As you describe, mobile phone signals today primarily rely on terrestrial hardware attached to the towers owned by Cellnex.
It's been a cash cow of a business over the past decade or more.
But is that changing with the rise of low Earth orbit (LEO) satellites. These satellites orbit much closer to Earth than traditional satellites, enabling them to communicate directly with ordinary mobile devices using standard mobile spectrum bands. Reducing costs are making it feasible for satellites to act as "cell towers in space," delivering signals direct-to-device to regular 5G smartphones without specialist hardware.
In the short-term, LEO satellites are expected to complement, not immediately replace, terrestrial towers. But in the medium-term we will likely see a growth in satellite networks, especially as 5G and 6G technologies mature, offering seamless global coverage and ultra-low latency.
Elon Musk has his Starlink satellites and Jeff Bezos is launching 3,500 satellites as part of Amazon's project Kuiper. These are just two of many. Is the Cellnex model about to be disrupted?
Fair question about satellites - I've spent a bit of time looking into this.
It seems like the satellite disruption thesis falls apart when you examine the physics and economics. LEOs simply can't match tower capacity in populated areas. Think about it: a single urban tower handles thousands of simultaneous connections in a small radius, while Starlink's entire constellation might support maybe a million concurrent users globally.
Where does Cellnex make most of its money? Urban and suburban areas - exactly where satellites struggle most with capacity density. It's perfect for someone that wants fast internet in the mountains or in the middle of nowhere. It's expensive to get towers there for just a few connections.
Let's look at the economics as well. Satellites cost less upfront but need replacement every 5-10 years. Towers cost more initially but last decades with minimal maintenance. The lifetime math heavily favors towers.
At the end of the day, mobile operators continue signing 20+ year tower contracts despite being fully aware of satellite developments. They understand network architecture better than anyone, and their capital deployment speaks volumes.
What we'll likely see is integration rather than replacement. Satellites will excel precisely where tower economics fail - remote regions with sparse population. For Cellnex's core urban assets, physics and economics remain firmly on their side.
It's less disruption and more ecosystem evolution - each technology finding its optimal niche.
I agree with your conclusions - I was only playing Devil's advocate.
In fact, I am invested in Brookfield Corpn. Its global telecom tower portfolio exceeds 230,000 towers, making it the third largest in the world.
Telecom towers are critical infrastructure assets characterized by long-term contracts with investment-grade counterparties and stable, predictable cash flows
Brookfield’s digital infrastructure assets under management (AUM) have grown to over $40 billion, up from less than $5 billion five years ago, reflecting its significant and growing commitment to the sector. The company has indicated readiness to invest further as opportunities arise, particularly in supporting 5G rollout and network densification.
I wrote an analysis on Brookfield recently: https://rockandturner.substack.com/p/brookfield-part-3-of-3 [It is behind a 45 day paywall, so initially it is a timely actionable idea for paid subscribers - up almost 20% since first published - but will open for anyone to read from the beginning of June.]
EBITDA - (minus) Maintenance CAPEX still yields poor ROI and ROE. Compared to their WACC, they aren't earning economic profits.
That means that unless EBIT margins increase 2.5-3.0x, and CAPEX and interest payments decrease (I'm always weary when capital intensive businesses start decreasing CAPEX), how is this investable long-term?
Excellent point about returns vs. capital costs. Your concern is definitely valid.
There's a disconnect between Cellnex's operational metrics and financial returns. While their EBITDA margins (70.42%) appear healthy, the translation to ROE (2.77%) is remarkably poor. This gap is wider than for American tower companies, suggesting structural inefficiency beyond just the growth phase.
Those two metrics get to the heart of whether Cellnex can ever achieve adequate returns:
1. Net debt to EBITDA (8.34x)
This ratio is significantly higher than what's typical for mature infrastructure businesses (3-4x). Even with Cellnex's contracted revenue, this debt burden consumes so much cash through interest payments that equity returns are severely constrained. For perspective, American Tower's net debt to EBITDA is around 5x. Unless this ratio dramatically improves, interest payments will continue to absorb too much of the operational earnings, regardless of margin improvements.
2. Actual maintenance CAPEX
This tests management's core claim that Cellnex will transform into a cash-generating machine. If the €200-300M estimate proves accurate, FCF would increase dramatically as build-to-suit concludes. But infrastructure businesses often underestimate true maintenance needs or defer them temporarily. The aging of tower assets, technology upgrades, and environmental regulations could all push this figure higher, invalidating the harvesting phase thesis.
I would look at those two metrics going forward.
