美国塔公司 (AMT) 2026年第三季度业绩电话会：CoreSite推动业绩指引上调lg...

核心要点

• 美国塔美（American Tower）公布，剔除非现金直线式收入和汇率影响后，2026年第三季度综合地产业务收入同比增长约5%。在对DISH一次性流失进行调整归一化后，增长率约为7%。
• 有机租户账单增长约1%，剔除DISH流失后增长约4%。管理层继续将2026年视为美国和加拿大有机增长的低谷年份。
• CoreSite仍是主要增长驱动力。数据中心现金收入增长约16%，促使美国塔美将全年数据中心收入增长预期由此前的15%上调至约16%。
• 公司将全年地产业务收入预期的中点上调了5500万美元，并将可归属AFFO指引每股上调0.02美元。调整后EBITDA指引保持不变。
• 美国塔美同意以约2.02亿美元出售其加拿大资产组合，对应约26倍的2026年预期EBITDA。公司还在第三季度回购了约3.15亿美元的股票。
• 管理层预计，随着DISH流失、再融资成本以及服务业务下滑的影响减弱，2027年每股AFFO增长将迎来显著拐点。潜在的卫星相关铁塔需求未纳入其长期增长框架中。

核心财务数据

业务与运营表现

铁塔资产组合

美国和加拿大的有机增长持平，但剔除DISH流失后约为4%。管理层表示，全年美国有机租户账单增长应在50个基点左右，其中包括约400个基点的DISH逆风影响。预计新业务全年将贡献约250个基点。

各地区表现参差不齐。非洲和亚太地区的有机增长超过8%，欧洲超过4%。拉丁美洲下降约2%，主要是由于巴西的客户流失率上升。管理层预计拉丁美洲的有机增长将在2027年加速。

公司指出了驱动铁塔长期需求的四个催化因素：5G容量扩容阶段、新频谱部署、向6G过渡以及AI驱动的无线流量增长。管理层还认为卫星网络和地面网络是互补的，并预计任何打造大规模美国无线网络的服务商都需要铁塔、光纤、回传、电力和边缘互联。

CoreSite数据中心

剔除非现金直线式收入后，CoreSite数据中心地产收入增长超过16%。云服务、网络、企业和AI客户的需求依然广泛，同时互联活动环比增加。

管理层表示，CoreSite的项目储备比以往任何时候都要强劲，并继续预计在可预见的未来将实现两位数的收入增长。预计全年现金按市值计价重定价升幅在6%左右，而新增有机开发项目的稳态收益率评估在15%左右或更高。

美国塔美计划在2026年投资超过7亿美元于以需求为导向的数据中心项目。根据客户需求、投资回报、许可证和使用权核准情况，其在建与待开发项目储备随时间推移可使CoreSite目前的容量翻近三倍。

资本配置

美国塔美在第三季度回购了约3.15亿美元的普通股，截至10月20日又回购了6800万美元。自2025年第四季度以来的总回购额已超过9.50亿美元。董事会另外批准了20亿美元的回购额度，使剩余回购额度增加至约31亿美元。

计划出售的加拿大资产组合预计将于2026年第四季度完成交割。当前的指引仍包含该业务全年的贡献：收入1300万美元、调整后EBITDA 800万美元以及AFFO 800万美元。管理层预计该交易及资金回收将立即提升每股AFFO。

管理层指引

管理层继续预计服务业务疲软和债务再融资将分别对2026年可归属每股AFFO增长造成约1%的逆风影响。预计在2027年出现拐点后，增长将恢复到长期中高单位数的区间。

风险与关注事项

• DISH的一次性流失仍是2026年的最大拖累因素，使美国有机租户账单增长降低了约400个基点。
• 客户项目的延期使网络服务对调整后EBITDA展望的贡献减少了约1500万美元。管理层将其描述为时间节点的推迟，而非业务的丧失。
• 巴西持续偏高的流失率继续给拉丁美洲的有机增长带来压力。
• 美国塔美预计截至2026年底，其针对AT&T墨西哥的收入准备金将达到约7000万美元。相关仲裁裁决预计要到2027年初才会出炉。
• CoreSite的扩张部分取决于许可证获取、使用权核准、电力供应、客户需求以及可接受的投资回报。
• 来自卫星运营商的潜在需求仍存在不确定性，未包含在管理层的长期AFFO增长框架中。

分析师问答要点

• 美国租赁业务：管理层表示，第三季度新业务贡献较低符合其最初对下半年放缓的预测。预计第四季度新业务将与第三季度基本一致。
• 卫星网络：美国塔美表示正在与有无线业务抱负的各方进行洽谈，但拒绝讨论具体的客户或部署时间表。卫星相关需求将代表其现有增长框架的潜在上行空间。
• DISH设备：公司表示，它拥有在破产过程中被遗弃的设备，并可以不受限制地提供使用权或处置这些设备。
• CoreSite回报率：管理层继续评估15%左右或更高的稳态收益率，并表示更高的利用率和定价可能会带来上行空间。
• 互联需求：交叉连接（Cross-connect）的增长得到了AI、云服务、网络和企业客户在中立托管设施内传输大量数据的支持。
• 并购重组：管理层尚未发现兼具卖家意愿、建设性监管条件和可接受经济效益的重大机会。股票回购、内部投资和偿还债务仍是资本配置的竞争选项。

业绩电话会议完整文字记录

---

完整财报电话会议逐字稿

管理层陈述

Operator

Ladies and gentlemen, thank you for standing by. Welcome to the American Tower Third Quarter 2026 Earnings Conference Call. As a reminder, today's conference call is being recorded. [Operator Instructions]

I would now like to turn the call over to your host, Spencer Kurn, Senior Vice President of Investor Relations. Please go ahead.

Spencer Kurn

Thank you, and good morning. Welcome to our Third Quarter 2026 Earnings Call. Joining me on the call today are Steve Vondran, our President and CEO; and Rod Smith, our Executive Vice President, CFO and Treasurer. Following our prepared remarks, we will open the call for your questions.

