Conferencia de resultados del 3T de 2026 de SBA Communications (SBAC): recompras, mayor guía de AFFO y pipeline de espectro

Conclusiones clave

• SBA Communications registró un AFFO por acción de 3,90 dólares en el tercer trimestre de 2026, un margen de flujo de caja de torres a nivel de toda la empresa del 80% y un margen de EBITDA ajustado del 69%.
• La empresa devolvió aproximadamente 430 millones de dólares a los accionistas a través de dividendos y recompras de acciones. Recompró 1,7 millones de acciones por 300 millones de dólares a un precio medio de 179,68 dólares, dejando 800 millones de dólares disponibles bajo su autorización.
• La dirección elevó el punto medio de las previsiones de AFFO para 2026 en 2 millones de dólares y la previsión de AFFO por acción en 0,08 dólares. Excluyendo las variaciones del tipo de cambio y las ventas en Chile y Ecuador, los incrementos fueron de 5 millones de dólares y 0,11 dólares, respectivamente.
• La facturación por nuevos arrendamientos y modificaciones a nivel nacional sumó aproximadamente 8 millones de dólares, mientras que la facturación por nuevos arrendamientos y modificaciones a nivel internacional alcanzó aproximadamente 5 millones de dólares.
• SBA acordó vender aproximadamente 1.800 emplazamientos en tres mercados de Sudamérica, completando su salida de la región a excepción de Brasil. Los activos representan cerca de 43 millones de dólares en ingresos anuales por arrendamiento y 20,5 millones de dólares en EBITDA ajustado anual.
• La directiva prevé que el nuevo espectro, incluida la banda C superior, impulse futuros despliegues de torres. También observa oportunidades a más largo plazo derivadas de la infraestructura terrestre vinculada a satélites y la computación distribuida de IA en los emplazamientos de torres.

Datos financieros clave

El consejo de administración declaró otro dividendo trimestral de 1,25 dólares por acción, pagadero el 10 de diciembre de 2026 a los accionistas registrados al 12 de noviembre de 2026. El dividendo es aproximadamente un 13% superior al del periodo del año anterior y representa una tasa anualizada de alrededor del 41% del punto medio de las previsiones de FFO por acción de la empresa para todo el año.

Desempeño operativo y del negocio

Arrendamientos e inversión en redes en EE. UU.

La actividad de los clientes en EE. UU. se situó ligeramente por debajo de la del primer semestre de 2026, pero se mantuvo en línea con las expectativas de la directiva. Los nuevos arrendamientos siguieron representando la mayor parte de la actividad a medida que los operadores abordaban las lagunas de cobertura y añadían capacidad. Los emplazamientos existentes también recibieron actualizaciones relacionadas con el 5G que incluyeron espectro de banda media, MIMO masivo y radios de mayor capacidad.

La directiva señaló que la actividad de los operadores sigue siendo desigual. Un operador se ha mostrado especialmente activo, otro ligeramente lento y otro muy lento. En general, los volúmenes de solicitudes se describieron como ampliamente en consonancia con el nivel medio de los dos últimos años, aunque la combinación de clientes ha cambiado.

La cancelación subyacente de clientes en el mercado nacional se vio afectada por factores importantes específicos de determinados clientes. La cancelación vinculada a Sprint fue de 16 millones de dólares en el T3, y aún se prevé que la cancelación de todo el año 2026 se sitúe entre 55 y 56 millones de dólares. La directiva prevé que la cancelación de Sprint caiga por debajo de los 20 millones de dólares en 2027. La cancelación vinculada a EchoStar fue de 14 millones de dólares en el T3 y se proyecta que alcance un total de 36 millones de dólares en 2026.

Excluyendo a Sprint y EchoStar, la directiva indicó que la cancelación normalizada en EE. UU. podría situarse con el tiempo en torno a aproximadamente el 1%.

Reorganización de la cartera internacional

SBA continuó con la revisión de la cartera iniciada a principios de 2024. Durante el T3, la empresa acordó vender aproximadamente 1.800 emplazamientos en tres mercados sudamericanos. Dos transacciones ya se han cerrado y la dirección prevé que la salida del mercado restante se complete a finales de año.

Las ventas en Chile y Ecuador redujeron las perspectivas para 2026 en aproximadamente 10 millones de dólares en ingresos por arrendamiento de emplazamientos, 6 millones de dólares en flujo de caja de torres y 4 millones de dólares en EBITDA ajustado. Las previsiones no incluyen la venta pendiente de las torres de la empresa en Perú, que sigue sujeta a la aprobación regulatoria y se prevé cerrar en el T4.

SBA construyó aproximadamente 210 torres durante el trimestre, más del doble de las 99 construidas en el T2, principalmente en América Central y África. La directiva espera que la construcción se mantenga en torno a estos niveles durante los próximos trimestres.

La empresa tiene previsto mantener e invertir en sus operaciones en África. La directiva señaló que Sudáfrica y Tanzania han generado sólidos rendimientos, y describió el retorno sobre el capital invertido como el más elevado de la empresa fuera de EE. UU.

Planes de despliegue de espectro

La dirección destacó la banda C superior como un catalizador clave de la demanda futura. Está previsto que la subasta comience en abril de 2027 y deba finalizar en julio. Esta pondrá a disposición 160 megahercios entre 3,98 y 4,14 gigahercios. Combinada con la banda C inferior, esto crearía 440 megahercios de espectro contiguo de banda media para uso inalámbrico.

SBA considera que parte de la actividad en torres vinculada a la banda C superior podría comenzar en 2028, potencialmente antes de la liberación completa del espectro. La directiva también espera que la actividad de despliegue a partir del espectro ya en posesión de los operadores, incluidas las frecuencias de 600 megahercios de AT&T y la banda C de T-Mobile, comience antes, potencialmente durante 2027. Los plazos y el impacto financiero siguen dependiendo de las decisiones de los operadores y de los acuerdos contractuales existentes.

