American Tower Corporation (AMT) Earnings Call Transcript & Summary

Let's get started. Good morning. Welcome, everybody. My name is Greg Williams. I cover the cable, wireless and telco space at TD Cowen. I'm joined today in this session by Steve Vondran, President and CEO of American Tower. Steven, thank you for joining us.

Thanks, Greg. Thanks for the invite.

Great. Steve, maybe you can just talk about your role as a CEO that you took on earlier this year. Now that we're over halfway through the year, can you just talk about strategic priorities, objectives for the company? Anything you're sort of doing differently or putting your DNA in the company.

Sure. Well, look, I've been with the company for over 24 years. I've been part of the executive team for the last almost 6. So long term, you just see a lot of continuity in terms of our strategy. Having said that, our kind of near to midterm priorities are, first and foremost, sales, driving organic growth on our assets throughout the globe. And that will always be a top priority for us. Second is complementing those sales with selective investments where we can earn highest risk-adjusted returns. And we talked a little bit on our second quarter earnings call about some of the strategic pivot that we've made to increase our investments in the developed world and decrease those in the emerging markets a bit. And that's not a new thing. We've been doing that over time, but we're just a little more explicit about that strategy. And there's actually a development on that that you may have seen recently, but there was an announcement between IHS and MTN, that some of the relocation sites are expected to go to American Tower are going to be retained by IHS and we view that as a very successful conclusion to some negotiations there. And so we'll still support MTN, which is a Tier 1 MNO in Africa, at about 2,100 sites there. And the exciting thing for us is that a much larger percentage of those will be [indiscernible] on our existing assets versus new builds that we originally [indiscernible].

[indiscernible].

Yes, exactly. And that's just consistent with the strategy we laid out at the quarter call about kind of going capital light in those emerging markets, but still driving growth there. So sales, selective investments, predominantly in our developed markets in CoreSite. And then you couple that with a fierce focus on cost control and we've been controlling costs this year. Our SG&A will be about $35 million less than it was last year. And we can continue to focus on driving best-in-class margins throughout the organization. We're using some of our globalization initiatives to look at how can we take costs out in the direct cost line. So things like site maintenance, land risk, things like that. And we'll continue to do that to try to expand that gross margin over time as well. So you take those kind of 3 items, then you add to that our investment-grade balance sheet continue to fortify that. We're very focused on getting our leverage down below our stated range of 3 to 5x. We believe that having an investment-grade balance sheet and having liquidity and the lowest cost of capital that we can sustainably have will be an advantage when M&A opportunities come up in the future. And if those opportunities don't present themselves, we'll be able to drive incremental shareholder value through dividends, share buybacks, internal CapEx program. And then kind of the fifth item that we're focused on is our board refreshment, something we've been doing for a few years. We've had 3 long-standing board members that didn't stand for reelection over the past 2 years. We've added 1 new board member, Neville Ray. Searching for new board members there, and you'll continue to see that refreshment taking place and we just changed all of our committee chairs. And that's part of our dedication to having the right people sitting around the table with the right mix of kind of historical perspective and kind of fresh eyes on the business. So those are really kind of the 5 things we're focused on in the near to midterm.

Great. That's 5 things to chew on. Hopefully, I have time to address all of it today, but we can start with sales. I guess, the U.S. carrier activity last year when we were speaking here, was subdued, and it seems like it's sort of continued that pace, but you did note that some activity is picking up. So maybe compared to where you're seeing activity this year where it was it compared to last year at this point?

Well, we're certainly seeing an uptick for the second half of this year from what we saw in the second half of last year. And we had guided for our services business to have an acceleration this year compared to last year. And we had a lot of questions about that and where the volume was coming from. And so for us, our Q1 application volume from our customers was about 70% over Q4. Now, Q4 was low. And then we saw another incremental increase in Q2. And we expect to see kind of solid activity from our customers for the balance of the year that enabled us to reiterate our services guide for the year. And in our business, the services volume is directly correlated to the activity. Now the property rental performance is a little bit decoupled from activity because of our comprehensive MLAs.

Sure.

But we are seeing healthy activity from our customers. It's pretty broad-based. It continues to be mostly amendment focused as they're rolling out 5G further and further out in the networks, but we are seeing some pockets of densification as well.

