Crown Castle: Setting the Next Capital-Allocation Policy
Download case PDF (4 pages)Request teacher’s pack
A strategy case for MBA students and working professionals
Information cutoff: October 2, 2026, 4:00 p.m. Eastern Time
Company: Crown Castle Inc. (NYSE: CCI)
A tower-only portfolio, several competing uses
In May 2026, Crown Castle completed the sale of its fiber solutions and small-cell businesses. The transaction marked a strategic reset: towers became the company’s sole operating segment. By June 30, the company also reported completing the previously announced $1 billion share repurchase and repaying more than $7 billion of debt. Those are completed historical actions, not a commitment to repeat the same mix in the next cycle. May 1 closing Form 8-K, Items 2.01 and 7.01; Q2 2026 earnings release, “Highlights from the Quarter”
For this classroom exercise, assume that a hypothetical CCI strategy group is reviewing capital-allocation criteria in late 2026. The public record does not establish an actual board meeting, a 2027 capital budget, project hurdle rate, or a decision to allocate a fixed amount. Your task is to recommend a policy for sequencing three uses: securing land and other site control; targeted tower upgrades and operating investments that could support tenancy or service; and preserving balance-sheet flexibility or making discretionary distributions. Do not calculate project returns from data the company has not disclosed.
The operating signal is mixed
CCI reported second-quarter 2026 site-rental revenue of $967 million, down 4.1% year over year, and adjusted EBITDA down 4%. AFFO rose 10% to $488 million. The company attributed the AFFO increase in part to lower interest expense and higher interest income following the sale. It is therefore possible for AFFO to rise while site-rental revenue and adjusted EBITDA fall; the quarter does not by itself establish operating growth. [Q2 release, “Results from the Quarter” and “Highlights from the Quarter”]
CCI also reported $38 million of adjusted organic site-rental billings growth, or 3.9%, while organic billings including DISH/Sprint effects were negative $16 million. The adjusted measure removes specified DISH/Sprint churn effects; it is a company-defined operating measure, not total site-rental revenue growth. The distinction matters when evaluating whether tower investment is creating new demand or offsetting churn. [Q2 release, “Highlights from the Quarter” and reconciliation tables]
At the July 22 release, management’s FY26 midpoint outlook was $3.855 billion in site-rental revenue, down 5% from 2025; $2.690 billion in adjusted EBITDA, down 6%; and $1.975 billion in AFFO, up 4%. These were forecasts as of July 22, not full-year results or a FY27 forecast. H1 continuing-operations capital expenditure was $116 million: $68 million for land interests, $34 million for tower improvements and other projects, and $14 million for sustaining capital. FY26 discretionary capex guidance was $150–250 million, less $30–50 million of prepaid-rent additions. These figures provide context, but do not establish a next-year envelope or comparative project returns. [Q2 release, “Outlook,” “Components of Capital Expenditures,” and “Outlook for Discretionary Capital Expenditures Less Prepaid Rent Additions”]
Site control and tenancy
At year-end 2025, CCI reported more than 40,000 towers, around 105,000 small-cell nodes and roughly 90,000 route-miles. The latter two businesses were part of the divested operations. CCI says its strategy includes growing tower cash flow through land interests, construction, acquisitions and improvements, alongside capital returns. Those stated priorities do not disclose which use has the better incremental economics. [2025 Form 10-K, “Business—Overview” and “Strategy”]
Land control may protect an important operating input. CCI reported that about 90% of tower adjusted site-rental gross margin came from towers on land it had owned or controlled for more than ten years, about 80% from sites held for more than twenty years, and around 40% from owned land or perpetual easements. These are portfolio measures, not evidence that a particular land purchase is attractive. Public materials do not identify a candidate site's price, current ground rent, lease expiration, tenant plans or incremental cash flows. [2025 Form 10-K, MD&A, “Outlook Highlights”]
Investment in tower capacity or operations may enable additional tenants, support amendments, or improve service. But project-level demand, cost, timing, and expected rent are not public. The H1 spending mix shows that CCI used capital for land and tower projects; it does not show their respective performance.