Before we begin, I need to call your attention to our safe harbor statement. It says that some of our comments today may be forward looking. As such, they are subject to risks and uncertainties described in American Tower SEC filings and results may differ materially. Additional information is available on our Investor Relations website.

I'll now turn the call over to Steve. Steve?

Steven Vondran

Thanks, Spencer. Good morning, everybody, and thanks for joining today's call. We delivered another solid quarter with our global tower leasing business performing consistently with expectations and our data center business continuing to outperform. The momentum we're seeing across the company combined with the benefit of our share repurchase activity, gives us confidence to raise our full year outlook for the third time this year. Our third quarter performance reinforces the durability of our business model and the quality of the portfolio we've built. We're operating from a position of strength, supported by resilient tower cash flows, a differentiated and rapidly growing data center platform, a healthy balance sheet, and a more focused portfolio following the actions we've taken over the past several years.

As we enter the final quarter of the year, our strategic priorities remain unchanged: driving durable revenue growth, enhancing operational efficiency and allocating capital with discipline. Starting with revenue growth, our tower business remains on track to deliver approximately 4% organic tenant billings growth, excluding onetime disrelated impacts. At the same time, CoreSite's continued outperformance is enabling us to raise our outlook for data center revenue growth to 16%, up from our expectation of 15% last quarter and 13% at the start of the year.

The long-term outlook for wireless infrastructure remains very favorable. Last quarter, I discussed 4 major catalysts that we believe can create overlapping investment cycles and support network demand well into the next decade. The evidence supporting each of those catalysts continues to strengthen. The first catalyst is the capacity phase of 5G. Early deployments were primarily focused on coverage. The next phase will increasingly be about adding capacity, improving performance and densifying networks as traffic continues to grow. That requires carriers to add equipment to existing sites and increased cell site density where demand is greatest, and we're already beginning to see that activity emerge.

Recent industry data reinforces the scale of that opportunity. CTIA reported that Americans consumed a record 159 trillion megabytes of wireless data in 2025, an increase of more than 20% year-over-year, and projects total wireless demand will grow fourfold by 2032. These trends point to sustained investment simply to keep pace with growing consumer and enterprise demand. The second catalyst is a new spectrum deployment cycle. Approximately 800 megahertz of additional mobile spectrum is expected to become available over the coming years, beginning with upper C-band in 2027 and potentially followed by additional deployments across bands such as 2.7 gigahertz, 4 gigahertz and 7 gigahertz.

Historically, new spectrum deployments have led to elevated amendment activity and equipment upgrades, and we expect the next cycle to do the same. The third catalyst is the transition to 6G. While commercial deployment is still ahead of us, the industry is increasingly aligned around network architectures that will be more intelligent, more distributed and more reliant on edge computing. Those characteristics point to increased equipment requirements, greater network density and expanding role for communications infrastructure.

The fourth catalyst is AI which we believe has the potential to significantly reshape wireless network demand. CTIA expects AI traffic to grow 3x faster than traditional wireless traffic and account for nearly 1/3 of broadband traffic by 2034. Importantly, AI is not simply creating more traffic. AI-powered devices, agents, connected vehicles, robotics and real-time applications are expected to create traffic that's more persistent, more interactive and increasingly uplink intensive. Supporting these use cases will require additional spectrum, network capacity, lower latency and greater density. In our view, AI has the potential to extend and deepen the wireless investment cycle beyond what's embedded in traditional traffic forecast today.

Taken together, these catalysts reinforce 1 another. The capacity phase of 5G, future spectrum deployments, the transition to 6G and AI-driven traffic growth, all point to a future requiring more capacity, greater density and increasingly sophisticated digital infrastructure. We believe that American Tower is uniquely positioned to support that evolution. We're also seeing growing interest from satellite providers seeking to participate more directly in the U.S. wireless market. We view satellite and terrestrial wireless infrastructure as complementary technologies. Satellite can extend coverage and support specific use cases, but delivering broad, reliable and high-capacity wireless service continues to require terrestrial infrastructure. Any provider seeking to compete at scale in the U.S. will ultimately require a terrestrial network layer that includes towers, fiber, backhaul, power and edge interconnection.

For American Tower, the emergence of additional participants represents an incremental opportunity. Our nationwide footprint and neutral host operating model allow both existing and emerging providers to deploy quickly, efficiently and at scale. Many of these same secular trends continue to drive strong performance at CoreSite. CoreSite delivered another strong quarter, with customer demand and pricing, both exceeding our expectations. Momentum remained broad-based across cloud, network, enterprise and AI customers, and demand has continued to accelerate into the fourth quarter.

As organizations deploy larger and more sophisticated AI workloads, they require greater connectivity between clouds, networks, compute environments and enterprise applications. This is driving increased demand for highly connected campuses and contributing to continued growth in our interconnection business. CoreSite's unique position at the intersection of cloud, network and AI ecosystems remains a key competitive advantage and gives us confidence in the platform's long-term growth trajectory. Given the strength of customer demand and leasing activity, we're once again raising our full year outlook.

We remain confident in our ability to execute against a development pipeline that can nearly triple our current capacity while continuing to evaluate opportunities to accelerate capacity additions where customer demand and expected returns support additional investment.

Moving to our second strategic priority, operational efficiency. Our tower operations continue to perform with the consistency and discipline that we expect. We remain focused on simplifying processes, scaling share capabilities and using technology to improve productivity. These actions support our objective of expanding tower cash EBITDA margins by 200 to 300 basis points by 2030, while maintaining the service quality our customers depend on. We also continue to evaluate practical applications for AI across the organization. Our focus is on opportunities that improve decision-making, automate repetitive work and make our teams more effective while maintaining the appropriate controls and oversight. We believe these capabilities can create incremental value over time and reinforce our culture of operational excellence.

Our third strategic priority is disciplined capital allocation. Over the past several years, we have reshaped American Tower around higher-quality assets, developed markets and businesses with attractive growth and return profiles. This quarter, we took another meaningful step forward by entering into an agreement to sell our Canada business. The transaction allows us to recycle capital towards opportunities with greater potential to drive long-term AFFO per share growth, including investments in our tower portfolio in CoreSite, debt reduction and share repurchases.