Computación en la periferia (Edge computing) e infraestructura de IA

A través de SBA Edge, la empresa está aprovechando la energía, fibra, seguridad, casetas de equipos y terreno disponible en los emplazamientos de torres para respaldar la computación distribuida. Varios emplazamientos ya están operando con empresas que desarrollan capacidad de cómputo e inferencia de IA distribuida.

La dirección señaló que la mayoría de los despliegues actuales utilizan aproximadamente entre 200 y 300 kilovatios, aunque algunas instalaciones requieren un megavatio o algo más. Los requisitos de capital incremental se han mantenido limitados debido a que los proyectos utilizan espacio vacante y exceso de energía. Se prevé que los ingresos a corto plazo sean modestos, pero la directiva ve potencial para escalar el modelo en todo el conjunto de torres de EE. UU.

Previsiones de la directiva

SBA elevó el punto medio de sus perspectivas de AFFO para 2026 en 2 millones de dólares y las de AFFO por acción en 0,08 dólares. Excluyendo las variaciones del tipo de cambio y las desinversiones en Chile y Ecuador, los incrementos respectivos fueron de 5 millones de dólares y 0,11 dólares.

La dirección atribuyó la mejora principalmente a un menor gasto neto por intereses en efectivo, menores gastos de capital no discrecionales y un menor número de acciones tras las recompras del T3.

Sin los cambios en los tipos de cambio y las ventas en Chile y Ecuador, la directiva afirmó que sus previsiones para todo el año sobre ingresos por arrendamiento de emplazamientos, flujo de caja de torres y EBITDA ajustado se habrían mantenido prácticamente sin cambios.

Las perspectivas para 2026 excluyen las recompras de acciones adicionales realizadas después de la fecha de la llamada de resultados y cualquier impacto de la venta pendiente de las torres en Perú.

Se espera que la cancelación internacional de clientes siga siendo elevada en 2027, mientras los mercados afrontan la consolidación de operadores, la reestructuración y la racionalización de redes. La dirección prevé que comience a normalizarse en 2028 y estimó un rango de cancelación internacional normalizada a más largo plazo de aproximadamente el 2% al 2,5%.

Riesgos y aspectos a vigilar

• Los volúmenes de solicitudes de los operadores siguen siendo desiguales, lo que genera variabilidad de un trimestre a otro en la actividad de arrendamiento.
• La cancelación de clientes de Sprint y EchoStar sigue pesando sobre la cancelación nacional reportada, mientras que se prevé que la consolidación internacional mantenga elevada la cancelación internacional hasta 2027.
• La venta de las torres de Perú sigue sujeta a la aprobación regulatoria.
• El calendario de despliegue de espectro depende de las subastas, los procesos de liberación, la asignación de capital de los operadores y las decisiones de lanzamiento comercial.
• La computación en la periferia sigue en una fase inicial, con unos ingresos previstos modestos a corto plazo y una adopción por parte de los clientes aún en desarrollo.
• Los elevados tipos de interés generan incertidumbre sobre los plazos de las próximas actividades de refinanciación, aunque la directiva afirmó que el acceso al capital se mantiene sólido.

Puntos destacados del turno de preguntas y respuestas de analistas

Equipos de DISH y su posible reutilización

SBA contaba con aproximadamente 3.000 contratos de arrendamiento activos con DISH. La directiva señaló que los equipos asociados han quedado prácticamente abandonados y ahora están bajo el control de SBA tras las rescisiones de contratos vinculadas al litigio por bancarrota. La mayor parte del equipo permanece instalado pero no está emitiendo.

Varias partes han expresado interés en los equipos. Cualquier nuevo usuario tendría que negociar por separado con SBA, ya que los contratos de arrendamiento anteriores no se transferirían automáticamente.

Redes satelitales e infraestructura terrestre

Al ser preguntada por los posibles acontecimientos relacionados con SpaceX y DISH, la directiva afirmó que no disponía de información específica que confirmar. Sin embargo, SBA considera que cualquier proveedor que busque una cobertura móvil de alta calidad y ubicua requerirá componentes de red terrestre, lo que podría generar demanda de infraestructura de torres.

Asignación de capital

La dirección prevé que las recompras de acciones sigan siendo significativas pero de carácter oportunista. SBA comparará la rentabilidad disponible de las recompras con adquisiciones, la construcción de nuevas torres y otras inversiones. Teniendo en cuenta la valoración actual de los activos de infraestructura en el mercado privado, la directiva indicó que la continua presión sobre las valoraciones en los mercados públicos podría favorecer un mayor número de recompras frente a la compra de activos.

Refinanciación de deuda

SBA prevé refinanciar su valor respaldado por activos (ABS) de torres de 1.165 millones de dólares a un tipo estimado del 5,25%. Para los vencimientos de 2027, la directiva prevé una mayor transición desde la financiación garantizada mediante ABS hacia instrumentos no garantizados de grado de inversión, que podrían incluir bonos sénior o un préstamo a plazo (Term Loan A) sin garantía.

Revisión estratégica

La directiva afirmó que la revisión estratégica va más allá de las desinversiones internacionales y abarca todas las líneas de negocio y actividades. El proceso no ha concluido, aunque se han logrado avances sustanciales. Ha incluido tanto salidas de mercados como inversiones adicionales, como la expansión en América Central mediante la adquisición de Millicom.

Transcripción completa de la llamada de resultados

Transcripción completa de la conferencia de resultados

Comentarios de la dirección

Operator

Welcome, and thank you for joining the SBA Third Quarter 2026 results. [Operator Instructions]

With that, I'll turn the call over to Louis Friend, Vice President, Finance and Capital Markets. Please go ahead.

Louis Friend

Good evening, and thank you for joining us for SBA's Third Quarter 2026 Earnings Conference Call. Here with me today are Brendan Cavanagh, our President and Chief Executive Officer; and Marc Montagner, our Chief Financial Officer.

Some of the information we will discuss on this call is forward-looking, including, but not limited to, any guidance for 2026 and beyond. In today's press release, and in our SEC filings, we detail material risks that may cause our future results to differ from our expectations. Our statements are as of today, November 2, and we have no obligation to update any forward-looking statement we may make.