Got it. And what's the time line from applications to actually seeing the revenue. Can you just remind us that when you're seeing the activity now picking up this year, how long do we actually see that [indiscernible] revenue.

Yes. Again, with the ones that are under our comprehensive agreements, those are decoupled -- that kind of smooths it out. The activity doesn't matter so much on those. But we do have one major customer that's not in the comprehensive agreement and we have our kind of vertical market segment. So typically, for a new co-location, you can think application to a lease signed being in that 3 to 6 months timeframe, depending on the carrier and their internal processes, amendments move, take about 1/3 of the time line off of that for amendments. But in the comprehensive agreements because of the administrative efficiencies we get there, you can go from an application to sign lease within a few days. They still have permitting and that type of stuff you have to go through, but it definitely shortens the timeline.

And you mentioned amendments are a little more heavy in the mix with this activity. Does that suggest then you're still in coverage mode, do you think? Because last time we spoke was Phase 1 versus Phase 2 and we're going to move more towards co-location later at some point towards densification.

We definitely -- we'll see both. If you just look at the ways that the carriers are built into the prior [indiscernible], the first push is coverage for a certain number of ops. When they get to that, there's a little bit of a slowdown. Then they come back and push that coverage out further out in the network. And so I think what we're seeing today is the carriers continue to focus on coverage, getting that mid-band 5G out further and further into the network, you will see some densification happening probably in more urban areas early. And that densification actually comes in both amendments and in leases because that first network that gets overlaid is really coverage network. You don't have many people using the network at that point. So we'll see amendments just to densify as they add more spectrum to those sites as well. And the handset penetration for mid-band 5G handsets, I think the last number I saw at the end of Q2 was about a little over 50% in the U.S. So as that handset refresh rate continues to drive more more people using it, there will be more traffic on the networks, the densification phase will come.

Right. And you're guiding $180 million to $190 million in new leasing guidance. But you've tracked about $90 million in the first half of year. So simple math says you're on the low end of that. But you may be coming to the middle of it because you've noted your service revenue is going up. That's a leading indicator, activity is picking up. So is that how we should think about it? Or are you thinking maybe tracking towards that lower end.

So the way to think about that, we've also said we don't see a lot of variability quarter-by-quarter. So that would imply that we're going to be at the lower end of the guidance. One of the variables, though we do have one of our carriers, it's not [indiscernible] comprehensive. So the timing of when those leases commence and the activity levels that we're seeing there could provide some variability in that timeline.

Got it. And -- and as we start to look at 2025, maybe remind us of your level of visibility given your holistic MLAs that you have? And how does that visibility compare to 2024?

It's too early to give guidance [indiscernible]. But when I think about '25, the way to think about it is, we're starting the year with kind of a similar level of contracted new business that we had previously in 2024. But the swing factors to think about is we do have a carrier that's not under [indiscernible] of MLA. So the variations in the volume and the timing of that volume is the swing factor there as well. There's also potential merger between T-Mobile and U.S. Cellular. So if that gets approved and closes, there could be some additional churn on that. Our overall exposure is pretty low. It's less than 1% of our U.S. revenues, less than 0.5% of our global revenues. But if there is some churn associated with that, that could be a swing factor as well.

Sure. And you mentioned there's one carryout that you do not have a holistic MLA with anymore. Can you describe the near-term opportunity to have a holistic agreement in place with all of your large customers then?

Yes. You can assume that we're always negotiating with the customers, the ink doesn't even dry on the paper before you're starting to talk about the kind of new network iterations. And we've actually only had 2 years in the last decade where we were under comparative agreement with all the carriers, and that was last year in 2021. So we've been successful under a variety of structures. Typically, when we have one of these holistic agreements with the carrier and we get to the expiration if we don't extend it, it's because we're negotiating, it's usually about what we think their expected activity is going to be and what they think -- there's usually a little bit of a mismatch there. And so you've seen us go for a period of time without one. You've seen us get back into one of the carriers. The key for us is making sure that we're capturing that long-term demand. And so whether it hits this quarter or next quarter, we're going to get the business at some point because they need to do the activity on the site. So we're very comfortable operating with that one. All things being equal, we'd like to have one. Well, the visibility, the administrative efficiencies, it's just best -- better for both parties. But at the end of the day, we're going to do whatever we can to actually capture our fair share.