During the quarter, we repurchased approximately $315 million of common stock. And subsequent to quarter end, our Board approved an increase to our share repurchase authorization. Together, these actions reflect our continued confidence in the value of the company as well as the flexibility provided by our balance sheet and cash flow generation. We continue to apply a disciplined capital allocation framework. We'll first fund the dividend and attractive organic investment opportunities, maintain leverage within our target range and then allocate incremental capital across share repurchases, selective M&A and further deleveraging based on relative returns and market conditions. The result is a more focused American Tower with a higher quality earnings base, multiple avenues for growth and the financial flexibility to invest through the cycle. We believe that combination positions us to deliver durable, industry-leading AFFO per share growth over time.

In summary, I am pleased with our performance through the first 9 months of the year and encouraged by the momentum that we continue to see across the business. Our premier global tower portfolio, differentiated interconnection-focused data center platform, strong balance sheet and disciplined approach to capital allocation positions us exceptionally well to capitalize on the next wave of investment in wireless, cloud and AI infrastructure. I want to thank our employees around the world for their continued dedication and execution as well as our customers, shareholders and business partners for their ongoing trust and support.

With that, I'll turn the call over to Rod to review the financial results and outlook in more detail. Rod?

Rodney Smith

Thanks, Steve, and thank you all for joining the call. As Steve mentioned, we've carried our strong momentum into the third quarter and are increasing our 2026 outlook for the third time this year. I'll start by reviewing our third quarter results, and then I'll touch on our revised full year outlook. .

Slide 7 shows a snapshot of our third quarter highlights. Consolidated property revenue grew approximately 5% year-over-year when excluding noncash straight line revenue and FX impacts. Normalized for the impact of onetime DISH churn, property revenue grew approximately 7% on a cash FX-neutral basis. Our growth was primarily driven by organic tenant billings growth of approximately 1% and or approximately 4% when normalized for the impact of one-time DISH churn complemented by data center cash revenue growth of approximately 16%. Adjusted EBITDA grew over 2% when excluding net straight line and FX impacts, normalized for the impact of one-time DISH churn, adjusted EBITDA grew over 5% on a cash FX-neutral basis.

Cash adjusted EBITDA margins declined approximately 50 basis points year-over-year, primarily due to DISH-related churn and lower services contributions. Excluding DISH-related churn, Cash adjusted EBITDA margins grew 10 basis points year-over-year. Attributable AFFO per share grew over 1% when excluding FX impacts. Normalized for the impact of onetime DISH churn and excluding the impact of refinancing costs, attributable AFFO per share grew approximately 6% on an FX-neutral basis.

Moving to Q3. Organic growth and data center growth on Slide 8. Across each of our tower segments, organic growth was in line with the expectations we laid out earlier this year, supported by solid demand across our global portfolio. In the U.S. and Canada, organic growth was flat and approximately 4% when excluding DISH churn, consistent with our expectations for durable growth in the mid-single digits. We remain confident that 2026 represents a trough year for U.S. and Canada organic growth and continue to expect organic growth to accelerate in 2027.

In Africa and APAC organic growth was over 8%, modestly ahead of our expectations for organic growth of 7% in the second half of the year. In Europe, organic growth was over 4%. And in Latin America, organic growth declined approximately 2%, primarily driven by elevated churn in Brazil, consistent with our expectations laid out at the start of the year. We remain encouraged by the prospects of an earlier-than-expected market repair in Brazil and the forthcoming acceleration in Latin America organic growth in 2027.

Finally, on the right side of the slide, data center property revenue growth was over 16% when excluding noncash straight-line revenue. As Steve mentioned, this quarter marked another very strong quarter of new leasing activity for CoreSite, and we remain on track for 2026 to represent another year of record leasing activity.

Now let's turn to our revised full year outlook. We are raising guidance for property revenue and AFFO per share while maintaining our adjusted EBITDA outlook, reflecting consistent growth across our global tower portfolio, data center outperformance and capital allocation decisions.

Starting with property revenue outlook on Slide 9. We are raising our outlook by $55 million at the midpoint representing an increase of approximately 1% relative to our prior outlook. Our revised outlook now implies over 4% year-over-year growth when excluding noncash straight line revenue and FX impacts. Normalized for the impact of onetime DISH-related churn, our outlook implies nearly 7% growth on a cash FX-neutral basis. The increase to outlook was primarily driven by approximately $10 million of data center outperformance and $45 million from other items, including pass-through and straight-line revenue.

We are reiterating our organic growth assumptions across all regions and continue to expect organic tenant billings growth of approximately 1% or approximately 4% when excluding DISH churn and raising our outlook for data center growth to approximately 16% year-over-year. As Steve mentioned, we recently entered into an agreement to sell our Canada portfolio for approximately $202 million which is expected to close in the fourth quarter at an implied multiple of approximately 26x 2026 expected EBITDA the transaction underscores the significant value embedded within our North America tower portfolio. Combined with the opportunity to recycle proceeds into share repurchases, debt reduction and other higher growth investments, we expect the transaction to be immediately accretive to AFFO per share while enhancing our long-term AFFO per share growth profile.

Consistent with our outlook policy, the impact of the transaction will be reflected in outlook once it closes. Our 2026 outlook, therefore, includes full contribution from our Canada portfolio, including $13 million of revenue, $8 million of adjusted EBITDA and $8 million of AFFO.

Moving to adjusted EBITDA on Slide 10. We are holding our adjusted EBITDA outlook flat to our prior outlook. Compared to our prior outlook, outperformance of approximately $10 million in data centers and approximately $5 million in other items, including straight line, was offset by approximately $15 million of lower network services contribution, reflecting timing shifts in certain customer projects that we had expected to commence in the second half of the year. Our outlook continues to apply over 2% growth year-over-year, excluding noncash net straight-line and FX impacts. Normalized for onetime impact of DISH-related churn, our outlook for adjusted EBITDA implies over 5% growth on a cash FX-neutral basis.