In addition, our comments will include non-GAAP financial measures and other key operating metrics. The reconciliation of and other information regarding these items can be found in our supplemental financial data package, which is located on the landing page of our Investor Relations website.

With that, I will now turn it over to Brendan to comment on the third quarter.

Brendan Cavanagh

Thanks, Louis. We had another great quarter with industry-leading AFFO per share, strong operating margins, continued progress on our portfolio review, over $430 million returned to shareholders through dividends and share repurchases, and a solid investment-grade balance sheet. In the U.S., the level of customer activity in the third quarter was slightly below what we saw in the first half of the year, but in line with our expectations. Our customers remain focused on investing in their wireless networks. The bulk of the activity continues to come from new leases as carriers address coverage gaps and add capacity.

In addition to new leases, payers are upgrading their existing leases with 5G-capable mid-band spectrum, including C-band and technological improvements such as massive MIMO and higher capacity radios. Fixed wireless access also continues to grow, with millions of new subscribers compared to just a few years ago. That growth, paired with steadily increasing mobile data traffic, is placing more and more strength on today's networks.

According to CTIA's 2026 annual survey released in September, in 2025, mobile data grew by 20% in the U.S. and expectations are for data demand to quadruple by 2032. Much of that fueled by AI and the continued growth in nonphone devices such as wearables and smart sensors. To help support this growth, we expect to see a new wave of spectrum come to market soon, including AT&T's 600 megahertz.

New spectrum has always been one of the most important drivers of our business. And looking further out, the spectrum pipeline continues to build, and I believe it is as strong as it has been in many years. The upper C-band auction is scheduled to begin in April of next year, with the FCC required to complete it by July. It will make 160 megahertz of spectrum between 3.98 and 4.14 gigahertz available and combined with the existing lower C-band create a large 440 megahertz of contiguous mid-band for wireless use. Whoever wins this spectrum, we'll have to put it into commercial service and putting it into service means new equipment on our towers.

Beyond the upper C-band, we see steady progress across several additional bands. Just lastly, the FCC proposed auctioning 25 megahertz of mid-band spectrum in the 1.6 and 2.0 gigahertz bands. Given the administration's plans to make up to 400 megahertz of spectrum available for commercial use by 2028, we expect proposed rulemaking for the 2.7 gigahertz and 7 gigahertz bands to come soon and continue to believe the FCC could auction the spectrum as early as 2028.

Additionally, the 4 gigahertz band is now being studied for full power commercial licensed use. And as the industry moves towards 6G, we expect a more balanced uplink and downlink network architecture, driven in part by AI applications, sending more data from devices back into the network. Each major spectrum auction over the last 2 decades has been followed by a meaningful deployment cycle on our towers. And our priority is to be ready for the next one with the right sites, the right people and the right processes to help our customers deploy quickly.

While the timing of carrier activity will move around from quarter-to-quarter, the long-term direction is very exciting. We believe this spectrum will be the foundation of sustained organic leasing growth well into the next decade.

Beyond our core customers, we see a real opportunity to expand our customer base to both satellite and edge providers. With regard to satellite providers, our view has not changed. Satellites are a complement to terrestrial wireless networks. This view is supported by a satellite provider's recent purchase of low band 800 megahertz spectrum, which is well suited for in-building coverage and enhanced capacity from the ground. We are confident that any provider looking to deliver ubiquitous, high-quality mobile service, particularly [indiscernible] loads that require low latency, storage capacity and a general need to sit closer to the end user.

The traditional data center industry is facing challenges around power availability, permitting and long construction time lines. Our tower sites are well positioned to address many of these needs. We already have power, fiber and security in place. Our sites are zoned and in many cases, they have existing shelters or ground space at the base of the tower that can be readily put to use. Through our SBA Edge entity, we offer access to the power and infrastructure needed so that AI customers can deploy quickly and efficiently.

We have several edge sites up and running with a handful of different companies building out distributed AI and print capacity. So far, the incremental capital remains minimal as these edge deployments use vacant space and excess power at our tower. It is still early, and the immediate revenue contribution will be modest, but the economics are very attractive, and the opportunity for much greater scale across our portfolio of U.S. towers is real. These initial sites give us an early framework for bringing compute to our sites. I'm excited about the potential of this business, and I look forward to updating you on our progress.

Internationally, we continue to execute on the portfolio review we began in early 2024. Our objective has been clear from the start to be a leading tower company in each market where we operate, to align ourselves with the leading wireless operators in those markets and to improve the overall durability of our cash flows. During the third quarter, we agreed to approximately 1,800 sites in 3 South American markets, representing a complete exit from that region outside of Brazil. Those 3 markets represent approximately $43 million of annual leasing revenue and $20.5 million of annual adjusted EBITDA.

Two markets have already closed, and we expect remaining market to close by year-end. This follows our exits from the Philippines and Colombia in early 2025 and from Canada last October. The steps we have taken allow us to have a more focused geographic concentration internationally. Our international strategy is focused on locking in long-term stable cash flow.

Internationally, we also built 210 new towers in the third quarter, more than double the 99 we built last quarter, primarily in Central America and Africa, and we expect our build program to remain at these levels over the coming quarters.

Turning now to capital allocation. It was a busy quarter. Following our inaugural investment-grade bond offering in July, which allowed us to fully repay our revolver and term loan, we resumed share repurchases, and we did so in a meaningful way. Since our last earnings report, we have repurchased 1.7 million shares at an average price of $179.68 per share for a total of $300 million. And as of today, we have $800 million remaining on our authorization.

Our priority is to allocate capital where it will generate the highest risk-adjusted return for our shareholders. And today, we believe our stock is a great value. With third quarter leverage at 6.4x net debt to adjusted EBITDA below the midpoint of our target range of 6x to 7x, we have the balance sheet capacity and cash on hand to opportunistically retire more shares. We believe in the strength of our business, our future growth potential and our ability to execute, and we will continue to allocate capital where we believe it will create the most value for our shareholders over the long term.