Got it. And you mentioned on your call that you have 1 carrier that's over 80% of our sites are deployed with 5G. Another carrier is a little over 60%, but there's a third carrier that's a little bit further behind. So is it fair to assume that, that third carrier is going to begin to upgrade their network more aggressively in order to sort of catch up from the leader.

So let me clarify, it's mid-band 5G. We've all had 5G for years. [indiscernible] Look, I'll let the carriers talk about their own specific deployment plans. But what I would say is for 5G deployments, we expect all of our carriers to get close to 100% 5G mid-band overlays. And the reason we think that will happen is the the efficiencies they get on that spectrum allow them to produce gigabytes of data much, much cheaper than 4G. And when you think about the demand driver for our industry, it's the 20% to 30% increase in mobile data usage that's happening every year. It's projected to happen for the next several years, probably as long as we can count. And that means our customers have to produce more gigabytes of data and to keep their margins, they have to do it cheap. And 5G is enabling them to do that. So we think they'll continue to roll out 5G ubiquitously across the U.S. until they get close to 100%.

Right. And we just spoke to Verizon's Chief Network operator last Friday, and they said sort of similar things. They said they're going to [indiscernible] C-band and a vast majority of their 70-odd-thousand sites, mostly by year-end 2025. So that [indiscernible]. I'm curious, you mentioned DISH, very little exposure. I think you said less than it 1%?

That was [indiscernible].

[indiscernible] Okay. But moving on to DISH then in terms of another one that you have very little exposure to. Maybe you can help us with the shot clock booming on the horizon. Have you had conversations in general on their build-out. And -- because as you said right here, we've got a long way to go. I'm just curious your thoughts on where you sit.

Yes. So with DISH, our exposure to DISH is it's less than 2% -- it's about 2% of our U.S. revenues and about 1% of our global revenues. It's pretty modest exposure. But we've been there with DISH from day 1 as they deployed this network, and we've watched them stand up a team and do a great job going to meet their timelines. I'll let them talk about specific plans moving forward. But -- and what we're seeing with them is a continued level of activity that we think is supportive of what they're saying publicly their plans are. And I would just remind you that our agreement with them is a comprehensive agreement with DISH and what we have baked into our longer-term guidance with DISH is a contractually guaranteed minimum levels of new business with them. So from our perspective, whether they go fast or small, it doesn't -- fast or slow, it doesn't make a big difference in terms of that property revenue. You will see it in our services guide as well.

Yes. I mean even on the earnings call, also was on last Friday, they did say -- they're doing a lot of things behind the scenes and so they really seemed confident in meeting that shot clock. We [indiscernible] and churn off. Do you think that maybe they're in the middle of a big bondholder agreement and they got to settle these November maturities, maybe to get that cleared that maybe that could be something that could help them unleash and just go back to business.

I think you got to ask them that [indiscernible] finding out things like that.

Fair enough. So just moving on to India. You mentioned in the second quarter call, that approval from the competition committee is the final hurdles and you can close the sale, I think could come 4 to 6 weeks later. Are there any other hurdles that we should note or should be aware of?

So we actually got approval from the Competition Commission last week.

Okay. So [indiscernible] 4 weeks from now.

Well, I'm not going to call it -- so look, the customary closing conditions [indiscernible] any transaction, so we are saying that, that closing, we anticipate happening in the second half. There aren't any other regulatory hurdles on it. It's just customary closing conditions [indiscernible].

Got it. On the second quarter earnings call, you mentioned that you expect higher SG&A and maintenance CapEx in India prior to close. So -- can you provide any additional color on the magnitude of these higher costs and the drivers behind these increasing costs?

Yes. It's about $15 million in SG&A and about $15 million in maintenance CapEx. And that's [indiscernible] just related as we're getting closer to closing. There's just some things that we need to do to kind of get there, take care of the teams and make sure that we're getting everything done that we need to. On the maintenance CapEx, we just identified a few things that needed to be done as part of the closing.

Okay.

They're relatively small, the overall [indiscernible].

Okay. And as it relates to the international portfolio and large American shipping, it's discretionary capital as you know towards developed markets and away from emerging markets. Maybe help us with that rationale and help us understand the definition really of a developed arketversus emerging markets for American Tower?

