it.
But speaking of ground control, you're replacing Major Tom Bartlett. At a time when SBAC has a CEO changeover, Crown Castle is doing an executive search as well. Interesting time in the tower industry to have that much turnover for what's a very stable business. What should we think? And Steve was there in Deer Valley this January. Appreciated him being out starting to talk with people. But how should investors in the room think about what's going to change or what's not going to change with a new CEO at American Tower?
Yeah. So for those of you that haven't met Steve Vondran, I mean, he's an incredible asset. He's been with American Tower for almost 25 years. He's a lawyer by trade. He operated our US business prior to assuming this role for the last five years. So he's been a part of the executive leadership for some time. And even extending beyond when he took the head of the US business, he was a critical part of the leadership team. And I always refer to Steve Vondran even before this move, but I'd always refer to Steve as kind of the IP of American Tower. I think I would challenge anybody to find someone that knows how to extract value out of a tower more than Steve. He's virtually written every Master Lease Agreement in our US business for the last 20 years.
He's got tremendous credibility with our customers. He's got tremendous credibility, obviously internally, among the employees at American Tower. We've also been getting him out there pretty proactively, even preceding this announcement, with our investors because he can obviously speak both to the CoreSite side and obviously the US Tower side and has tremendous insights into how you extract tower both in the near term but also sustainably over a long-term period. The vast majority of our investors are long-term investors, so they really appreciate that perspective. The fact that Steve's been on the leadership team for some time, and he sat there with Jim Taiclet and Tom Bartlett, I don't think anybody should really anticipate any major changes in terms of how we operate.
Steve did lay out a set of kind of priorities, and they kind of stem from how we manage the US business for the last five years. I mean, one is maximizing organic growth and making sure we have contract structures that facilitate a level of monetization that we think is appropriate and adequate for the tower business. That means monetizing amendments and real estate rights, collocations, attractive escalator terms, but really monetizing organically and making sure we have the right quality assets in our portfolio to drive the sustained growth that I think investors in towers and infrastructure really command. Two is being very selective in terms of where we put that next dollar in terms of capital deployment. I think here in 2024, a lot of the focus for us is accelerating that pathway to balance sheet strength. So we are investment grade. That's non-negotiable for us.
We want to get down to our 5x leverage limit by the end of 2024. We've kind of been hovering above the 5x since doing the CoreSite transaction at the end of 2021. But we work very closely with the rating agencies to make sure our financing plan and our pathway to de-lever is really supportive of the investment-grade credit that we have. But with all that said, we do want to accelerate the pathway to get down to a level of financial flexibility because we do look across our footprint, the opportunities that we have in front of us, the opportunities that we anticipate into the future, and a level of flexibility to further buy back stock or other options that could be at our disposal. We look forward to the opportunity to accelerate to get there.
Being very selective with our CapEx, we've brought down the intensity over the last couple of years, $2 billion in 2022, $1.8 billion in 2023, and our guide called for $1.6 billion here in 2024. But one of the beauties, I'd say, of our CapEx program, $1.4 billion of that $1.6 billion is really truly discretionary. And I think one of the attractive elements for American Tower, and I'm sure we'll get into this at a time when we don't necessarily find M&A attractive, we have a very nimble CapEx program where we can put that next dollar to where we think we can drive the best risk-adjusted rates return. This year, it might be a little bit more heavily weighted towards CoreSite, just given all the demand that we're seeing within that ecosystem and that platform.
But we also balance that with deploying capital towards built-to-suits across the more emerging markets as well. So I think that's really critical area number two. And I think critical area number three, and it kind of feeds into that, is just making sure we continue to strengthen and reinforce our balance sheet as a key asset. Again, M&A is not important today, but I think we kind of look at where assets have been acquired over the last decade and whether or not those could come to market over the next 5-10 years. And we want to be in a position to strengthen either to execute, buy back stock, or if attractive CapEx opportunities arise, being able to execute on those as well.
One more. What lessons did you learn from India now that you're getting closer to exiting?
Yeah. I got to check the time here. But I mean, it's all.
Yeah, two minutes. Hit it.
Okay. So we entered India in the 2007, 2008 time frame, Rick. And I think at the time, what you're looking at is a market with 12-15 carriers. It was ripe for consolidation. It was a market that had fixed escalators kind of 2%-2.5%. We obviously seek CPI-linked escalators in the vast majority of our markets that we operate in outside of the US It was low barriers to entry. It's a low-cost opportunity to build towers. And I think when you combine that with a rather captive tower, meaning the carriers largely own the tower companies, and low barriers to entry and low cost to deploy networks, those were certainly the challenges that we knew at the time.
But you also look at the market and the data growth that we're seeing across the footprint, a billion-plus people in terms of population, a very data-hungry demographic. And we saw a lot of opportunity. But I think admittedly, it's not a dog-eat-or-home-earth situation. There's a lot of lessons to be learned. Some are unfortunate. Some we can be a little bit more proactive in how we take those learnings and deploy capital. I think, one, we're not going to bet on consolidation. I mean, it's not a secret, I think, where we've probably deployed a lot of our capital in the market has ultimately kind of resulted in a level of turbulence from a counterparty perspective. We did acquire Viom in 2016, and 2016, and Tata exited the market shortly thereafter. And we've acquired the Vodafone Idea assets in 2018.
And obviously, there's been a degree of collection challenges there as well. We were probably naive in thinking we could change the escalator construct given the barrier to entry and the low barrier and cost to deploy networks and captive tower element. But it's not something we're learning today. We're obviously going through the process of selling India. We signed the deal at the start of the year. We think there's a tremendous opportunity to redeploy that capital and drive better sustained growth longer term. But it's learnings that we've actually been employing in our capital allocation over the last five-plus years. And you look at Telxius and the fact that we were seeking out CPI-linked escalators, tier one counterparty and M&O, developed markets with a little bit more stability, markets that have largely gone through the consolidation process. That's not a coincidence.
So even though we're kind of going through the process of, I would say, reoptimizing or right-sizing the risk profile of our portfolio and actually positioning it with better quality of earnings and anticipated growth longer term, that's happening now with India. But the lessons learned from India and how we've actually taken those lessons and deployed capital, that's been going on for years now.
Yeah. We're out of time, but I want a one-sentence answer. Dividends are flat from 2023 to 2024. What should we expect longer-term dividend growth can be?
Yeah. So we're modestly growing it. I think longer term, the expectation should be, and what we look at internally, is growing the dividend with AFFO. And I think the way we've kind of looked at the AFFO algorithm plus a 3.5% yield, I think, is a really attractive TSR and one we remain committed to. And I think the dividend continues to be a really strong piece of that TSR.
Great. Let's wrap it here. Appreciate the time. We'll go down to the breakout session. Thanks, everyone. Have a good rest of the conference.
Thanks, everyone.
I've got the microphones off.