As I look ahead, I'm excited about where our industry is going. The network -- sorry, the networks being built today are only the beginning and 6G is around the corner. Add to that, the largest pipeline of new spectrum in many years new networks being built to work alongside satellites and AI workloads moving to the edge. SBA is well positioned for that future. We have a high-quality portfolio, an investment-grade balance sheet, a disciplined approach to capital allocation and the best team in the industry.

And finally, I'd like to thank our team members around the world. Their hard work and commitment are what allow us to pursue our vision to be our customers' first choice provider and the industry leader in quality infrastructure solutions, and they deliver on it every day. I'd also like to thank our customers for their continued trust and partnership and our shareholders for their ongoing support.

With that, I will now turn it over to Marc to comment on the third quarter results and 2026 outlook.

Marc Montagner

Thank you, Brendan. Our third quarter results were in line with our expectations and if not for changes in FX and the sale of our operations in Chile and Ecuador, our full year 2026 outlook for site leasing revenue, our cash flow and adjusted EBITDA would have been mostly unchanged. The sale of our towers in Chile and Ecuador reduced our full year outlook by approximately $10 million of site leasing revenue, $6 million of total cash flow and $4 million of adjusted EBITDA.

We're increasing the midpoint of 2026 guidance for AFFO, FFO per share by $2 million and $0.08, respectively. Excluding both FX and the sales of our operations in Chile and Ecuador, the midpoint of AFFO and FFO per share is up by $5 million and $0.11, respectively. These increases in AFFO and AFFO per share, mostly driven by lower net cash interest expenses, lower nondiscretionary capital expenditure along with the benefit of a lower share count from our third quarter share buyback.

Our outlook does not yet include any impact from the pending sales of our tower in Peru, which remains subject to regulatory approval and is expected to close in the fourth quarter.

In the third quarter, we continue to operate efficiently, controlling different costs and achieving company-wide tower cash flow margins of 80% and an adjusted EBITDA margin of 69%. AFFO per share was $3.90 and we paid a cash dividend of $1.25 per share.

Operationally, we added approximately $8 million of domestic new lease and amendment billings in the quarter. Spring-related churn was $16 million in the quarter, and we continue to expect $55 million to $56 million for the full year. We expect pre-churn in 2027 to be less than $20 million. EchoStar related churn was $14 million in the third quarter, and we expect $36 million for full year 2026.

Internationally, we added approximately $5 million of new lease and amendment billings in the quarter. We are seeing healthy leasing demand broadly across our international portfolio. The midpoint of our full year outlook for international churn remains at $38 million as key areas work through consolidation, restructuring and network realizations. We expect international churn to remain elevated in 2027, but to begin to normalize in 2028 when the impact of carrier consolidation is behind us.

Turning to investing activities. During the third quarter, we did 213 towers, and we spent $108 million to acquire 7 sites. We also spent $13.7 million to purchase land, Eastman and extend ground lease terms. Total cash expenditure for the quarter was $105.2 million which 14.1% was nondiscretionary.

Moving to the balance sheet. I'm very pleased with the progress we have made as an investment-grade company. In July, we issued a noncore $3.5 billion investment-grade bonds. In addition, on September 25, we established a [ $2 billion ] commercial paper program. These programs give us another flexible low-cost source of short-term liquidity. As of today, we have no amount outstanding under either our revolver for the commercial pay per program.

As we approach the maturity date on our $1.165 billion ABS tower security, we anticipate a complete refinancing of those securities at an estimated rate of 5.25% in the coming days. Our access to capital remains very robust and we will continue to look at ways to reduce our overall cost of debt. Our 2026 outlook does not include the impact of any additional share repurchase beyond those completed for the date of this call.

During the third quarter, we declared a cash dividend of $131.3 million or $1.25 per share. For our third quarter share buyback and dividend payout, we returned approximately $430 million of capital to our shareholders. And today, we will now our Board of Directors declared a quarterly dividend of $1.25 per share, payable on December 10, 2026, to shareholders of record as of the close of business on November 12, 2026. This dividend represents an increase of approximately 13% over the dividend paid in the prior year period in an annualized rate of approximately 41% of the midpoint of our full year FFO per share outlook.

Let me now turn the call over to the operator for Q&A.

Operator

[Operator Instructions] All right. Let's go to our first caller, Ric Prentiss with Raymond James.

Preguntas y respuestas

Ric Prentiss

A couple of questions if I could. Obviously, number doll and Tower land, rumors today, could there be a space and a dish Walk us through a little bit about what your thoughts as far as what the future might hold. And I got to admit, universe of 1 here, I have a hard time imagining an MVNO-only strategy and wanting to achieve 60%, 70% EBITDA margins. But just help us kind of unpack what you think might be happening there?

Brendan Cavanagh

Yes. Well, first, before I do that, Ric, let me just say, I think this is your last earnings call, and you've probably been on every earnings call that SBA had over the last or so. So we're going to miss you, and thank you for all of your service over the many years. We've enjoyed working together.

So to your question on SpaceX and DISH, I mean, obviously, I have no specific intel or insight into exactly what they may or may not be doing. So it's not really my place to comment on that. You should look to those parties to comment. I think if there's truth to it, I think what's good from our perspective is that it is an indicator -- another indicator, another data point that suggests the necessity of terrestrial components of any network that's going to be put out there to provide mobile coverage.

And that we feel like is ultimately going to be a good thing for us if that's the way that it goes, and that's our belief on what would be necessary for anybody that would look to provide that kind of coverage. So beyond that, it's hard for me to say what their specific plans are and whether there's any truth to that or not, but I would say ask those parties for their comments.

Ric Prentiss

Sure. It's actually hard for me also a vision what is bad news for having that SpaceX might be doing here. One other question related. We keep getting questions on the potential of DISH former sites. Can you let us though how many DISH sites you had actually gotten on air? And if, as we call it, the equipment formally known as DISH, is that still on your towers? And if so, has it been abandoned? And do you guys have done enough in on?

Brendan Cavanagh

Yes. We had approximately 3,000 leases that were on air with DISH. And the -- in terms of the equipment, it's basically been abandoned, and we do own it as we terminated those leases as part of the bankruptcy litigation. So it's there, and it's in our control at this point.