Sure. For us, the developed markets are the U.S. and Canada, including CoreSite in Europe. So think of our emerging markets being Latin America, Africa and then India was the vast majority of APAC. We have Bangladesh and Philippines, emerging markets as well. And the pivot we made was actually a few years ago. We haven't done an M&A transaction in an emerging market in several years now. And most of the capital we've deployed has been in the U.S. and Europe, CoreSite in the U.S., Telxius in Europe. And you've seen us also take our internal CapEx program and shift that. So the amount that we're spending in Africa this year is dramatically down from what it was [indiscernible]. And the reason we've made that shift is the macroeconomic conditions that are kind of affecting the globe have an outsized impact on the emerging market. And so the challenges we've seen there, when you go to these emerging markets, you're trying to overcome operational challenges, but also financial challenges. The teams have an amazing job on the operational front. I mean it's so difficult to operate some of these environments, and our teams have the best operators in every [indiscernible]. So the challenges we're really seeing are more economic challenges. And so we've had FX has been a particular challenge in [indiscernible]. We've had a little bit of care consolidation that we worked through largely through most of that. And then in Latin America, it's really been care consolidation as well. And so in particular, we have Oi in Brazil, that's about half of the churn that we're seeing in Brazil is from that Oi transaction. And that [indiscernible] going to continue for a few years there. So as we kind of looked at where we can put our capital and the opportunities we had, we think we're going to drive better long-term shareholder value by pivoting to our developed markets, where we have a lot of great opportunities to invest. We have build-to-suits in Europe, where we have great day 1 NOI yields. We have CoreSite in the U.S. for underwriting mid-teens yields on the new capital we're deploying there as well. And so we think those are just better places for our capital right now given that kind of umbrella of activity.

Yes, sure. I mean, even at the IHS MTN deal, I guess, the narrow pressures in the last couple of years and the opportunities you have in front of you, especially in the data centers, hopefully, will get to. It seems like there's more there. But as you look at the international portfolio, which markets or regions would you be most bullish in? As you mentioned Europe would be part of the developed markets and over the medium term -- and what are the drivers behind that view?

Yes. I'll just touch on all 3 because they have different reasons to be excited about them. So in Europe, we're seeing some healthy growth in Europe. It's primarily being driven by 5G. And while they've got a combination of mid- and low-band 5G deployed pretty ubiquitously, there's still a good runway for mid-band 5G to be deployed there. And we're also seeing a dynamic in Europe where the governments are really pushing for more rural coverage. So you're seeing some incentive plans going in place that are promoting additional builds, [indiscernible] city centers that were originally built there. That's creating opportunities for both new builds and co-locations on the existing portfolios. And we also have the new market entry in Germany of 1&1, which is also something that was not as much [indiscernible] when we are acquiring the Telxius assets. So all those are adding up to a higher growth rate from new leases and amendments in Europe. And we're excited about that going forward. We think those trends are going to continue. Now if I look at Africa, we had the FX challenges, but the [indiscernible] for assets in Africa is huge. We're writing new business in Africa at near record new levels of business every quarter in some of our geographies. And that demand driver there it's helping us weather some of the offsets that we're getting from the FX. But there's no reason I think that's going to change anytime soon. If you think about the continent of Africa, they're so dependent on wireless technologies, their banking system. They do telemedicine there. Like they need to deploy more and more infrastructure. So those sales are going to continue.

It's just a higher hurdle rate.

It's a higher hurdle rate to overcome the macroeconomic challenges there. It's not a demand issue. In Latin America, we're going to have escalated churn for the next few years. But once you get through that, you'll have fewer carriers, they'll be healthier and better balance sheets. And again, demand is spiking there. You're seeing similar growth rates of the U.S. and in some cases, even higher. So they have to invest in those networks to meet that demand. So we think that Latin America will return to a growth rate that's accretive to our overall growth rates once we get through this churn period.

I guess, there's 2 questions there. One is on the 1&1 you mentioned. What's the visibility on 1&1 going forward?

Look, we continue to work with them to deploy their network. And Germany is a challenging market. We've had some of the same challenges that other folks have had with permitting, getting power connected and things like that. That's we're having a global scale is coming really handy. We've got some really good people across the globe who can [indiscernible] help fix some of that. And we've definitely hit our stride there, and we're one of their top vendors delivering more sites than I think anybody else right now. And so that's a good pace. And we see them continuing to build that network to meet the regulatory commitments as well.

