All right. Welcome, everybody. Thank you for joining. I'm David Barden. I head up communications, infrastructure and telecommunications research for the U.S. and Canada for Bank of America. I'm kind of moonlighting here at Jeff's Global Real Estate Conference, but I do have the pleasure of being here with Adam Smith, who heads up investor relations and other finance duties at American Tower. Adam, thank you so much for joining.
Yeah, great being here. Thanks.
Do you have any safe harbor that you need to rattle off?
Nope, I don't think so.
Okay.
I think we'll get right into it.
Perfect. So I want to start off by congratulating Texas Christian University and the Horned Frogs, for their overwhelming underdog victory against Long Island University, 45 to nothing. How did it happen?
Yeah, no.
How did you guys do it?
I can go into this in depth.
Okay.
But, but we are tied right now for the best record in college football. We're two and oh, so I will-
With Wisconsin.
I will take it. Yeah.
With Wisconsin. Okay, this is a long time rivalry here, so, So I guess what I'd like to do is to-- Jeff is asking us to ask all the companies, a few questions about the general kind of demand environment around real estate, and I know American Tower is not your typical real estate company, but so I'll try to couch this question: Do you think that sector M&A will grow when the Fed cuts, yes or no?
Um, yes.
Do you think that that would happen in the second half of this year, or the first half of next year, or maybe the second half of next year?
I don't know. I guess from where I sit today, sometime next year.
Okay. Would you say that demand for tower space is increasing, steady, or decreasing?
I think this year we've seen it increasing for us, 2024 versus 2023. And what I would say, and I think we'll obviously take a double click into this, is based on the conversations we're having with our customers, I think there's a very constructive dialogue that gets us pretty enthused as we look ahead towards the densification phase.
Perfect. And then last year, the majority of companies said that they expected a ramp-up in their AI initiatives in twenty twenty-four. How would you characterize your plans, and your outlook?
I'm trying to work myself out of the business, so I think they'll just put a machine in to replicate me at some point, but.
Me too.
But no, I mean, American Tower has been ahead of the game on that for a while. I mean, we've not just AI recently, but we've actually been investing pretty significantly over the last five-plus years in terms of data quality, digital twinning. And I think, quite honestly, it's given us a bit of a differentiated value proposition back to the customers that's been really appreciated, and we'll look to expand that and enhance it. I mean, we have a lot of different offerings that solve for a lot of complex solutions for our customers, whether it's power in Africa, backup power here in the U.S., but also, you know, utilizing data to create a best-in-class experience for our customers on the leasing side.
So I think there's a lot more still to be done, but it's an area we're definitely focused on.
I think that that's interesting because, you know, I think a lot of people think about the tower business and one of its strengths as being its simplicity.
Yeah.
But when you dig in, you realize how not simple being a tower company is. All the contracts, all the data points, all the elements of the SLAs and the agreements that you have. I mean, there's a gigantic data lake inside these companies that I imagine, you know, could lend, lends itself to an idea of being automated and greater efficiency.
You know, and it's. And I'm not just saying this as we come up on comp season, but, you know, Steven Vondran is the IP of American Tower. I've been saying that for a long time, and you know, Steven's been with the company for 25 years. I'd challenge anybody to find somebody that knows how to extract value from a tower better than Steven Vondran. And when we talk about growth for the tower business, and specifically for American Tower, you know, I think when you look at our new business on a per-site basis, you see pretty attractive growth for a sustained period of time. You also see growth when you go from 4G incrementally to 5G, probably about a 10% increase on average.
It's not so much just about the contracts that we've signed over the last four or five years. It's also the contracts that we signed ten years ago, and fifteen years ago, and twenty years ago, and making sure that we have flexibility to monetize on every incremental generation of investments. It's... You know, I get the question a lot: "All right, we're under comprehensive agreements with two of the Big Three. What's going on with the third? Are you gonna get a deal done? Aren't you at a disadvantage? What's going on?