Turning to AFFO on Slide 11. We are raising our attributable AFFO outlook by $0.02 per share, primarily reflecting share repurchase activity during the quarter. Our revised outlook now implies growth of over 3% year-over-year. Normalized for the impact of onetime DISH-related churn and excluding the impact of refinancing costs, our outlook for attributable AFFO per share growth implies approximately 6% growth on an FX-neutral basis.

We continue to expect our services business growth to represent an approximately 1% headwind to attributable AFFO per share growth this year. Similarly, we expect debt refinancings to represent an approximately 1% headwind to attributable AFFO per share growth this year. We continue to believe this year represents a trough for attributable AFFO per share growth. Heading into 2027, we're confident that we can deliver a meaningful inflection in growth and return to our long-term expectation of AFFO per share growth in the mid- to high single-digit range.

Turning to capital allocation and our balance sheet on Slide 12. Our capital allocation strategy remains focused on balance sheet strength, disciplined investment and long-term value creation. In 2026, our growth capital plan remains largely consistent with our prior outlook. We continue to expect to spend approximately 85% of our discretionary capital within our developed markets platforms including over $700 million in success-based investments in our data center portfolio to replenish elevated levels of capacity, purchase of land beneath our tower sites and continued investment in our European newbuild program. Relative to our prior outlook, we are reducing discretionary capital spending by approximately $40 million primarily reflecting the timing of certain international tower development projects. This adjustment does not reflect any change in our long-term investment priorities or demand outlook.

As Steve mentioned, during the quarter, we entered into an agreement to sell our Canada business as we continue to simplify our portfolio and support our strategy of recycling capital into higher growth opportunities across the business. We also repurchased approximately $315 million of American Tower stock during the quarter, plus an additional $68 million through October 20, bringing our total share repurchases, since we started buying back stock in Q4 of 2025 to over $950 million. And as Steve mentioned, subsequent to quarter end, our Board authorized an additional $2 billion under our share repurchase program, bringing remaining repurchase capacity to approximately $3.1 billion.

The actions we've taken over the last several years have positioned us with significant financial flexibility. We ended the quarter with leverage of 4.9x, comfortably within our target range of 3 to 5x, while maintaining the highest credit rating among our peer group. As demand for digital infrastructure continues to grow, our balance sheet remains a key strategic asset that enables us to pursue opportunities from a position of strength.

Turning to Slide 13. Our third quarter results reflect the durability of our business model and the consistent execution of our strategy we continue to see favorable demand trends supported by rising mobile data usage, sustained investment in communications networks and the increasing need for critical digital infrastructure. Together, with our high-quality portfolio, disciplined capital allocation strategy and financial flexibility, these trends reinforce our confidence in our ability to generate durable growth and create long-term value for our shareholders.

With that, operator, please open the line for questions.

Operator

[Operator Instructions] And our first question comes from Batya Levi of UBS.

分析师问答

Batya Levi

Can you provide a bit more color on the application volume you're seeing in the U.S. today? There was a bit of a step down in new leasing in the quarter. What would you attribute that to? And how should we expect new leasing to pace as we exit the year and into '27. And maybe another question on -- we're hearing a lot of discussions on potential Starlink terrestrial wireless network build. Can you provide maybe a bit more color in terms of the level of engagement you're seeing from them today. And how would you characterize how you're positioned to capture this opportunity? Do you anticipate new business to follow all the tower companies in the U.S. or maybe be positioned to 1 or 2, and when would you expect that opportunity to be meaningful.

Rodney Smith

Batya, this is Rod. I'll take the first part of that question, and then Steve will address the second part relative to Starlink. So in terms of our new business level, you're correct in seeing that our new business contribution numbers have stepped down in Q3, and I'll talk about that in just a bit. .

I'd first highlight that the activity level that we're seeing and the new business levels that we're seeing for the year remain at very healthy levels. We continue to see the carriers deploying 5G kind of the late stages of 5G and they're also transitioning to the early stages of network densification.

I'd also highlight that not all carriers are active at the moment. And I would highlight here that, that means that what you are seeing is that the new business as it paces through the year is front-end loaded, and we are experiencing a step down in the second half of the year. That is entirely consistent with the outlook that we put out at the very beginning of the year. So it's in line with our expectation without a doubt.

And it also for the full year, new business is expected to contribute about 250 basis points towards organic tenant billings growth in the U.S. And that is consistent with the level of new business contribution to organic tenant billings growth we saw in 2025 ex-DISH, so there is an element of consistency there regarding the new biz as well as its contribution to organic tenant billings growth. With that said, we're firmly on track regarding the full year OTBG in the U.S. that we expect as well as, again, that new business contribution.

And that puts our U.S. full year OTBG outlook at about 50 basis points. And I'll just highlight that, that includes about 400 basis points of headwind from DISH churn. Relative to going forward, I'm not going to address '27 specifically at all, of course, as you would expect. But looking beyond 2026, we do expect OTBG in the U.S. to increase to more normal levels as we move forward. This is primarily 2 broad reasons.

One is the elimination of the nonrecurrence of the DISH churn and the corresponding headwind it has on OTBG as well as we anticipate increased carrier activity in an environment going forward where all carriers would be active in the future as well as some of the catalysts that Steve mentioned earlier in the call and in the prior call, that could be catalyst for upside from that.

So as a result, you put all this together, we are confident that 2026 represents a trough year for both organic tenant billings growth as well as AFFO per share growth and that our portfolio, global portfolio of tower assets and data center assets are well positioned to have an inflection point up in both of those metrics going forward.

Steven Vondran

Yes. Batya, I would just add, when you kind of zoom out and think about the carrier environment and what's going on today, I focus back on the mobile data growth that you're seeing in the U.S. And when you look at those numbers that are coming out of CTIA and the growth that we saw year-over-year in total mobile data growth consumption going up about 20% year-over-year this year and projecting that to accelerate even further in the next few years. That really tells you that the carriers will have to continue investing in their networks to provide more capacity.