Ric Prentiss

Great. And one final one for me. Any thought on the litigation and the arbitration process? And does it feel like we might be getting close to something that it feels like maybe the industry is getting closer to understanding how that escolar might be divvied up, but it's complicated I get it. Just wondering any updated time line or thoughts on that.

Brendan Cavanagh

Yes, it's getting complicated. I mean, I guess, every day that goes by, we're a little closer to some resolution. But at this stage, we're in a stage of mediation. I think that's pretty well known as an industry group. And that's the process. We'll see how it shakes out and whether I can't guess exactly where that's going to end up. But it does allow for at least a dialogue to take place, and we'll see what the results of that are. But it's a little premature at this point, Ric, to say what might or might not happen.

Ric Prentiss

Great. Well, certainly, never dull. And my last 1 was that you guys was certainly no difference there, but I appreciate working with you guys all the years as well.

Operator

Moving on to our next call are Batya Levi, UBS.

Batya Levi

Can you provide a little bit more color on the application volume that you're seeing from the carriers? And if there is any indication that there could be an inflection in activity into next year? And maybe just a little bit more on your conversations with CSX. Would you characterize them as they have picked up more recently? And to the extent that you have to some look into what they might be looking to do, how quickly do you think the tower companies could benefit from that?

Brendan Cavanagh

Yes. On the application volume with the carriers, it's been -- I guess the way I'd characterize it is that each of the carriers are in a little bit of a different phase. And sometimes that happens where certain ones are perhaps more active than others, and that's kind of the environment that we're in right now. We have one that's been particularly active with us and one that's been modestly slow and 1 that's been really slow. And so the application volumes kind of took that as well.

So on an overall basis, our application volumes are, I'd say, relatively the same on average as they've been for the last couple of years. But the mix in terms of the makeup is a little bit different. And so we'll have to see how that continues to develop as we go into next year. And obviously, it's a little premature to talk about next year's numbers. I think next call, we'll be able to tell you exactly what we're thinking about that.

On the SpaceX piece, I can't really comment on any specific conversations that we have other than just to say that we have conversations with anybody that has an interest or a use for the infrastructure that we offer and the services that we offer. And you should expect that we talk to all parties that are interested in that. And that's what we will continue to do.

I mean timing, I would kind of refer you to their comments about what they're thinking about in terms of time line to build networks if they plan to go that route as a guide. But I can't really say at this stage what the time frame would be. Obviously, if anything ever really happened, at that point, we'd have something to talk about that's tangible.

Operator

Let's move on to Richard Choe, JPMorgan.

Richard Choe

I'll say Ric is the universal one for tower coverage. So [indiscernible]. Just following up, though, like where -- given where we are with DISH and the process, is there any way that another entity could back into leases or the equipment or adjustments? That the rights they had? Or was it just -- is it better to assume that it's just going to be a new negotiation with new priorities and new types of equipment?

And then it seems like we're in an extended period where there's been a lot of spectrum and efficiency on technology. But in the past, given the long history of towers and wireless technology, do you expect that to kind of pass at some point and this to pick up, and it seems like it's pretty long in the in that sense?

Brendan Cavanagh

Yes. I'm sorry, Rich, on your second question there, you're talking about specific technology development.

Richard Choe

Just where we are today with 5G and -- have we seen this before? Have we gone through this before? And do you think it's different this time versus each other with 3G and 4G, there were kind of pauses in 1 time?

Brendan Cavanagh

Yes. So I'll take that one first because we ask for a clarification, and I'll go back to the other one on DISH. I think that there are -- as an overarching generational technology rollout, they tend to follow similar patterns. And I think that this isn't materially different in that regard. But each one has nuances that make them a little bit different.

And I think with 5G, some of the development of the applications and the timing of their availability and uptake by the end subscriber has perhaps been a little bit different than it was with 4G, for instance. And as a result, the behaviors of the carriers around the deployment of those networks have perhaps shifted a little bit.

But overall, the same dynamics exist. You kind of have an early significant rush to kind of go out and overlay existing networks with new technologies and then you have infill that takes place. And that's sort of where we are today. And we're seeing that, particularly with one of the customers that we have, we're seeing significant infill taking place. A lot of the growth that we're seeing today is new leases, and that's mostly infill and some expanded coverage. So I don't think it's that different than the others. I think other factors in terms of even cost of capital sometimes have an impact on the timing of deployments, but the general gist from a technology standpoint is the same.

In the case of the DISH question and what could or couldn't happen. I mean at this stage, it's hard to say for sure what might happen, but the equipment, as I said earlier, is now ours and within our control. And so anybody that wants to use that would obviously have to deal with us on that and the existing lease agreements wouldn't necessarily apply to a new party that would come in. So that would have to be just a separate discussion if somebody wanted to use it. And we have many parties that have expressed interest actually in that equipment. So we feel like there may be some opportunity, but we'll see.

In terms of other rights that may exist around stuff that DISH had, I can't really speak to that. That would be between the other parties in DISH. And obviously, they're in the middle of a bankruptcy proceeding as well, so that I'm sure that would complicate discussions too. So a lot remains to be seen. But I'm sure, over time, we'll see these things work out one way or another.

Operator

Moving on to Nick Del Deo from MoffettNathanson.

Nicholas Del Deo

First, with respect to the edge computing opportunity, you talked about Brendan, can you share what sort of experiences you've had with the utility companies when you go to them asking for power augmentation to a site?

Brendan Cavanagh

Yes. It's still somewhat early days for all of this stuff, Nick. And in many cases, what we've had available at the sites where we've done these smaller edge computing facilities has been adequate already. So we haven't always had to go back to the utilities. But in cases where we have. It's been -- we've gone through kind of a normal process that we would go through at any stage when building a new tower, for instance, and we're trying to bring power there. It's really not been materially different than that. So I wouldn't say there's anything that stands out. It's just a typical process of applying for the incremental power needs at a site and going through the steps at that particular utility requires. So nothing really to call out that's any different than what we usually do.