You know, is this a sign of friction? You need to be extremely careful and clear as to how you think about not just what your exit rate is with these agreements in year five, six, seven, four, whatever the duration of the contract is, making sure you're retaining a set of real estate rights that allows us to monetize when we get to 6G, 7G, 8G, and so on and so forth. And I think that's. We've done an exceptional job of that. I think our US leasing team and Steve have done an exceptional job of that. And, yeah, it does get complex, and to adequately evaluate what the opportunity cost is of going a la carte, you need a substantial and comprehensive data set of information.
You know, we've invested in drone flights that take millions and millions and millions of pictures of all these different assets. We've got an extremely comprehensive data set that we can draw on to evaluate, all right, you know, what on average is the customer equipment across all of our customers on every one of our sites? What are the envelopes? Where is there available space? Where do we think ultimately they need to kind of upgrade to adequately propagate 5G or 6G, 7G? That all goes into the decision-making.
It's not quite as simple as just saying, "All right, you know, P times Q, this is what we ultimately want to agree to." You gotta make sure you're retaining a set of real estate rights for the long term, and I think that's, you know, for American Tower, that's, that's the bread and butter.
So before we jump into the business, remind us, you know, there's a debate out there. I think that we all maybe know that whoever wins the presidential election kinda doesn't matter to the tower industry. But, you know, what if there is a recession? What if there's hyperinflation? How did the tower sector respond to these types of variances in the economic climate?
So I think admittedly, and what we've seen over the last three years, it'd be a bit naive for me to say we're not overly sensitive to the rate environment, right? We do have long-term cash flows. We do have predictable cash flows, and you know, movements, quite honestly, in the interest rate have a material effect on our discount rate that you're discounting those cash flows back to for value. As it relates to inflation, you know, we have a set of contract structures globally, and I kind of break it down between the U.S. and that of international. In the U.S., we do have fixed escalators, kind of in that 3% fixed average range, which you know, honestly, as you can probably imagine, has served us exceptionally well over the last 15 years.
What we saw over the last couple of years was obviously seeing the inflation kind of exceed that of what our fixed escalators are. There isn't, like, an opportunity at the time to really renegotiate that, and quite honestly, I don't think we'd get a lot of sympathy from the carriers to come back based on the last decade plus. And I say there's not flexibility because those are terms that are underwritten in the underlying Master Lease Agreement, the ones that we signed in twenty seventeen, eighteen, nineteen, that go for 10 to 15 years. So I think I get the question a lot, would you want to renegotiate to CPI? For us to renegotiate those, you're probably bringing new things to the table to renegotiate, to open that up.
And I think over the long term, we still think 3% is a very compelling term, and one that we have fought to maintain, and one we're gonna continue to fight to maintain over the long term. Now, in the U.S., we mitigate a lot of that because probably our largest cost element is land rent in our direct expenses. It probably makes up around 65%-70% of our direct expenses in the U.S., and that also has a fixed 2.5%-3% escalator term. So we have a pretty good matching mechanism when you get down to the margin in terms of how we, how we counter that term at, on the top line. And then internationally, India is certainly an outlier.
India, we did have fixed 2% escalators, kind of a contractual norm in the market. To be quite honest, naively, we thought we probably could have changed over time and weren't able to. But those are the type of learnings that very much shaped the focus on executing CPI-linked escalators in other parts of the globe. And with the exception of France, and there is a piece of Nigeria leases that are pegged to USD, that have different types of escalator terms, virtually the remaining international leases are all tied to CPI. So we do have pretty good insulation there. But you know, getting back to the macroeconomic environment, I do think we've proven to be a bit of a recession-proof type of equity.
Connectivity and the importance of connectivity to probably every economic agenda is extremely critical and top of mind. You know, we've done a lot over the last three years to further strengthen the balance sheet as well. We did the CoreSite acquisition in 2021. We've since delevered from about 6.8x , basically at about our 5x target today. We've very much been focused on enhancing our quality of earnings, pro forma for India, which hopefully will close here in pretty short order. We'll probably reduce our emerging market exposure, probably to around 25% of our unlevered attributable AFFO. And I think that that commitment to quality of earnings, durability of earnings, and balance sheet strength, I think, has certainly been rewarded. You might have seen, we actually...