And that's exactly what we thought we would see at this stage of that evolution. And we are seeing them start to evolve more into a capacity play. We're seeing all the carriers becoming more interested in densification, adding more sites. And so we believe that we'll see the back half of 5G play out just as we talked about in the past on that.

And then when you think about the other catalyst that could accelerate that even further, you've got the spectrum auctions that are coming up, and the deployment of that spectrum plus there's some other spectrum kind of out there being held in private hands that could come into play as well. So you've got a lot of spectrum that could come on board. Traditionally, more spectrum means more equipment and more leasing for towers.

And then pretty soon, we're going to be talking about 6G transitioning into that. And we haven't even started seeing a lot of equipment going on site yet for AI. And we think as AI becomes more prevalent, you're going to see an uptick in that as well. So when you kind of zoom out as to what's going on in the overall ecosystem, there are a lot of positive tailwinds on that really make us feel confident that you're going to see that new leasing environment continue to be healthy like it is today, but actually increase over time as well.

And then I'll address your questions on Starlink. We are really excited about the prospect of maybe having a new customer at scale in the U.S. And we're really encouraged by their public statements and what they're saying that they plan to do and also by the concrete steps that they're taking to build that layer cake of spectrum that you need to be a viable player in the U.S. market. So those are all really encouraging signs that we're seeing there.

It also reinforces what we've been saying all along, which is satellites are a complementary technology to terrestrial networks that you need a terrestrial presence if you're going to try to operate in the wireless space in the U.S. And so this again, it's playing out the way we thought it was going to you're seeing some positive momentum there.

In terms of specifics with us with Starlink, we're not going to get into any specific discussions that we're having. But you should assume that anybody who wants to build a network in the U.S. at scale is going to come to American Tower as one of their premier tower vendors, if not the premier tower vendor. We have an unmatched portfolio in the U.S. in terms of where our sites are located, the quality of those sites, and we can get them on air faster than anybody else. So we should be the first call that everybody makes. So you should assume that that's kind of what's happening there.

And again, we're excited about the prospect. And in terms of timing and things like that, we'll let them speak to what their network plans and their deployment cadence would be.

Operator

And our next question comes from Rick Prentiss of Raymond James.

Ric Prentiss

Appreciate the questions. First, I want to follow up on Batya's comments a little bit. How many sites did you guys have out of the 24,000 sites that DISH has on air?

Steven Vondran

Yes, Rick, we're not going to put out specific numbers on that. We were a significant player in their build and very proud of our contribution to that. And it was a significant piece of that, but we're not going to put out specific numbers on that.

Ric Prentiss

Okay. But use your revenue maybe as a guide, but you've already kind of reported when you took DISH out. .

Steven Vondran

We got a bunch of sites, Rick.

Ric Prentiss

Yes. Okay. Cool. Who owns that DISH equipment that's sitting on the towers? Obviously, the DISH had it, you terminated the agreement then it got it in bankruptcy court. Has it been officially abandoned, and could that be of interest to some potential players, the equipment that's on the towers because we think the 800 spectrum might actually already beyond some of them. .

Steven Vondran

There is an abandonment order in the bankruptcy. So yes, we own that equipment as of the date of that order. As to the interest, the other people may have that equipment. I'll let them speak to what they want to do with that. But we do own it, and we can provide access to that to all interested parties.

Ric Prentiss

Great. That makes sense. And as we think about your earlier comments that satellite providers need terrestrial. We've seen a lot of satellite companies, amazon has their LEO, ASTS -- you've got even Rocket Lab buying Iridium, ASTS, -- are you having discussions with a broader universe of satellite companies that might want terrestrial beyond just what's kind of been mentioned about StarLink SpaceX? .

Steven Vondran

Rick, you can assume that anybody who wants to play in the wireless space is exploring all their available options including building, buying or renting a network and that we're talking to all those parties. So anybody who has wireless ambitions is probably somebody that we're talking to.

Ric Prentiss

Great. And last one for me. Obviously, the data center business continues to outperform. You raised guides because of data centers, revenue growth rate there. Can you help us understand a little bit about the return on capital that you're seeing from spending that almost $700 million to continue to increase the supply out there. .

Steven Vondran

Yes. Thanks, Rick. I'll touch on some of the demand drivers and if Rod wants to give any specifics he can on this. But when you think about that CoreSite business, again, I want to distinguish it from other types of data centers. It is an interconnection platform that brings together clouds, networks, enterprises and now AI companies so that they can communicate together.

And what that allows us to do is be very specific in curating our customer mix, but it also gives us the ability to price the way we need to, to continue to underwrite mid-teens or better stabilized yields on the incremental capital that we're investing. And so we continue to underwrite that, quite frankly. We're outperforming it in many instances. But that's kind of the baseline assumption that we have for all the capital that we're deploying in CoreSite is kind of in that range for all of our organic builds.

Rodney Smith

And Rick, I would just add a couple of metrics. You've seen our double-digit revenue growth that we posted for the quarter and the outlook continues to look at double-digit revenue growth expectations for CoreSite. That also comes with expanding margins and some really interesting results and cash mark-to-markets, which is also double digit, which is a significant increase from the traditional 3% to 4% increase.

So we are having a fantastic Q2 and Q3 relative to cash mark-to-market. I would just caution you that for the full year, we do expect that to be more in the mid-single-digit range in the 6% range. We're also -- we're experiencing low churn in the second half of the year here. And the business is performing exceptionally well. And just as Steve said, the capital that we're putting out in this business, we expect mid-teens stabilized returns.

And as the environment, if the environment continues to drive the way it is and stabilized yields, we get beyond that and have more utilization and continue to have the pricing leverage, there's certainly upside from there.

Operator

And our next question comes from Cameron McVeigh of Morgan Stanley.

Cameron McVeigh

Just had a couple. First, on the U.S. services, so there was a $15 million headwind to EBITDA this year versus the prior outlook. Can you just give any more color on the timing shifts you had mentioned, and I might have missed it, but is there a services revenue outlook? And then secondly, on the CoreSite cross connects, it looks like there was a step up in cross connects this quarter, is that just lumpiness? Or do you think AI is driving an increasing level of cross connections going forward?