The real test will be 2026-2027, when we'll see if the operational leverage is real or if Cellnex has simply built an empire that can't deliver adequate returns regardless of scale.
That was a great piece on Cellnex. One aspect you didn't mention is that Cellnex can execute an aggressive leveraged buyback model. If they keep the leverage ratio at 5.5 times EBITDA, they can take on 5.5 times of new debt for every extra dollar of EBITDA. They already mentioned that they want to spend ~10bn € for expansion and/or accretive buybacks in the next couple of years.
Thanks for the thoughtful comment! You've touched on an interesting financial lever Cellnex could theoretically pull.
The math does check out - each additional euro of EBITDA at a 5.5x leverage target would indeed create €5.5 of debt capacity. But there are a few details worth fine-tuning in your thesis:
First, Cellnex has been actively reducing leverage (from 7.7x to 6.4x EBITDA) rather than maintaining a steady ratio. Patuano has emphasized their "strong commitment with material deleveraging" alongside shareholder returns.
As for the €10bn figure - management talks about cumulative cash generation exceeding €7bn from 2026-2030, which is substantial (about 30% of current market cap), but not quite €10bn. They haven't explicitly outlined plans to deploy this level of capital for expansion and buybacks specifically.
What they have confirmed is the €800M buyback starting this year (a "floor" for future years) and dividend plans starting in 2026. Patuano was quite clear: "Next year is going to be something more than what we do this year. And this is without making any further disposal."
The key insight from management's commentary is they're pursuing a progressive, sustainable approach rather than maximising leverage. As Patuano noted, future capital allocation will depend on "where the share price is" - suggesting they won't pursue buybacks at any price.
It's less an "aggressive leveraged buyback model" and more a balanced approach that increases shareholder returns incrementally while maintaining financial discipline and that precious investment grade rating they just secured.
That's a great comment, thank you very much. They mentioned on their CMD that they aim for a long-term target leverage of 5.0-6.0x Net Debt / EBITDA (5.5x midpoint). They should arrive at their target leverage in pretty short oder. The RLFCF in combination with additional debt capacity will enable Cellnex to deploy a pretty big chunk of money each year in order to increase FCF per share.
As you describe, mobile phone signals today primarily rely on terrestrial hardware attached to the towers owned by Cellnex.
It's been a cash cow of a business over the past decade or more.
But is that changing with the rise of low Earth orbit (LEO) satellites. These satellites orbit much closer to Earth than traditional satellites, enabling them to communicate directly with ordinary mobile devices using standard mobile spectrum bands. Reducing costs are making it feasible for satellites to act as "cell towers in space," delivering signals direct-to-device to regular 5G smartphones without specialist hardware.
In the short-term, LEO satellites are expected to complement, not immediately replace, terrestrial towers. But in the medium-term we will likely see a growth in satellite networks, especially as 5G and 6G technologies mature, offering seamless global coverage and ultra-low latency.
Elon Musk has his Starlink satellites and Jeff Bezos is launching 3,500 satellites as part of Amazon's project Kuiper. These are just two of many. Is the Cellnex model about to be disrupted?
Fair question about satellites - I've spent a bit of time looking into this.
It seems like the satellite disruption thesis falls apart when you examine the physics and economics. LEOs simply can't match tower capacity in populated areas. Think about it: a single urban tower handles thousands of simultaneous connections in a small radius, while Starlink's entire constellation might support maybe a million concurrent users globally.
Where does Cellnex make most of its money? Urban and suburban areas - exactly where satellites struggle most with capacity density. It's perfect for someone that wants fast internet in the mountains or in the middle of nowhere. It's expensive to get towers there for just a few connections.
Let's look at the economics as well. Satellites cost less upfront but need replacement every 5-10 years. Towers cost more initially but last decades with minimal maintenance. The lifetime math heavily favors towers.
At the end of the day, mobile operators continue signing 20+ year tower contracts despite being fully aware of satellite developments. They understand network architecture better than anyone, and their capital deployment speaks volumes.
What we'll likely see is integration rather than replacement. Satellites will excel precisely where tower economics fail - remote regions with sparse population. For Cellnex's core urban assets, physics and economics remain firmly on their side.
It's less disruption and more ecosystem evolution - each technology finding its optimal niche.
I agree with your conclusions - I was only playing Devil's advocate.
In fact, I am invested in Brookfield Corpn. Its global telecom tower portfolio exceeds 230,000 towers, making it the third largest in the world.