Actually, I know you didn't see because I read the transcript with Steve, but the S&P upgrade to BBB . Moody's recently extended the tolerance for leverage as well, and I think all of that is a great reflection of the efforts that we've been putting in over the last three years, and I think it's being recognized.
But in addition to interest rates weighing on the stock, I think the 5G rollout was slower than hoped in terms of lease- up. Can you talk about where you are now and-
Can you repeat the question?
Yeah. So the question is kind of, I think, articulating the 5G cycle. How do we kind of view the 5G cycle? And I think the underlying question is really, you know, on the fundamental side, has that been a drag on the stock in terms of monetization of 5G on the tower side? You know, really, for us, 5G is playing out very consistently to what we thought five years ago. We always looked at 5G as being a decade-long investment cycle. We always looked at 5G as being two peaks of investment, so kind of like a sine wave. The first peak is gonna be coverage-focused. Carriers rapidly deploying mid-band spectrum, covering their POPs.
There's gonna be a level of grooming, and then as you continue to see strain on the networks, you're gonna see a need for densification or what we would kind of call that capacity-focused investment. I think the peak for coverage was 2022. The carriers probably invested around, you know, low-to-mid $40 billion CapEx in 2022, and if you want to kind of put that in a historical context, they probably average around $30 billion in 4G and probably around $25 billion in 3G. So a very material step up. We benefited. We actually had record levels of organic new business in 2023, probably commenced around $230 million of new business, which is around a little over 5% contribution to our organic growth number. So again, record levels. So we were certainly able to monetize.
And again, I get that. I point back to our contracts and our long-term focus on how we structure contracts with upgrades in mind and being able to monetize on those investment cycles. What we're seeing in the last couple of years, I would tell you, twenty twenty-three, the pullback in spend was far more abrupt than we anticipated. In the same way, we've gone through cycles where we've had to increase our services expectations, which is kind of a leading indicator for leasing. We had to abruptly pull it back last year. We went into twenty twenty-four with an expectation for a re-acceleration in activity. We looked at it through a host of lenses. One, we have pretty good visibility into how the carriers think about their networks because we're indifferent between doing what we call a comprehensive or a holistic agreement versus a la carte.
But at the same time, it gives us very good visibility into how the carriers are thinking about their networks over an upgrade cycle. So we knew that there was a high need or expectation to get to kind of that 90-95% type of coverage or upgrade on their portfolios. And I think when we cross-checked that to just conversations that we directly had with the field teams, our historical experience, our knowledge based on, you know, the strain on the networks and ultimately what's underwritten in these agreements, we had a pretty good expectation that we'd see a re-acceleration here in 2024. That's played out the way we anticipated. You know, Q1, I think we increased our applications around 70% versus Q4 of last year.
We saw another acceleration in Q2, and our expectation is to continue to see acceleration over the course of this year. This is still really focused on the upgrades. We haven't really gotten to that densification phase, but I would tell you, even those customers that we have comprehensive agreements with, the conversations have become more constructive on the densification side, and, you know, our largest customer is probably on roughly half of our sites, and use that to say there's a lot of capacity for us to utilize our portfolio beyond the existing footprint and play a bigger role, either in co-locations or, you know, I think hopefully we might find it economically attractive to play a part in new builds at some point, too, over the next several years.
But, you know, we're we've been a little bit insulated from the ebbs and flows of activity over the last two or so years because we have had these comprehensive agreements. Maybe just to level set with everyone, we have an MLA with everybody. An MLA establishes what's the escalator terms, churn rights, what's the rate card. So how am I going to price any incremental square inch that you're going to upgrade on an existing site? But then I could elect to do what we call a comprehensive or a holistic, usually shorter term, shorter duration, because it's focused on a network need at the time. So if a carrier needs to upgrade and go from 0% to 90% over a five-year period, we kind of price that out. We're not going to leave anything on the table.