Rodney Smith

Cameron, thanks for joining the call. I'll start with the services piece. As you see in our outlook, we did update our outlook that reflects now a modest pullback in services revenue down to about $220 million of revenue. The first thing I would highlight is that number is still a very compelling strong number up against historical standards. It wasn't that long ago that our services revenue was typically in the $100 million range.

The pullback itself is related to the timing of certain activity projects from a certain customer that we now expect to be delayed and not happening in Q3 and potentially not in Q4. So that $220 million pullback in revenue, it does have roughly a $15 million reduction in the EBITDA levels that within our outlook.

The one thing that I would also highlight here is that the services revenue pullback is not directly related to any organic new business reduction. Those 2 metrics at this point really are decoupled in the services is activity exactly at the site services work where our new business is more tied to terms and conditions within holistic agreements that we have in other MLA types.

Over the years, we have expanded our network services operations to now include end-to-end solutions, including zoning, permitting and construction management. And the increase in revenue that we've seen in the services business over the years really reflects and demonstrates the value of our network services capability to our customers in terms of helping them get new equipment deployed on our sites efficiently and quickly.

Steven Vondran

And just touching on the interconnection. We are absolutely seeing an increase in interconnection kind of quarter-over-quarter. I don't think it's lumpiness. I think that you're seeing a different demand characteristic. And it's definitely driven partially by AI, but it's not just AI. You're seeing an increase in these ecosystems where more and more customers are talking to each other. And so it's enterprise to cloud, it's enterprise to AI. It's AI to cloud. It's network to AI. It's really all of those different components are seeing an increase in cross connects.

Operator

Our next question comes from Michael Rollins of Citi.

Michael Rollins

Just a few follow-ups, if I could. So first, Rod, going back to some of your comments on domestic leasing. So if it's 250 basis points of core leasing contribution this year, and you take that against the nearly $5.1 million you ended with last year, that's about $127 million, and you did $93 million to date. So is your indication that these dollars are expected to bounce back towards that $34 million level just by that math, that you saw quarterly in the first half of the year. And then I'll just after that follow-up with a couple of others, if I could, please.

Rodney Smith

Yes, Michael, thanks for the question. I would just highlight kind of that the new business is in line with our expectations. You are correct in terms of the math that you're laying out when you look at the new business numbers in our full year expectations. So we continue to believe that we'll be broadly in that range of the 2.5% contribution from new business. .

In terms of the actual new business number for Q4, I mean you know we've kind of moved away from guiding to a new business dollar amount. But we do expect that new business to be consistent in Q4 as we saw in Q3 and that is where it's consistent with our original outlook that we provided, which is we did expect the second half to be lower than the first half.

Michael Rollins

I appreciate the extra color on that. Secondly, I was just curious if there's more of an update on just some of the litigation processes that you're going through. So the situation in Mexico with AT&T Mexico and if the pending divestiture by AT&T has any impact on that process? And then also, in terms of the potential mediation that you're in with DISH, is there just a range of outcomes that you can update us on in terms of like what the high watermark is and like maybe where the downside risk is just to sort of range bound what could develop on that front? .

Steven Vondran

Sure. Well, there's really not much new to report. So we did -- as we had told you guys we had the arbitration hearing in Q3. We don't expect a ruling on that until 2027, early 2027, and so there's really no update on that. And the potential sale of that business doesn't really affect that at all.

In terms of DISH, there is a mediation which is common in these types of scenarios for that to happen. I'm not going to speculate on a range of outcomes on that. I think the range of outcomes is kind of consistently been the same options that are out there. I will just remind everyone that we have derisked our business by taking DISH completely out of all of our numbers and projections and things like that.

And so whatever happens there would be incremental to any of the projections and the numbers that we've given you guys. And just a reminder to everyone that this churn that we saw this year is what has been weighing on OTBG. And so that will be gone next year. So we will see an inflection in growth next year as a result of the nonrecurrence of that churn. And so those processes will just keep playing out. And when there's a resolution, we'll let you guys know.

Rodney Smith

And Michael, I would just ask you to keep in mind, too, that relative to AT&T in Mexico, we did make a reserve of about $30 million in 2025, up for the revenue there. In 2026, we're currently reserving $8 million to $10 million per quarter. So think about '26 is adding about $40 million to that reserve. So at the end of this year, we expect revenue reserves regarding AT&T Mexico to be at about $70 million. So you can think about that in terms of run rates and eventual resolution here and how that may fit into the picture.

Michael Rollins

That's helpful. And one last one, if I could, please. And this may be a tougher question. I appreciate the opportunity to ask it. So you mentioned earlier that part of what was happening with the services business was a delay from a customer. And another tower company talked about a customer delay where that customer was purchasing low-band spectrum that process got delayed in terms of approvals. Is that a similar delay that you're experiencing? And does that also then infer some conviction on the ramp for 2027?

Steven Vondran

If we get too specific about our customer builds, they get very grumpy with me. So we're not going to touch too much on what that was. But what I will say is the delay that we're seeing is just a change in timing, not expectations. So we fully expect to get that business, both the services and the rental revenue associated with that when the customer builds, not if the customer builds.

Operator

And our next question comes from Nick Del Deo of MoffettNathanson.

Nicholas Del Deo

First, you've been experimenting with some smaller multi-megawatt purpose-built data centers at some of your tower sites. One of your peers has talked about potentially supporting deployments in order of magnitude smaller, maybe a couple of hundred kilowatts or so per site. I guess have you received any interest from potential customers about the latter concept? And how would you view its appeal versus the model that you've been developing in recent years? .