Nicholas Del Deo

Okay. Okay. And to confirm, to the extent that you're supplying power to these customers there, are they're planning, call it, couple hundred kilowatts. Is that the sort of magnitude you're dealing with?

Brendan Cavanagh

Typically, yes. Typically, it's -- I would say it's 200 to 300 kilowatts roughly. But we have some facilities that are a megawatt or even slightly more. So most of them though will be in the range, the 200 to 300-kilowatt range.

Nicholas Del Deo

Okay. Okay. Great. And then maybe switching to U.S. churn. Obviously, your underlying churn is much lower than your headline churn. I think it's about 1.4% if we take out Sprint and EchoStar. If we look at your peers, they're probably around 1%. Should we expect to see your churn kind of converge at that level? Or are there reasons why we should expect your return to remain different from what others are saying?

Brendan Cavanagh

Yes. I think that, that's a reasonable expectation. When I get the detail and I get all the details of churn, as you might imagine, and obviously, we have the 2 big ones that kind of distort things right now. But the rest of it, it really is a list of cats and dogs. I mean there is all kinds of small stuff that's going on there. Most of it is not related to our core big customers. There's occasionally some here or there. We do have some U.S. Cellular T-Mobile overlap, for instance, that's in our churn right now that started.

So there's some of those things, but I think over time, I would expect to see that start to really fall off because there's really only so much of it -- so many of those customers that exist and there's not even that many of those smaller carriers that have some level of overlap that's existed in the past that remains. So I would expect somewhere in the 1% range to be a normalized level after we get through these bigger Sprint and DISH numbers.

Operator

Okay. Let's move on to Michael Rollins from Citi.

Michael Rollins

Just a few follow-ups. So the first is, I was just thinking about your answer earlier about. I think you said about 3,000 sites with DISH before they stopped. And I think you mentioned it was about $56 million of annualized rent. So just in very rough terms, it's a little under $1,600 a month, recognizing there's probably some rounding and stuff in there. But just curious, like, is that the right kind of Zipcode for that number?

And how do you think about rent for new colocation? Whether it's for -- with respect to that number, like whether it's should rents for new colo from the big 3 carriers typically trend significantly above that? Are there reasons for that? And if you have a new entrant coming in, is it -- do you get like a new entrant discount, so we should expect it being more like $1,600? Or were there like specific circumstances that drove the rent to where it was? And I have a couple of other follow-ups, if I talk afterwards.

Brendan Cavanagh

Yes. I mean there's a variety of factors that go into setting the rents. I mean one of the biggest ones is what are they actually doing? What's the usage at the tower sites. So as you might imagine, when it comes to DISH, when they were doing their deployments, they were materially less in terms of the capacity of the towers that they were taking up, the size, weight, surface area, all of that at the tower site and even on the ground was much, much smaller than the incumbent MNOs have at our towers on average. So that obviously results typically in a lower rent than you might see for a normally sized incumbent based installation because the capacity is just simply less. Also the height on the tower that they've taken if they're materially at a lower spot that sometimes has an impact as well.

But beyond all that, the way that the agreement is negotiated also matters. There's other terms that can have an impact as well. And some of those, I mean, for instance, you know we have an MLA with AT&T that we signed a few years ago. Well, the structure of that was very different and certain things were prioritized both by AT&T and by SBA that resulted in a structure and a price point that was different than might have happened at a different point in time. And so with the -- in the case of DISH, when we negotiated our agreement with them, certain minimum commitments that they might have made affected pricing certain terms that they agreed to, volume commitments, those sorts of things all have an impact.

And I would expect with anybody that we negotiate with on a master agreement type of level, there's some modest trade-off of volume minimum commitments for price. But it really -- there's so many factors that go into it that it's hard to say. But the basic one is the first one I mentioned. What are they actually using at the site? How much capacity they are taken up and that's probably the biggest one in terms of having an impact.

Michael Rollins

Really helpful. And just the second question is, just given where is in terms of just the relative maturation of the business, when you look at the amount of annual discretionary cash flow or AFFO that you generate, do you think that your company is in a better position to maximize shareholder value in the public markets? Or do you see a case where value maximization over time could be better in the private markets?

Brendan Cavanagh

That sounds like a trick question, Mike. I think that depending on a moment in time, valuations sometimes vary between public and private markets. And all I can tell you is that when we look at private assets in the U.S. specifically today and granted, they're smaller and so perhaps can be more competitively a bid after. The valuations are much, much higher than where SBA and the other public companies for that matter, are valued. But there was a time in our history here where I remember it being the opposite as well.

I think all of that is a little bit of trying to guess at a point in time. It doesn't really matter. We are a public company. We have the balance sheet that we have. It provides us plenty of flexibility. We do produce a lot of AFFO. So we have a lot of capital to put to work into the company, and I think it provides us the opportunity to use that capital effectively to create value for our shareholders, some by returning it directly to shareholders as we did this past quarter and in other cases, by investing it into incremental assets that we think are going to drive outsized returns. -- by being smart about how we allocate that capital. So I think there's plenty of opportunity for us ahead, and I look forward to continuing to run this business in a way that generates a lot of returns for our shareholders.

Operator

Moving on to Brendan Lynch from Barclays.

Brendan Lynch

I wanted to touch on international markets. So you exited everything in South America other than Brazil. Can you discuss your appetite for maintaining your exposure in Africa? And what the primary differences you see are in the African markets versus the South American markets that you've exited?

Brendan Cavanagh

Yes. Well, for -- just in terms of SBA-specific portfolio, we were much lower scale in the Andean markets that we just announced our exit from than we are in the 2 African markets where we operate. So we have -- we definitely have a greater scale in those 2 African markets. That's one difference.

The actual, I mean, the returns in those -- in both South Africa and Tanzania have been very, very strong for us. The return on invested capital is, frankly, the highest in the company outside of the U.S. and really that's only because the U.S. has a much longer standing history and much lower entry point back at the beginning. But it even rivals that in terms of incremental investments today, it's actually been better. So as long as we see our way clear to having a meaningful position in the markets where we're operating and we have relationships with the top carriers in those markets where we can feel comfortable in their future and their viability going forward and we can generate strong returns, we're happy to continue to operate and invest in those markets.