They like the speed to market advantage. It's administratively easy, enhances the customer experience. We're all for it, but it does give us a very high degree of visibility into what the growth path is going to be, and when you see a little bit of a pullback or, you know, acceleration in spend, we have still a high degree of visibility into exactly what that growth is going to be, regardless, so over the last couple of years, we've largely had comprehensives in place with the big three plus Dish. One of those agreements rolled off at the end of 2023, so that customer is actually going a la carte, but we still have a high degree of visibility.
I think, you know, if you kind of think about those agreements focusing on upgrade, we've retained the monetization opportunity when you get to densification, when that time comes, which we're enthused about.
In terms of the carrier demand dropped off, now it's picked back up. Do you have any visibility into 2025? Where are we even having densification?
I think there's still a lot-
Just to repeat, what's the outlook, you know, the rate of growth that we think might happen in twenty twenty-five coming out of twenty twenty-four?
Yeah. So there's still a lot of work to be done just on the upgrade cycle. So I personally, from where I sit, I don't see the densification taking scale in 2025. I think there's green shoots that kind of start to show that acceleration. You know, internally, we've kind of thought about that second peak, that capacity focus peak, to probably be more 2026, 2027, 2028. And while the pickup in activity this year, it's not really influencing our leasing revenue because we do have these comprehensive agreements, it's a very positive sign. You know, we saw that abrupt pullback in 2023, and obviously, a lot of questions are, you know, is this a priority for the carriers? Is 5G going to follow the same type of cadence as 4G?
What we're seeing now is, I think it continues to reinforce our initial outlook on 5G. I mean, maybe there were other reasons for priorities in 2023, but I think we're seeing a re-acceleration in that commitment towards what's been consistent in what we've observed in 3G, 4G, and ultimately reinforcing how we thought about 5G, and so as we look to 2025, we still have a very high degree of contracted visibility into growth, even with one customer off the comprehensive. We are working through our final tranche of Sprint churn, so we have kind of accelerated the impact of Sprint churn, and that was really underwritten in our T-Mobile MLA, executed a few years ago, so we'll do about $70 million of annualized churn in Q4 of this year.
So that'll have a year-over-year growth impact. Everyone should be aware of that for twenty twenty-five. So I kinda think twenty twenty-five will probably be a relatively similar profile to what we see here in twenty twenty-four. And I think over time, we'll continue to see a bit of a shift between the upgrade and more into the densification as we get kinda to the later end of that multi-year guide that we put together.
A follow-up question on the... You, you mentioned the idea that you weren't-- it wasn't just new contracts to your benefit from the upgrade, but the older ones as well. And I guess my assumption would be that that's every time they touch the tower or put on new equipment, that's how you protect your... So I'm just wondering, like, because you talked about how, you know, as you're writing new contracts, you kind of set the table for moves to 6G, 7G. How do you-- how do you protect your real estate rights? Is it simply every time they touch it and every time they put more equipment on? I guess, what if we come to a cycle where there's equipment is smaller? I mean, how does, how does that sort of think about it?
Okay, I got to repeat the question. So the question is, how do we kind of maintain the ability to optimize the monetization of our sites through incremental investment cycles? And every contract's gonna be a little bit different. I think one of the things that American Tower really benefited from is how we've built a portfolio in the U.S. over the last two decades. You know, we have largely you know, we have forty-three thousand sites in the U.S. and Canada. A lot of it's come through the form of acquiring developer portfolios, third-party independent portfolios, and then strategically adding in alignment with, you know, a customer like Verizon, when we did that portfolio back in two thousand and fifteen.
But it allows us to kind of aggregate a developer, third-party type of portfolio that probably comes with a level of monetization in the underlying contracts, and then we're able to kinda bring it into the American Tower system, whether we kinda roll it into another MLA with the likes of T-Mobile, AT&T, Verizon. But one of the critical focus items, and when we sit at the investment committee and review a portfolio, is understanding what are the underlying contracts and what are the real estate rights over the long term, and you know, when you kinda give up those real estate rights, and granted, you would value those differently. Not to say you can't make money, you just have to underwrite them differently.