Steven Vondran

Yes. Thanks for the question, Nick. We're excited about everyone by getting excited about Edge. It's something that we've been working on for a few years now. And edge means something different every person you talk to. And so there are a number of proof of concepts that are going on with various parties and various sizes. So yes, we're seeing interest from various parties for smaller deployments on sites as well.

our Raleigh facility, which is kind of the first that we've built of its kind. I'm happy to say that the first megawatt leased up already, and we're about to go live on the second megawatt and it's almost completely early step as well. So we're absolutely seeing demand for a little bit larger scale deployments as well at the edge as well as the smaller scale deployments.

And so when you think about the edge ecosystem, the momentum is very positive. And that's something that we're excited about. We think the TAM is expanding because our multiple parties interested in different models there. And we continue to believe in our thesis that connecting those edge data centers back to CoreSite gives us a right to win in that space, and it makes us a more interesting partner.

So we'll continue to iterate on that. And when there's something more material to announce, we'll tell you guys that, and we'll keep you up to date on our efforts with Raleigh. And if we decide to build a couple more of those small ones to iterate on that concept, we're confident those are going to lease up as well because we are seeing a lot of demand in that space.

Nicholas Del Deo

Okay. That's great to hear. Maybe pivoting a bit to capital allocation. You've obviously been leaning more heavily into share repurchases. I think for reasons that you've described in your past calls, I guess I was wondering if you've been seeing any changes in seller expectations in the M&A market, given rising interest rates and whatnot, that might suggest that the relative positioning of opportunities with capital work might be shifted at all in your priority list? .

Steven Vondran

What I would say is we haven't seen material opportunities where you had a willing counterparty, the constructive regulatory environment and the economics all came together for us. And you really have to have those 3 for more material M&A. In some of the smaller tower deals in the U.S., we're seeing a little bit of rationality in some of the multiples, but nothing that's been of scale and nothing that we've actioned on recently -- have up to scale in that space. So I'd say that there are opportunities out there. There are people talking about things, but we haven't been able to bring those 3 components together on anything material.

Rodney Smith

And Nick, I would just add to that and maybe update the listeners here on our overall capital allocation strategy and approach. And I think that entire process is underpinned by management. Steve and I and the executive team as well as the Board being very disciplined in terms of our approach to capital allocation. And I think you've seen that.

First and foremost, we do allocate capital towards our dividend. As a REIT, we are required to dividend out about 90% of our REIT taxable income. We choose to dividend out about 100%. And we think that's most efficient for our shareholders. And we enjoy returning capital to our shareholders.

Next, we look at internal capital programs and you see us this year investing between $1.8 billion and $1.9 billion pretty consistent level. Within that, 85% of that is being allocated towards developed markets, which is exactly in line with the strategy that Steve outlined a couple of years ago in terms of leaning into developed markets, high-quality economies with high-quality counterparties, which gives us a more stable return outlook and an easier, clearer path to value creation for our shareholders.

Next, we look at M&A. And in the U.S., we're doing some small M&A. You'll see that tucked into our cash flow statement and our balance sheet, but nothing material. But we do look at M&A in the U.S. and in Europe and CoreSite and towers. But we're very disciplined. So we don't feel a need to jump in, in a way that would be undisciplined. We don't mind allocating capital towards continued debt reduction in this environment. And we also don't mind buying back shares. We actually like that and at certain times, it can be -- it can be the best use of capital depending on all the factors that come into play. So we're constantly scanning the dashboard when it comes to capital allocation, being very disciplined. We have multiple levers to pull. So that is good.

We have a lot of flexibility, and it allows us to make really good decisions on behalf of our shareholders. With that said, even with the nearly a little over $900 million that we've allocated back to share repurchases, the $1.8 billion to $1.9 billion that we look to deploy in internal CapEx this year as well as the $3.3 billion we're dividending back to our shareholders. We have an incredibly strong balance sheet, stronger than all of our peers. We have a BBB+ credit rating. The lowest leverage out there. And we do view that as a strategic asset that puts us in a really strong position to be flexible and opportunistic as we go forward.

Operator

And our next question comes from Michael Funk of Bank of America.

Michael Funk

Thank you again for the question. So want to go back to a potential LEO terrestrial build and get your thoughts on time required to build out nationwide network using existing equipment, if possible, versus doing more of a greenfield build using your expertise, I think about timing around that. And then looking at the cost line and the services business, didn't come down as much as they would have expected given the drop-off in revenue there. So what does that tell us about anticipation of maybe increased work in the services side?

Steven Vondran

Sure, I'll take the first part, and Rod, you can take the second part. We do not want to speculate on the scope of what people are going to do in terms of their network. So it's very hard to guess at what a time line would be for deployment on that. So again on that, I think I have to direct you back to the parties who might build it. There's just too many variables for us to start speculating on time lines. What I can tell you is that American Tower can get them on air faster than anybody else, and we can offer them services, which kind of leads into your second question. Part of what we do for our customers is take them from application and zoning all the way through construction. And that will lead into the second question with Rod.

Rodney Smith

Yes, Michael, what I would say about the services is that you really have to consider the mix, and I don't want to get into too much detail there in terms of all the different end-to-end services that we provide in the different margins but of course, you can imagine construction management is a much different margin profile than the zoning and permitting we do with the engineering.

So it is a mix of the services revenue and particularly where the reduction came primarily. And I would just highlight that we now have construction management runs at about half of our overall services revenue. So that's been increasing quite a bit. And you'll see that reflected in the overall margins from services to gross margins, which are running now north of 40% from a level that used to be around 50%. That is primarily the mix and the increase in volume that we're seeing in the construction management.

Operator

And our next question comes from Richard Choe of JPMorgan.

Richard Choe

Great. I wanted to follow up on the data center opportunity. What are you seeing in terms of your pipeline of deals? And then in terms of the renewals, you mentioned the 6% mark-to-market. But are you seeing different customers asking for different things in terms of terms and conditions that you think is a good sign going forward for your existing base?

Steven Vondran

Sure. Well, the demand is really higher than we've ever seen before in terms of CoreSite. And that's giving us a lot of flexibility in terms of curating the customer mix and making sure that we're optimizing the yield on those assets.

The mark-to-market is -- and remember, that's on top of kind of annual escalators that we get in that business. So that's just a reflection of pricing in the industry going up, which is a factor of the supply and demand dynamics that are there.