And particularly in Tanzania, you probably can see in our numbers that we built a lot of sites there this past quarter and really this year-to-date, and we continue to invest there because we see great opportunities. So there's some differences on an individual market basis but also on the existing portfolios that we have in those markets.

Brendan Lynch

Great. That's helpful. And you mentioned international churn should remain elevated in '27, then kind of normalize in 2028. Looking back over the past couple of years, it was around 5% to 7%. So maybe just help us think about what is a normal level of churn in international markets after you kind of get through this period.

Brendan Cavanagh

I would -- it's hard for me to say exactly because now we're looking out a little ways. But I think I'm going to take a guess, which I might be held taking a task for later. But I would say somewhere in the 2-ish percent, 2% to 2.5% range would be a normal level.

And the reason it's a little bit higher than, I think, the normal level here in the U.S. is that there's still a level of immaturity that exists in those markets and there's still some level of rationalization, I think, with smaller carriers and so forth that's taking place, and that has to kind of work through the system. So that's the only reason I think it would be a little higher than the normalized U.S. churn, but that's the number I'd circle.

Brendan Lynch

Okay. That's helpful. And then maybe just one other one on edge. Can you give us some sense of how you think the pace of edge deployments might scale and maybe a little bit on the nature of the companies that are doing trials. Are these start-ups or hyperscalers, neoclouds, et cetera?

Brendan Cavanagh

Yes. I think we will see a very steady level of growth there. I think as we mentioned in our prepared comments, the contributions -- the financial contributions at this stage are relatively modest and probably will be for the near term, but I think the pace of growth will be at a healthy pace.

The types of customers that we're seeing. There's a mix of start-ups, I would say. Neoclouds are probably the primary users that would be interested in the types of facilities that we have because what we're really talking about here is the same kind of provisioning of compute power for, in particular, AI inferencing applications. The value of this is that it's able to get on air and available faster because of the distributed structure of it, where power behind-the-meter power is available more quickly. And while it's much smaller, tinier chunks, the accumulation of those and the fact that you can do it quicker can get you to meaningful contributions and there's a lot of demand for that today.

So I think who the end users are, who the customers are, and we have several different companies that we're engaged with today, I think that will continue to evolve over time. And it will become clearer who's kind of doing the best with that. And they'll start to have a higher -- bigger presence with companies like us as we help them roll it out.

Operator

Let's move on to Cameron McVeigh, Morgan Stanley.

Cameron McVeigh

Marc, I was just curious how we should think about the margin expansion opportunity over the next few years. You're buying land and ground leases, and your competitors have outlined a target for something of 200 to 300 basis points of expansion by 2030. So just curious how you're framing the opportunity and what the puts and takes might be.

Marc Montagner

Well, I think we have been doing land buyout for probably 15 years now. And I think we have been very aggressive earlier than our competitors. It's a way to derisk our business going forward, but also given the multiple arbitrage a way to increase margins over time. So you should expect us to keep doing this, both in the U.S. and internationally.

Brendan Cavanagh

Yes. And I think that actually applies generally, not just on the land buyouts, but our overhead base, our cost base. We've actually been a leader in our industry with regard to those items. And that's good. And I think there's a lot of attention being paid to it by our peers, which fine. But I think what that means is they have a little more opportunity to squeeze some of that out than we do because I feel like we've actually operated as efficiently as anybody. Despite being the smallest company in the industry, we have been actually the most efficient from an overhead perspective. So we'll continue to do what we've always done.

Cameron McVeigh

Got it. That's great. And then secondly, just one on capital allocation. We repurchased $300 million in the quarter with $800 million remaining under the authorization. Just curious how you're thinking about the hurdle rate for buybacks versus some of the acquisitions, new builds, that reduction?

Brendan Cavanagh

Yes. I think we did a large amount of buybacks this quarter. I would expect us to continue to be a meaningful share repurchase. But really, in many ways, it's opportunistic in terms of where we're valued relative to alternative uses of capital. And we have historically always had a mix of investing in assets as well as doing buybacks. And I think we'll continue to see that going forward. If our valuation remains under pressure as it's been relative to the value of assets I mentioned earlier in response to a different question about the high valuations of assets, particularly here in the U.S., I mean if that dynamic remains, I would expect we'll do more buybacks, then we will do new asset purchases.

But ideally, we'd like to do a mix of both, and we're going to have to just work harder to kind of find those selective opportunities we feel good about in terms of new assets. But it's really a comparative analysis.

Operator

Let's move on to Eric Luebchow, Wells Fargo.

Eric Luebchow

Brendan, you talked a lot about all the new spectrum that's coming to market, but I think some people still have questions around the time line of when you'll see activity on the tower side between AT&T 600, T-Mobile C-band and then the new auction spectrum that's coming -- starting with upper C-Band next year. So when do you think we'll start to see that activity on the site? Do you think it could come as early as 2028 sooner than that? Or will it be pushed a little further out?

Brendan Cavanagh

Well, I think that certain bands, we will start to see some activity in all likelihood in 2028. I mean it's hard to say for sure. But for instance, if you took the first upcoming auction that's ahead of us, which is the upper C-band, that's going to take place and be completed by middle of next year. Obviously, there's a whole process to clear that spectrum. But what we've seen in the past is some level of activity that actually runs ahead of that clearance process in order to be ready so that at the time that it's cleared, and I think there will be an effort to accelerate that clearance by the way. that leads to some level of activity around our sites that happens earlier.

So I'm hoping that we can see that by some time in '28. In terms of the spectrum though that's already in the hands of some of the parties like the 600 megahertz that AT&T has or the C-band that T-Mobile has, I expect that we'll definitely start to see those before that time frame. Hopefully, some of that will start to see next year. It's hard for me to say for sure, you'd have to ask the carriers, but I know AT&T has been vocal about starting to get out there and roll out their 600 megahertz. Its impact on us. will have to be seen. We obviously have an MLA that will have some level of restriction on how much we see out of that, but that MLA is up in about 1.5 years. So I think in some ways, the longer it takes maybe the better from our perspective.