Got it. And going back across the Atlantic, what's your visibility on the Oi churn.

So Oi -- Oi makes up about half of the churn that we're seeing this year. So if you look over the next few years, there are 2 different [indiscernible] churn. They've got a wireless business and wireline business. The wireless business was sold to the 3 incumbents, and we would expect to continue to see churn for the next several years from that. We'll retain a lot of those leases, but they'll continue to be churn at similar levels to what we saw this year on the wireless side. On the wireline side, there was a judicial recovery process where we entered into an agreement with them, where we gave them some discounts. That's in our churn this year. We're also deferring some payments in exchange for getting some assets. And then we shortened the term of that agreement to 2027. So we'd expect to see wireline churn in '27. If you think about the kind of headwinds we see in Latin America, you can expect us to be in kind of the low single-digit growth rates through 2027. But after that, we should see that market returning to a nice healthy growth.

Got it. That's helpful. And as I think about the emerging markets, again, we saw India evolve, are there any other markets that sort of have similar characteristics to India where maybe it made sense at one time, but now it's evolving and contemplating, exiting.

Look, the Board and my management team review all of our markets all the time. And I can tell you we don't see another India on the horizon.

Okay.

India has some dynamics that made it unique. We talked about -- I just had a couple because we've talked about the length, but we were stuck with a 2% escalator instead of CPI. We have CPI pretty much everywhere else. You have captive tower companies that are well funded there. You also have regulatory scheme that does not prevent any type of build-arounds. And you've also got a very low ARPU market [indiscernible] focused there. When you look at the rest of the geographies, those conditions are different. We have CPI-linked escalators. We don't have robust captive tower companies that are well funded that are competing with us. And the regulatory schemes, we have the build around. We just had a judicial process in Nigeria that kind of highlighted that where they have a regulatory scheme where you can't build sites near each other. So we don't see another India out there. Having said that, we've got some markets that are subscale, and we've got some businesses that aren't kind of core to our business. And if we got the right price, we might look at doing something with those. But we've been able, through our kind of globalization efforts to support those from different markets. So they're not a distraction, they're cash flow positive, so there's no pressure [indiscernible].

Moving on to data centers. How would you describe the health of the enterprise buyer for data center capacity? Are you seeing that elongated sales cycle within that vertical?

We're really not seeing an elongated sales cycle. We had tremendous demand for our assets in the enterprises, and we don't see that tailing off anytime soon. CoreSite's business is really keyed around people who need to connect to the cloud. And there's a large number of businesses that are either [indiscernible] computers in their offices, they went cloud native or they all cloud and was too expensive. And so they're looking at these hybrid cloud deployments. That's our bread-and-butter. And those customers have done nothing to kind of accelerate [indiscernible] even when the economy was looking a little shaky, that's when they step up their efforts to automate and use machine learning and things like that. So we see that demand cycle continue to accelerate. So CoreSite has had -- we had a record year of new business last year. I can't yet say we're going to this year, I'm pushing the team. It's going to be close. But we're having another great year this year. And we're building twice as much inventory as we ever had in the past. So the demand drivers there are just, they just continue to accelerate.

Right. And as you think about the CoreSite data centers or all your data centers and the GenAI opportunity, obviously, very topical. A lot of the data center opportunities seem to be a training data centers now as they build [indiscernible] and NAS. And then they'll evolve to the inference notes. But of course, it's more of the interconnect facility. So just wondering how do you see CoreSite benefiting in the GenAI opportunity?

Well, there's 2 ways we're benefiting today. The first is, there's so much demand for compute for AI, that's taking up a lot of the available capacity of the market. That's creating favorable supply demand dynamics or pricing is going up. And so that's going up across the board. So we're able to still secure the business with those key customers that we want to, those enterprises at higher rates because the market rates have gone up. The second area is there is demand for inferencing [indiscernible] on the CoreSite.

Okay.

Because once you generate all that data, you've got to talk to people. So you need a distribution channel, and we're the perfect partner for that. So we have seen a lot of interest in inferencing in CoreSite. We've written some agreements with customers for that. We're also being very selective. We create our customer mix because we want to make sure we have the right people, but they're not taking any counterparty risk on kind of a startup that doesn't have.