Concentration and changing carrier networks
T-Mobile, AT&T and Verizon accounted for about 90% of CCI’s 2025 site-rental revenue. Tower contracts generally start with terms of five to fifteen years, with tenant renewal options of five to ten years; weighted-average remaining term was about six years. CCI disclosed approximately $23.7 billion of expected future cash inflows under tower contracts, excluding DISH. These portfolio disclosures do not reveal which individual sites face renewal, expansion or nonrenewal. [2025 Form 10-K, “Customers” and “Competition”; “Business—Overview”; MD&A, “Outlook Highlights”]
Carrier consolidation can create both risk and opportunity. CCI attributed roughly $200 million of 2025 tower nonrenewals to the T-Mobile/Sprint combination and expected additional nonrenewals through 2034. Consolidated carriers may remove duplicate equipment from overlapping sites, while a remaining carrier may colocate or expand equipment elsewhere. No public site-level map in the reviewed materials resolves the net effect for a proposed CCI investment.
The DISH matter requires similar care. CCI reported a default and termination dispute, asserted more than $3.5 billion of remaining payments, and described related balance-sheet amounts and its recovery expectation. The claim was not an adjudicated liability or realized cash. FY26 guidance excluded DISH revenue and reflected an expected rent reduction tied to termination. [2025 Form 10-K, Item 1A and Note 18; Q2 release, FY26 outlook bridge]
Passive tower colocation—multiple carriers leasing space on a tower landlord’s infrastructure—is not the same decision as active coordination by rival carriers over network investment, quality, or pricing. Academic models of network sharing address carrier coordination and do not estimate CCI’s landlord returns. They caution that sharing may strengthen investment incentives under some joint decisions and weaken them under separate decisions through free-riding. The model cannot tell CCI which tower to upgrade.
What the research can—and cannot—say
Historical REIT evidence offers competing mechanisms rather than a capital-allocation answer. Breuer, Nguyen and Steininger analyze 192 U.S. equity REITs and 1,653 firm-year observations from 1999–2015. Their fixed-effects results associate tangible assets with higher debt-to-assets and operating risk with lower leverage; reported growth-opportunity measures are not significant in the REIT regression. The study pools property types and is observational; it does not estimate CCI’s financing capacity or optimal leverage. [Journal of Financial Research 46 (2023), §4.1 and Table 3; §7.3]
Ghosh and Sun begin with 794 complete annual firm-year observations for 1999–2009. Requiring three consecutive years and averaging variables leaves 519 nonmissing observations representing 107 REITs; these are three-year-averaged observations, not 519 untransformed annual records. Their Table 6 reports positive associations between excess dividends and measures of externally financed growth. Dividend and investment policies are jointly determined, so the result is not causal proof and does not compare CCI’s debt repayment, buyback, land purchases or tower upgrades. The study’s excess-dividend measure is not the statutory REIT distribution minimum. A tax-qualification minimum distribution, discretionary dividends, and share repurchases are distinct choices. [Journal of Real Estate Finance and Economics 48 (2014), Data section, pp. 674–675; Tables 1 and 6, pp. 695–696 and 702–703]
The policy decision
The hypothetical review must choose how the company should rank and sequence opportunities when the public evidence does not permit project-level comparison. One policy could emphasize site control to protect long-lived tenancy. Another could release capital for tower upgrades when specific operating or tenant evidence supports them. A third could preserve greater financing flexibility and make distributions only after defined operating and investment needs are addressed. A blended policy is possible, but it still needs explicit gates and reasons for moving capital between categories.
Each approach carries a cost. Land purchases may reduce future ground-lease exposure but tie up capital at an unknown price. Upgrades may improve capacity or service but could precede carrier demand or fail to prevent churn. Retaining capital can support flexibility but forgo a distribution or a timely project. Distributions can return excess capital to owners, while leaving less discretionary funding inside the company. The public disclosures do not rank these outcomes for a particular site.
Assignment
Prepare a concise decision memo for the hypothetical late-2026 strategy group. Recommend a capital-allocation policy across site control, targeted tower investment and financial flexibility/discretionary distributions. Include:
- criteria and sequencing gates for land purchases and tower upgrades;
- what evidence would justify releasing, staging, pausing or redirecting capital;
- how carrier concentration, renewal exposure, consolidation and service needs enter the decision;
- the financing and distribution principles that apply after the completed repurchase and debt repayment; and
- the missing information most likely to change your policy.
You may recommend a blended policy. State assumptions plainly. Do not invent a fixed budget, project return, hurdle rate, site-level demand, tenant commitment or future result. Do not present the hypothetical review as an actual board vote.
Source note
This case uses information public by October 2, 2026, 4:00 p.m. ET. The May sale, repurchase and debt repayment are completed actions; July FY26 guidance remains a forecast. See the research source register, evidence extraction, and literature review for source locators, reading depth, definitions and limitations. No later outcome is used.