I think one of the frustrating things, probably from our seat, is on the investor side, maybe you say, "All right, well, that looks like a cheap deal. That looks like an expensive deal. Why - that was a low multiple. Why didn't you acquire that?" Unless you really know the underlying terms and conditions, was it a fixed escalator at 1%? Are there no real estate monetization capabilities? Are they restricting co-location and lease-up? All right, that's gonna be a lower multiple deal versus something that actually you're able to monetize. But we take a very long-term view. I mean, there's probably an opportunity to maximize 2025 growth or 2026 growth, but it can't come at the detriment of giving away real estate rights longer term and prohibiting us from ultimately leveraging our capabilities as an operator and from a legal perspective.
I think that's a lens that we've taken over the last 15 to 20 years, and that's probably one of the most critical elements to evaluating whether or not you go à la carte or you go comprehensive with a customer. You need to. It's not just about what's the exit rate in 5 years to their monthly run rate, it's also, have we maintained a level of optionality and flexibility based on a perspective of what 6G, 7G, and 8G? We don't know exactly what it's gonna look like, but that's also why these comprehensive agreements are typically shorter term in nature as well, because you wanna focus it a little bit more on what's the task at hand.
You know, if an underlying master lease agreement is 10-15 years, a comprehensive is usually 3-5 years, and that's because both parties want to focus on the task at hand. So if you kinda go back to twenty eighteen, twenty nineteen, twenty twenty, when we signed a lot of these, we knew mid-band spectrum was going to be rolled out, so there was very much a very tangible, what's an upgrade cycle gonna look like, and, you know, do we wanna touch 90%-95% of the sites? But co-locations would largely be outside of it. Neither party really wants to sign up for a use it or lose it use fee in twenty twenty-nine associated with a co-location, right?
You want, you wanna make sure it's kind of focused on what I have visibility into today and maintaining a level of flexibility for the future. You know, customer A doesn't want to pay for a co-location if they don't see the densification need yet. In American Tower, we don't necessarily want to give up monetization of a co-location if it means AI to the device is gonna drive significant amount of new bandwidth. FWA is kind of a dedicated CapEx deployment for the carriers. You know, these are all things that you want to maintain a level of optionality and flexibility, and, you know, it kind of extends not just to the underlying existing contracts, but how we think about monetizing that new wave of, of leasing as well.
I guess part of what I'm asking for is the mechanism by which... discussed philosophically, how do you approach these and potentially value them? But just what are the mechanisms that you use?
Yeah, we haven't gotten into the mechanisms, right? I mean, that's something and quite honestly, it's gonna vary by customer. Every contract's a little bit different. We're very careful not to get into the details of our specific customer contracts, so. But I mean, it's a very detailed and intensive effort. I mean, even if an agreement expires two years from now, you can probably assume we're sitting down with the customer, going through that agreement. You know, how are they progressing with their network deployments? What's kind of that anticipated exit rate, and what's their view for new opportunities in the future? And it's constantly evolving. And, I mean, at the end of the day, we-
Maybe in that question, too, I think, is there technology risk, you know, that threatens the inherent value of the relationship between the tower company and the wireless company?
Yeah, I mean, no, no, no crystal ball. But I think the reason we wanna do shorter duration contracts as it relates to a comprehensive is certainly to retain a level of optionality. I mean, I can't tell you, is there gonna be a new technology 10 years from now that materially changes the footprint on a tower? But from where we sit today, I mean, when you look at a 10- to 15-year contract, I mean, I think it gives us pretty long visibility into the importance of a macro tower and the installations that they have on it.
I also think that, you know. Let's just zoom out a little bit, right. Number one is there's only three ways to create wireless capacity. If you make an assumption that wireless needs will grow, there's three ways to do it. You've got technology XYZ, you've got the number of cell sites, and you've got the amount of spectrum. The most efficient way to broadcast wireless spectrum is probably a satellite, but the next most efficient way is a tower. The economic reality of the situation kind of hinges on this idea that there's rising wireless demand, there's finite ways to address that demand, and towers are the most economic way to address that demand. You own the largest portfolio of domestic towers, a worldwide, publicly traded company.