Right. So one of the big hyperscalers that have the inference.

Yes, just somebody that we feel good about being there for the long term. So we will see more business coming in from inferencing, but we're benefiting across the board from -- and actually, I guess, the third place I would say is the enterprises that are using AI, a lot of them are in these hybrid cloud deployments. So they're adding AI in to their stack there as well. So it's driving the enterprise demand as well.

Right. It's a flywheel hyperscale and the enterprise themselves, right?

Yes, it's a whole of it.

And then talking about power, I mean, it's become a power constraint huge focus in the U.S., and it's only a growing headwind to growth. Are you running into any issues procuring additional power? And as part of that, how do you think about securing access to incremental power in your current markets over the long term?

So the good news is we have power secured for everything under development today and for a large part of our forward-looking development pipeline. Just like everybody else, we're in some markets where power is constrained, and we've got a great team that's been -- we've always had a long view on that, and they've been planning it for at least 10 years in advance for a long time. A couple of months ago, I approved a substation for a site that won't be built for 10 years. We're looking forward to that. So there are some constraints out there. We think we have [indiscernible] hand. There are some deployments we'd like to accelerate if we can. So we'll be looking for alternative power sources for bridge capacity and things like that. But at the end of the day, we feel good about the power we've secured to date, and we think we have a solid pipeline, future power. I would just mention one reason we feel really good about that is our Northern Virginia campus is not [indiscernible] County. So we're in [indiscernible] and so we're not subject to some of the same limitations that they have there.

Okay. That's good to know. And how do you think about potential to expand your geographic footprint at CoreSite.

Look, our first priority is investing in our campuses. That is the lowest risk investment we can do. At the end of Q2, 60% of everything we had under construction was pre-leased. So that's just a low risk investment for us. But we do see opportunities to do some tuck-ins. We bought a small data center in Miami. And you might see us do a couple of small tuck-ins like that. And the reason for that is we think that we might have an opportunity to create a campus-like effect there as well. [indiscernible] take years to grow. So there's not going to be a huge campus in Miami in a couple of years, but 10 years from now, there will be. Obviously we see a couple of those. We also have leveraged the partnership that we have at Stonepeak to be opportunistic here in Denver, and we're building E3, which is the -- we have 2 smaller data centers here. It's not quite a campus effect. The E3 will be our first purpose-built data center here and by utilizing the partnership with them helps accelerate that as well.

Okay. And you started the conversation with your 5 priorities, and we'll bring it back to capital allocation. You mentioned, I think, the last one was deleveraging and focusing on that. But how do you think about your capital allocation strategy when the company gets below 5x leverage, more delevering versus buybacks. You just mentioned possible tuck-ins. So how do you balance it all.

So Rob and I are both very math-based. We're both very focused on long-term shareholder value. So when we get below the 5x, every incremental dollar of capital has to compete. And so if you think about the uses we can do with that. We can further delever, we can increase the dividend, we can do share buybacks. We can increase the internal CapEx program or we can look at M&A. And every dollar that we allocate -- we'll look at all 5 opportunities and whatever creates the most long-term shareholder value is where that money is going to go. That's just kind of what we've committed to and where we'll be looking opportunistically.

The disciplined spreadsheet aspect to it.

Very much.

Right. And when you look at the M&A landscape, have you observed any notable changes in seller expectations over the last year in U.S. and then we talked about Europe and [indiscernible] international.

The short answer is no. There's still a dislocation between private and public multiples. We've seen some of the private multiples tick down a little bit on certain transactions, but some of them are still kind of a bit in higher territory. So at this point, there's nothing that we're seeing in our M&A pipeline across the globe, and we look at everything and [indiscernible] teams they may like to buy stuff. So they're going to look at everything that comes across the [indiscernible] there. But there's nothing that we're seeing that's compelling enough to take us off of our path of delevering and focusing on that.

Okay. I want to [indiscernible] back to just the tower space in the U.S., [indiscernible] the earlier conversation about densification, call-over [indiscernible] that we're talking about. If we're assuming the steepest ramp of densification in '26 and 2027, I mean, first of all, do you push back with that assertions when you think densification would happen?