Telecom towers are critical infrastructure assets characterized by long-term contracts with investment-grade counterparties and stable, predictable cash flows
Brookfield’s digital infrastructure assets under management (AUM) have grown to over $40 billion, up from less than $5 billion five years ago, reflecting its significant and growing commitment to the sector. The company has indicated readiness to invest further as opportunities arise, particularly in supporting 5G rollout and network densification.
I wrote an analysis on Brookfield recently: https://rockandturner.substack.com/p/brookfield-part-3-of-3 [It is behind a 45 day paywall, so initially it is a timely actionable idea for paid subscribers - up almost 20% since first published - but will open for anyone to read from the beginning of June.]
Amazing writeup, thank you so much.
Just some back of the envelope thoughts...
EBITDA - (minus) Maintenance CAPEX still yields poor ROI and ROE. Compared to their WACC, they aren't earning economic profits.
That means that unless EBIT margins increase 2.5-3.0x, and CAPEX and interest payments decrease (I'm always weary when capital intensive businesses start decreasing CAPEX), how is this investable long-term?
Thank you again!
Excellent point about returns vs. capital costs. Your concern is definitely valid.
There's a disconnect between Cellnex's operational metrics and financial returns. While their EBITDA margins (70.42%) appear healthy, the translation to ROE (2.77%) is remarkably poor. This gap is wider than for American tower companies, suggesting structural inefficiency beyond just the growth phase.
Those two metrics get to the heart of whether Cellnex can ever achieve adequate returns:
1. Net debt to EBITDA (8.34x)
This ratio is significantly higher than what's typical for mature infrastructure businesses (3-4x). Even with Cellnex's contracted revenue, this debt burden consumes so much cash through interest payments that equity returns are severely constrained. For perspective, American Tower's net debt to EBITDA is around 5x. Unless this ratio dramatically improves, interest payments will continue to absorb too much of the operational earnings, regardless of margin improvements.
2. Actual maintenance CAPEX
This tests management's core claim that Cellnex will transform into a cash-generating machine. If the €200-300M estimate proves accurate, FCF would increase dramatically as build-to-suit concludes. But infrastructure businesses often underestimate true maintenance needs or defer them temporarily. The aging of tower assets, technology upgrades, and environmental regulations could all push this figure higher, invalidating the harvesting phase thesis.
I would look at those two metrics going forward.
The real test will be 2026-2027, when we'll see if the operational leverage is real or if Cellnex has simply built an empire that can't deliver adequate returns regardless of scale.
Good insights, thank you.
That was a great piece on Cellnex. One aspect you didn't mention is that Cellnex can execute an aggressive leveraged buyback model. If they keep the leverage ratio at 5.5 times EBITDA, they can take on 5.5 times of new debt for every extra dollar of EBITDA. They already mentioned that they want to spend ~10bn € for expansion and/or accretive buybacks in the next couple of years.
Thanks for the thoughtful comment! You've touched on an interesting financial lever Cellnex could theoretically pull.
The math does check out - each additional euro of EBITDA at a 5.5x leverage target would indeed create €5.5 of debt capacity. But there are a few details worth fine-tuning in your thesis:
First, Cellnex has been actively reducing leverage (from 7.7x to 6.4x EBITDA) rather than maintaining a steady ratio. Patuano has emphasized their "strong commitment with material deleveraging" alongside shareholder returns.
As for the €10bn figure - management talks about cumulative cash generation exceeding €7bn from 2026-2030, which is substantial (about 30% of current market cap), but not quite €10bn. They haven't explicitly outlined plans to deploy this level of capital for expansion and buybacks specifically.
What they have confirmed is the €800M buyback starting this year (a "floor" for future years) and dividend plans starting in 2026. Patuano was quite clear: "Next year is going to be something more than what we do this year. And this is without making any further disposal."
The key insight from management's commentary is they're pursuing a progressive, sustainable approach rather than maximising leverage. As Patuano noted, future capital allocation will depend on "where the share price is" - suggesting they won't pursue buybacks at any price.
It's less an "aggressive leveraged buyback model" and more a balanced approach that increases shareholder returns incrementally while maintaining financial discipline and that precious investment grade rating they just secured.
At least, that's what I see.
That's a great comment, thank you very much. They mentioned on their CMD that they aim for a long-term target leverage of 5.0-6.0x Net Debt / EBITDA (5.5x midpoint). They should arrive at their target leverage in pretty short oder. The RLFCF in combination with additional debt capacity will enable Cellnex to deploy a pretty big chunk of money each year in order to increase FCF per share.
(FD: I own some shares.)
Nice write-up on Cellnex 🗼
An excellent resource, if you want to dive deeper into Cellnex and the tower sector:
https://www.towerxchange.com/