Eric Luebchow

Great. And then just one final one for me. So you talked about handling the $1.1 billion debt maturity in the next couple of days at a 5.25% rate. As we look at maturities for 2027, I guess, given there's been a lot of volatility in interest rates the last couple of months, I mean, how are you thinking about kind of the refinancing wall that you have next year in terms of how you want to refinance it with like-for-like instruments or paying down some of the debt principal. What are kind of the options on the table?

Brendan Cavanagh

Sure. Yes. There's definitely a number of instruments that are coming up soon. In terms of the types of instruments we use to refinance it, in most cases, would not be like-for-like because a number of those maturities are ABS securities and with the shift to investment grade, you will see a little bit more of a shift towards unsecured debt. And obviously, those are secured instrument securities that will be maturing in many cases. And so with those, that would likely be replaced with some sort of unsecured instrument.

We have an unsecured senior note that is maturing in the first part of next year, during the first quarter. So that would be replaced with a similar unsecured senior note, I would expect. Although, again, it would be IG-rated and instead of the high yield instrument that it was before. We look at other potential securities that might work, too. They would all likely be unsecured instruments, for instance, perhaps an unsecured term loan A is something that we're looking at as well.

Really, what we're dealing with here is not so much what instruments to use. It's really the timing, and we're trying to manage this as finely as we can because, obviously, rates are elevated. It's not a certainty as to when that might change or if it will change. But if we can go a little bit shorter dated in terms of the tenor of these instruments and give ourselves an opportunity to perhaps catch a better window on refinancing, that's something that we ideally would like to do. But you can never beat some losers there on the timing. But in any case, we have plenty of access to capital, and we feel very good about refinancing. We'll just try to manage it as efficiently as we can.

Operator

Next caller is David Barden from New Street Research.

David Barden

Brendan, we get a lot about the auction spectrum that's coming up there's actually quite a bit of spectrum that's held in private hands that's coming up with build-out requirements that really start in the 2027 to '29 time frame. Specifically, I'm thinking about the 600 megahertz and 3.5 gigahertz. Have you seen conversations happening around those bands, people entertaining those possibilities of coming in to strategics and taking advantage of that opportunity?

And then in an adjacent question, similarly, in the 900 megahertz, it's a little bit more speculative, but there's also a lot working there as well. So kind of augmenting the entire spectrum buildup picture. There are a lot of other places, too. I was just wondering what, if anything, you're seeing there.

Brendan Cavanagh

Yes. I can't say too much about that, David, but I will say that in -- with some of those spectrum bands, we have been seeing some applications around those. But I don't know if there's enough there to point to as a holistic deployment yet. So I guess I'll be able to come back to you more on that as we get a little bit down the line. And then there -- some of the bands that you mentioned that we haven't really seen anything to date. So I would imagine that some of those parties are thinking about what their next steps are with regard to.

It's hard for me -- and this is just me speculating, I really have no idea for sure with any of these folks. But my belief is that it's hard to imagine that some of the parties would actually have any interest in actually building out something other than to do what they need to, to preserve their spectrum. So I think the FCC has become a lot more strict about deployment requirements around spectrum, the parties are holding and the penalties have become more severe. So I think you'll see a lot more firm decisions being made. People are either going to build it out and use it or they're going to sell it.

David Barden

Right. Sell it. And then it gets built out.

Brendan Cavanagh

Right. That's an important part, right, is that eventually, that environment is very good for us because whether it's those parties or somebody else and maybe better somebody else who's actually serious about it, it's a good dynamic for us to have because it will get built out.

David Barden

And so I just wanted to follow up Brendan, on a question I asked a couple of quarters ago, which was in response to what the decisions that DISH had made regarding its leases? Why a company like SBA wouldn't have sent out an army of engineers to go snip wires and start ripping down gear? And at the time, you said, "Well, you don't know what we're doing." And so now that you own whatever is left of whatever this is, could you tell us what have you done and what is left?

Brendan Cavanagh

Most of the sites still have the equipment left at them, David, but they're not actually -- there's nothing being broadcast through that equipment. So it's actually there. And to the extent that it is useful to another party that would find it helpful in their own deployments. And as I said, there are several that have actually contacted us in that regard. It's available, and it's something that we obviously can work through a discussion on how that would work.

Operator

And our final question from Matt Niknam, Truist.

Matthew Niknam

Brendan, can you hear me?

Brendan Cavanagh

Yes, I can hear you. Go ahead.

Matthew Niknam

Okay. Great. Great. So my question is more around the strategic review. So I know this was initiated a while back. Is the strategic review process largely done? Are there other regions or businesses you'd consider exiting? Or are there other initiatives you'd look to target across the businesses?

Brendan Cavanagh

Yes. I mean the strategic review has tended to kind of become a name for international divestitures, but that's not really what it is. That's just one component. I mean it was really a review of all aspects of our business, which is what I think you're alluding to. It wasn't just international markets. it was other smaller business lines that we're in and other things that we're doing. We don't necessarily come out and talk about all of that on our earnings call. But we're looking at a wholesale review of everything that we do. We have done over the last several years. I wouldn't say it's done, but we've made a lot of progress. And so as we've gone through and check things off, I would say there's certainly less to do.

As it relates to the international markets, we've clearly exited a number of markets. We also, as part of that, made the decision to expand meaningfully in Central America, and we did that the Millicom acquisition. That was really part of it as well. So the decisions were not all about shrinking. They were also, in some cases, about investing more.

We continue to look at a couple of other things that we're involved with as well, and we'll see what the end result is. sometimes it's not just about the decision that's made, the steps to actually kind of effectuate that decision can take some time as you've seen. And so we'll continue to take the steps that we're taking. But I feel good about the direction that we've gone and our ability to now focus and prioritize the things that we think are going to be the most value creative for our shareholders.

Thank you all for joining the call, and we look forward to reporting our results next year.

Operator

Thank you to our speakers and everyone in the audience for joining us today. The call has concluded, and you may now disconnect.