China's probably got something to say about that. But, I think that's where the inherent real estate value comes from, is that the amount of people, you know, tenants that are going to be wanted within these very finite number of beachfront sites around the country, is gonna rise.
Yeah.
And so then the question, another question would be: To what degree do you believe you have defensibility around the fact that there's 125,000 freestanding metal structures in this country? Why are there not 250,000? Why not? Why aren't there 500,000?
Yeah.
What's gonna happen there?
I mean, look, I think it's. I think we kinda look across the landscape, and I think there is gonna be a need for incremental infrastructure when we get to a densification phase.
You believe that?
I mean, I think for us, scale is extremely important. There is a very clear value in terms of having a nationwide portfolio of 43,000 sites, getting into the system of long-term comprehensive agreements, and I think just recognizing the disruption that goes along with actively moving antennas within a network. And I think you cross-check that up against, you know, the network needs that the carriers are working through today. And, you know, I think historically, like, obviously, there's a cost for anything. Like, I think if lease rates get wildly spread and there's a delta, like, there's obviously a break-even cost to moving.
But between zoning, permitting, and the regulations that you have in the U.S., and cross-check that up against, I think, the level of service that American Tower gives our customers, I think we routinely get very high grades in terms of the eyes of the customer, as well as you can kinda get on a landlord-tenant relationship. There's always some healthy tension there. But we very much pride ourselves, and I think what we've invested as a company into our operational capabilities also provides a level of stickiness in terms of operating with American Tower as well. But obviously, there's a competitive moat when you think about the stickiness of a network, the disruption of a network, to really augment the RF design and try to move it.
Obviously, there's a degree to which, you know, I'm sure there's a price at which it becomes economically unattractive, but I think having nationwide scale and a leading portfolio of assets, I think gives you a lot of optionality to kinda solve for those type of disruptions longer term.
Just a quick question on the fiber transactions recently between Verizon and others. Any thoughts about how that, why they're doing these kind of fiber transactions and fixed wireless and impact on towers, I think?
Yeah. So the question is, weighing in on the Verizon acquisition, and maybe more broadly, you know, fiber in general among the carriers and kind of the importance and the implications on towers. You know, I'll leave it to Verizon and others to kinda talk about the strategic importance. I mean, I think for us, convergence represents an expanding playing field for infrastructure, I think, across, you know, certainly the U.S. You know, our long-term guide certainly doesn't consider any sort of emergence of cable, and quite honestly, it doesn't really consider any emergence of fixed wireless either. I think, you know, we certainly observe, you know, the deployments that the carriers have done to date, and I think it's where typically you see a level of fallow spectrum.
So I can't sit here and tell you that there's an inflection in our growth associated with FWA. Most of that is probably covered within the comprehensive agreements that we have today because they're existing installations. I think the big question for us is: Is it economically viable to begin dedicating capital intensification to monetize FWA? You know, I think the carriers have put out probably collective targets of 11 million-13 million subscribers in 2025. I think from where we sit, we kind of think that can be accommodated with, you know, the spectrum that's available today. I think from where we sit, we'll be paying close attention to any revisions to those targets longer term. But certainly, you know, we hope it could be one of many catalysts that ultimately could drive a level of densification over the long term.
You know, I think it's always a positive when you see a level, I think, of convergence and new opportunities for other players to, you know, maybe play in the backyard of others, and whether or not distributed infrastructure, 43,000 points of presence could be attractive to an operator. I think we're kind of in wait and see there. I think as it relates to our long-term expectations, we've kind of been in wait-and-see mode on FWA in general. I think we're cautiously optimistic, but, you know, we'll be waiting to kind of understand how the targets kind of evolve among the carriers. We'll leave it to them.