I don't think I would push back on that. I think it all depends on, again, the handset [indiscernible] users. We're seeing that 5G handset users are using twice as much data as 4G. And so I think that, that's pretty good time frame to start seeing some needs to [indiscernible].

Do you see a correlation between increased adoption of AI and the need for additional mobile infrastructure? I mean we're just trying to think about how AI is going to the edge and maybe then the micro edge. Is that something -- I'll stop there, maybe use cases or anything to think about.

Let me start with the first one is the demand on the wireless network. I think it's a little early to know what AI is going to do. I think it depends on what the usage factors are. I know the way I use microphone, if I'm talking to ChatGPT, it's text or maybe a photographic. Those aren't data-intensive kind of applications. You start seeing video if you're able to start [indiscernible]. If I see you on TikTok dancing with Beyoncé, I'm going to know it's AI created. But I think that's where you're going to see -- see the heavier data usage coming from it, but we also don't exactly know how AI is going to be embedded in the phones yet. We're hearing -- we're hearing that -- that's coming. And so you don't know what the bandwidth demand is going to be on the network until it actually starts happening. And so I think there's opportunity there.

Yes, there's a lot to chew on. I guess, if the AIs in the hardware with Apple talking or if it's out in the cloud.

I do think the video will come.

Yes. But the video is interesting just because AI is creating video, are we going to be watching more video. There's only so much doom scrolling we do today and just going to be created by AI, not by TikToker.

The mobile data usage is going up 20% to 30% today is predominantly video. Video or those traditional kind of channels, social media, it's Netflix, it's all the usual suspects that are driving it today. So we're watching more video and we're watching more high-definition video today. So I think anything that drives more video usage is going to continue that demand and maybe even steep curve on it.

Sure. Okay. And then last few topics is just on the -- one is Africa. You mentioned in the second quarter call that you're looking into using the Power as a Service business model in Africa for other markets. Help us provide an update regarding that comment and have you exported the model to any other markets?

At this point, we haven't exported the primary power to any other markets. Africa, it's a little bit of a different situation because we have an unreliable power grid there. So we have to provide primary power for a large portion of the day there. But we have a lot of learnings there. That team has become a real expert in both power storage and power generation. And they've also been able to take advantage of a lot of renewable power there. So as we're looking at that and how do you export that to a developed market that has a more reliable grid becomes the question of, can you do more renewables? And can you actually store power so that you're not paying peak rates when you have power fluctuations in the grid. So we haven't done it yet, but we are continuing to try to work with our customers to see what appetite there is for that, what opportunities there are. Because again, we've developed an incredible expertise there. So [indiscernible] not to be able to use that in other place.

Got it. And the other opportunity would be and we spoke about M&A in Brazil, a comment I wanted to ask about is the company is underwriting new standards in Brazil? Or is it not from the environment that Brazil is in? And how much of a factor does consolidation play?

Yes. So we've raised our underwriting in all of our emerging markets. It's not just Brazil. And so the way we've looked at it is we're trying to learn from the lessons of the past. And some things that worked very well for us in our international markets, having [indiscernible] escalators, being able to monetize amendments with the anchor tenant, being partnered with Tier 1 MNO, so the top 1 or 2 customers [indiscernible]. Those are things that have worked really well for us. When looking at the things that haven't worked as well, foreign exchange has been a challenge. So we need those CPI-linked escalators and we may need some contractual mechanisms like some dollar pegging which we have in some of our markets. We also need to be -- when we partnered with MNOs that are Tier 1. That's where we've seen more care and consolidation churn. So across that entire emerging footprint, we've just looked at what worked well, what hasn't worked well, and we said, these dynamics have to all be in the works well category, nothing from that doesn't work well. And because there's some inherent risk in those economies, we've raised the hurdle rates because the CPI-linked escalators weren't fully offsetting the FX. And so we've done that across that portfolio.

Have you updated your sort of [indiscernible]? You've articulated publicly, for example, in the past, you used to say that Mexico was x bps above U.S. and then Brazil is x bps above that, and then Africa was x above that. And India was in the past, but [indiscernible].

Not yet. I'm not really going to do that today, but nice try.

That was a good try. Thank you. And with that, we're out of questions. So thank you.

Great. Thanks.