Business Model
| Revenue stream | Mechanics | Economics |
|---|---|---|
| Site rental -- 95.9% | A carrier leases space at a specific height on a specific tower for its radios and antennas. Master lease agreements (MLAs) run 5-10 years with renewal options and typically ~3% fixed annual escalators. Non-cancellable in practice | The tower is a fixed-cost asset. The first tenant covers the ground lease, maintenance and overhead; each incremental tenant carries roughly 100% incremental margin. This is the entire economic engine |
| Amendments | An existing tenant adding or swapping equipment -- new spectrum band, more antennas, heavier loading -- pays more under the same MLA | The highest-quality growth: no new tower, no new ground lease, no acquisition cost. Amendment revenue is why a new spectrum band matters more than a software upgrade -- AT&T's 600 MHz needs physical radios; 3.45 GHz did not |
| Services & other -- 4.1% | Site installation, tower modifications, construction management for carriers | Low-margin, cyclical, fragmented, and shrinking -- guided down $20M intra-year on carrier decision paralysis. Not a moat business |
| Edge compute -- ~0% | Trials to host distributed inference/compute at tower sites needing under 0.2 MW | Zero incremental capex by management's account. Optionality only -- no revenue, nothing quantified, and the street carries zero |
A single tower is best understood as a small, highly operationally geared property. Revenue per tower is roughly $96.7K on an LQA site-rental basis as of 2026Q2 — down from ~$107K in 2024Q4, because tenancy churn is running ahead of new leasing on a fixed 40,000-tower base.
The critical structural facts:
- Tower count has been flat at 40,000 for three years. There is no unit growth in this business. All growth must come from putting more tenants and more equipment on existing structures, or from raising price.
- Ground cost is the main controllable expense. CCI leases the land under a large share of its towers rather than owning it. Management has quantified a ~11-point land-ownership gap versus AMT and SBAC and is buying land to close it — this is the source of the targeted ~200bps of EBITDA margin expansion, and it is why capex is deliberately rising even as the business shrinks.
- Escalators provide a contractual floor. ~3% fixed annual increases on the installed base mean revenue grows without any new leasing activity — which is why "organic growth ex-churn" of 3.4-3.9% can coexist with a reported revenue decline of -4.9%.
A macro tower is a zoning-and-siting monopoly at a specific latitude and longitude. Four things make it un-replicable at economic cost inside a carrier's search ring:
- Municipal permitting and local zoning — often the binding constraint, and increasingly restrictive.
- Environmental review and FAA clearance — mandatory, slow, and occasionally disqualifying.
- Ground-lease control — whoever holds the land rights holds the site.
- Time — relocating a radio to another structure requires a new search ring, approvals, a ground lease and construction: 18-36 months and capex 30-40% above simply leasing.
This is why AT&T's master lease lets it add radios at 30-40% below new-build economics, and why the honest answer to "could a customer replace CCI within 12 months?" is no. The only 12-month replacement path is decommissioning genuinely redundant sites after spectrum consolidation — which is precisely what Sprint and DISH churn is, and it is contractually bounded rather than open-ended.
The practical consequence shows up at renewal. CCI sets holding-capacity and amendment pricing and ~3% escalators unilaterally within existing MLAs — genuine pricing power on the installed base. But at master-lease renewal against a three-buyer market it behaves closer to a price-taker. The CFO's own framing when asked about the 2028 AT&T renewal — "what kind of money we might or might not be leaving on the table, competition, et cetera... craft win-win outcomes" — is negotiation language, not monopolist language.
That 2028 renewal covers roughly $774M of annualised rent struck under 2013 sale-leaseback terms with modest escalators. It is more plausibly a downward mark-to-market than an uplift, and management has declined to quantify it in either direction.
The company that exists today is materially different from the one that existed eighteen months ago:
| Change | Effect on the business model |
|---|---|
| Fiber / small cells sold (closed 2026-05-01) | $8.4B net proceeds. Removed the capital-hungry, low-return segment -- but also removed the only diversification. CCI is now a one-segment, one-country, three-customer business |
| $7.2B of debt repaid | Net debt from $24,577M to $17,099M; leverage to 6.33x, inside the 6.0-6.5x IG target. Interest expense fell $35M YoY. A permanent reduction in the cost structure -- but a one-time event |
| $1B of stock repurchased at $88.66 | 11M+ shares retired, cutting the annual dividend obligation by $47M. Share count 437M to 434M |
| ~20% workforce reduction to ~1,250 FTEs | $65M of annualised run-rate cost removed, plus a further $15M in-year at Q2'26 |
| DISH default (Jan 2026) and bankruptcy | Removed ~$220M of FY2026 revenue that had been carried as contracted. CCI is an unsecured creditor pursuing a $3.5B claim against a $2.4B escrow contested by up to $13B of total claims |
| Dividend rebased | DPS $6.26 (FY2024) to $4.75 (FY2025) to a $4.25 annualised run-rate -- held flat since, at a 92.6% AFFO payout |
This is the single most underappreciated feature of the model today. At a 92.6% AFFO payout, CCI retains roughly $164M of cash flow against $200M of discretionary capex. That means:
- No organic deleveraging. Leverage sits at 6.33x, at the top of the target band, and can only fall on EBITDA growth — which is currently negative.
- No organic buyback capacity. The $1B Q2 repurchase was funded by divestiture proceeds, not free cash flow. That well is dry.
- Land buy-ups compete with everything else. The ~200bps of margin self-help requires capital that the payout ratio does not comfortably leave.
Management has committed to holding the dividend flat until the payout falls to 75-80% of AFFO — which, absent a dividend cut, requires AFFO to grow roughly 15-20% from here. That is the arithmetic underpinning the "2026 is the trough" claim, and it is why the trough call matters far more than a single quarter's organic growth print.
The model's fragility is concentrated in three places, all covered in more depth on Valuation, Concerns & Risks:
- Customer credit and concentration. 89% from three carriers, and the DISH default demonstrated that "contracted" revenue is only as good as the counterparty.
- The renewal cycle. A five-year weighted-average remaining contract term (down from six) and a $774M AT&T book renewing in 2028 against a three-buyer market.
- Technological substitution. Satellite direct-to-cell is the live narrative. Management's rebuttal is quantitatively serious — ~90% of usage is indoor or in-vehicle, satellite signal is ~10,000x weaker, terrestrial supports ~30x more users per MHz — and asked whether carriers had changed behaviour because of it, the answer was "No. Nothing." The fundamental risk looks overstated; the multiple drag is real.