And so when you look at our pipeline, Richard, it's very robust. And again, we try to curate that customer mix. We want people who build the ecosystem who have a lot of cross-connects who are going to be there for the long term. And because of the demand cycle that we're in, we're able to be very selective on our counterparties and curate that customer makes even better than we have in the past.

So we're very confident in the long-term trajectory and the durability of both that revenue and the ability to grow it. When you think about the growth that we're seeing, what we've been very confident in saying is we still see double-digit growth for the foreseeable future in terms of revenue in CoreSite. And that's not coming at the expense of the quality.

We're increasing the quality of the earnings there, we're putting better counterparties in. So I would say it's a very healthy environment, and we're excited to be growing and investing in that business.

Richard Choe

And 1 follow-up on the interconnection side. You talked about it a little bit, but how are you seeing interconnection kind of evolve or connectivity evolve? And then with your more AI kind of centric customers or enterprises that are adopting it? Is there some question on how many cross-connects would end up taking? How would you monetize it? Can you give us a sense of what you're seeing?

Steven Vondran

We're just seeing an explosion in growth when you -- there are massive cross-connects to people who are operating LLMs. But it's really -- the way to think about it is everyone is becoming more connected to each other. And so you're seeing these cross connects between parties that historically didn't cross connect in the past because they're transferring petabyte data back and forth. And you just can't backhaul that across the country over the Internet and things like that. And that's really what's driving demand.

People want to be in a neutral host colocation facility next to other people where they can get direct cross connects and that's really what's feeding this ecosystem growth and the cross-connect growth.

Operator

And our next question comes from Brendan Lynch from Barclays.

Brendan Lynch

Rod, you remain confident that 2026 is going to be the trough year in organic growth and also mentioned that AFFO per share should be accelerating in 2027. Is that dependent on contribution from the satellite operators that could deploy traditional networks? Or would that be incremental?

Rodney Smith

Yes. Thanks for the question, Brendan. I would say -- let me step back and just look at our AFFO per share growth algorithm that we've laid out in prior calls. I'll start by just highlighting that this year for 2026, in our most recent outlook, that has a 3% AFFO per share growth rate. On an FX neutral basis, that's much tighter to 0%. That includes contributions from DISH headwind of about 400 basis points. It also includes about 100 basis points of headwind from refinancing costs and the services step down is another 100 basis points of headwind.

So if you look at that approximately 0% FX-neutral reported AFFO per share growth for 2026. And you look at DISH churn as being nonrecurring, that adds 400 basis points sort of going forward on an average kind of run rate basis, we are moving through the refinancing curve and the refinancing headwind is becoming muted relative to where it was a couple of years ago. So we're growing into the higher interest rate environment, but that's another 100 basis points. And then the services reduction is another 100 basis points.

That puts us at an AFFO per share number equivalent to approximately 7% if you think of those items as nonrecurring and going forward. So outside of those items, our global portfolio of assets is really well positioned for that mid-single to better than mid-single-digit AFFO per share growth. And that comes with a few distinct expectations. One is the developed markets, both U.S. and Europe, we expect revenue growth in the mid-single digits from those businesses, very high-quality economies, very consistent activity levels. We expect higher growth from our emerging market tower businesses, and we expect double-digit growth from CoreSite as we move forward into the next few years.

You combine that revenue growth with our high operating leverage, our high financial leverage and very smart capital allocation and disciplined capital allocation, that makes it very attainable to have an AFFO per share growth in that mid- to upper mid-single-digit rates before you account for the impacts of FX and interest rates.

So that algorithm is fully intact. And I would say when you think of satellite players, we're not really building any of that into the long term. So there could be some upside there. If that's modest, I wouldn't expect a lot of changes to our overall algorithm. But depending on the nature of those investments and the ambition of any particular satellite carriers, some of that could certainly represent upside.

Brendan Lynch

Great. And Steve, just to follow up on your comment about the DISH equipment. I understand that you own it now. But is there any dispute about that in the legal proceedings with DISH? Or are you free and clear to sell that equipment to whoever you choose right now?

Steven Vondran

No, we are good. We can do whatever we want to do with it.

Operator

Our last question comes from Eric Luebchow of Wells Fargo.

Eric Luebchow

I wanted to touch on the data center CoreSite business a little bit. Maybe you can talk about any steps or things you're evaluating to get the higher multiple data center business even more reflected in your stock price today. Are there opportunities to ramp CapEx spending even further as we look into '27 or '28? Are there any inorganic opportunities out there, whether it's portfolios or specific assets that could drive an even higher contribution to your attributable AFFO? .

Steven Vondran

Sure. Thanks for the question. We are continuing to look for opportunities to invest in CoreSite. And if you kind of reflect back on what we said about our pipeline, we have line of sight into almost tripling the megawatts under service that we have today over time. And we'll continue to look at adding scale there. We are exploring some new market entries there from an organic build perspective. We haven't announced any yet because we're still working through the process of making sure that we've got all the entitlements on those before we announce that. But we do think that, that's an avenue.

We will continue to invest in our existing campuses. We're pushing the teams to accelerate that build pipeline. We have all the entitlements we need for our near-term build. But as you may have heard, it's a little bit difficult to accelerate some of the time lines where you don't have permits. So we're working hard on that. And we've got a 30 year track record on the tower side of navigating those headwinds. So we think we're better positioned than anybody else to do that.

But we are trying to accelerate that investment, and we do think that, that will be something that gains a little bit more attention for CoreSite. When it comes to inorganic opportunities, like we would do on the tower side, we'll always evaluate what's out there. And if we can find an opportunity that makes sense, then we'll take action on it. But nothing to announce on that front today. But we certainly believe in the business. We believe in the connection between towers and data centers. And we think that that's going to continue to evolve and become more apparent, and we think it's a great asset that we intend to continue growing. So thanks for the question.

Operator

This concludes our question-and-answer session and today's conference call. Thank you for participating, and you may now disconnect.

原文链接