I would maybe just throw out there that, you know, Sampath, the CEO of Verizon Wireless Consumer, last week at our conference, you know, highlighted that, when they have a fiber connection, the wireless churn is about half what it would be without the fiber connection. Which I think, you know, if the question is: Does this newfound interest in fiber distract the wireless players, for instance, from spending on the mobile networks? Or, I think the answer is no. I think what the interest in doing this is to make the wireless business more valuable, and it accrues more value to the mobile, which gives them more resources to invest in the mobile business. I think that that's really what the impetus of that is.
It's not a distraction from the mobile business. It's actually part of being interested in being a successful mobile operator. And what's interesting about that is that the more fiber that we put out there, does that put more pressure on the cable industry to maybe take this idea of building out their mobile networks from an owner's economic standpoint more seriously? They have not really taken it very seriously, but you know, that's potentially super interesting. We only got, like, three minutes left. What do you want to talk about, man? So I think maybe just while we wrap up real quick with American Tower's pitch on the data center business, because we're going to be talking to the data center companies later.
We'd love to hear your perspective on how business is going, what is the opportunity, and how much, how much of your resources, now that you're reaching your leverage target, are you prepared to commit incrementally to the data center business?
Yeah. So, so CoreSite has been performing exceptionally well since we acquired it at the end of 2021. I would just remind everybody the rationale for CoreSite. It's not a diversification play for American Tower. I think if you follow AMT for the last five-plus years, and this precedes Steve, it goes back to Tom, it actually goes back to Jim Taiclet, you know, kind of this thought of: How can we utilize our distributed points of presence to play a larger role in a 5G ecosystem? And one of the clear opportunities in our mind, as you continue to see latency-sensitive use cases and applications evolve, would be this idea of the mobile edge. You know, we own the land under or have perpetual easements under 35% of our sites.
We largely have fiber, we largely have adequate power, we have distributed points of presence, and it's always kind of appeared to be a natural synergistic opportunity for AMT. We did a couple of, call it, pilots preceding CoreSite, and I think what it really kind of told us is, if you can't solve for interconnection, you really don't have much of a value proposition. It's kind of like owning a piece of property along a highway and just hoping a tower winds up there someday, so if you go read the tender offer documents from CoreSite, you can see that actually our relationship started with us talking about a partnership, and you can assume we had a lot of those types of conversations, but I would really highlight, not all data center companies would have made sense for us.
It really does require an interconnection hub like CoreSite, that kind of has that campus hub-and-spoke model that we can really kind of extend out to the distributed points of presence. We underwrote it on a standalone basis. We thought it was largely underfunded, so giving them a level of financial flexibility was certainly an upside, but what we've seen to date is really phenomenal new leasing across the business. We had record sales in 2022, broke that record again in 2023, and I think we're on a great track here in 2024. It's, you know, we've largely raised the capital we're putting towards it. I think we probably underwrote around $200 million-$300 million of annual capital.
When we did the deal, we've probably raised that up to north of five hundred, and I would probably anticipate that'll kind of stay elevated the next couple of years, but it's really success based, and you know, I think an interconnection hub like CoreSite is playing a really valuable role in terms of performance and latency-sensitive applications today, but I think as we get into an AI environment, I don't think the large language models today necessarily fit. They're putting a lot of strain on the overall ecosystem, but I think when you start to get to AI inferencing, that could certainly be either an opportunity within CoreSite or I think as we kind of play out this edge thesis. In the near term, I think we're very happy keeping an elevated profile in terms of capital spend at CoreSite.
These are mid-teen type of yields that we're underwriting. It doesn't necessarily change how we view whether or not we want to materially expand the data center platform. I think investors, they can do their own diversification. They don't necessarily need American Tower to. This is an enhanced option for American Tower to play a larger role at the edge, and I think over the next several years, we'll continue to try to use this enhanced positioning to kind of prove out that one plus one equals three value proposition, and in the near term, you know, we're really excited about giving CoreSite an opportunity to continue to grow and enhance that value.
It's a great place to leave it. Thank you so much, Adam.
Thank you, guys.
Thank you, everybody, for joining.
Thanks.
Appreciate it.
Thank you.