Crown Castle Is Being Priced as a Broken Fiber Company. It Sold the Fiber.
For most of the last decade Crown Castle was the odd one out among the three big American tower REITs — the one that invested more than $15 billion building a fiber and small-cell business its owners never wanted, cut its dividend, and cycled through four chief executives since December 2023. On May 1, 2026 that experiment formally ended: the company closed the sale of its fiber and small-cell businesses for $8.4 billion of net cash proceeds, with EQT taking small cells and Zayo taking fiber solutions .
What remains is the thing investors always liked — roughly 40,000 U.S. cell towers leased to the three national carriers on long-dated, escalating contracts — trading near $74.90, close to a five-year low and within a whisker of its $69.72 52-week trough . The stock has round-tripped to price levels it last saw in the mid-2010s, and the market is discounting the entire enterprise as though the tower franchise itself were impaired. It isn't. The only thing that was impaired was fiber, and fiber is gone.
My argument is that three things are true at once, and the market is pricing only the first:
- 2026 is the disclosed low-water mark for organic growth, with the churn that caused it finite and rolling off;
- The balance sheet and cost structure are now demonstrably repaired; and
- There is a multibillion-dollar contractual claim against a bankrupt tenant that carries zero in guidance.
None of that requires heroic assumptions about a spectrum super-cycle. It requires only that a de-risked tower business be valued like one.
Two segments became one, and the survivor is the good half
Through 2024 Crown Castle reported in two pieces: Towers, which produced $4.46 billion of 2024 revenue, and Fiber, which produced $2.11 billion . The Fiber half is the one that generated the write-downs — the FY24 impairment that pushed reported operating income to negative $2.94 billion and diluted EPS to negative $8.98 — and it is the half that has now been sold and reclassified into discontinued operations through April 30, 2026 . What remains is a single reporting business: U.S. towers. That collapses the whole "sum of a good business and a bad one" debate that hung over the stock for years. There is no fiber capex line to fund, no small-cell lease-up to wait for, no cross-subsidy to argue about. The company describes itself now as the only publicly traded pure-play U.S. tower operator , and while that is a marketing line, the accounting behind it is real.
The economics of what's left are what a tower investor wants to own. Second-quarter site rental revenue was $967 million against $723 million of adjusted site rental gross margin — roughly a 75% gross margin on the rental book . That margin exists because a tower's incremental tenant costs almost nothing to add: the ground lease and the steel are fixed, so each additional carrier amendment drops to the bottom line. And the capital intensity is now trivial. Second-quarter capital expenditures were $59 million, of which only $7 million was sustaining ; full-year discretionary capex is guided to $150–250 million . Against roughly $2.69 billion of guided 2026 adjusted EBITDA , this is a business reinvesting a low-single-digit percentage of its cash flow — the opposite of the capital-hungry profile it carried with fiber attached. That distinction matters for how you value it: the return on incremental capital here is not the question, because there is barely any incremental capital. The value driver is escalators plus modest new leasing landing on a near-fixed cost base, which is exactly what a per-share AFFO lens captures and what an EV/EBITDA screen on the consolidated trailing numbers badly distorts.
The trough is contracted, not hoped for
The bear framing treats Crown Castle as a melting ice cube: site rental revenue fell about 5% in 2025 to roughly $4.05 billion, and is guided down another 5% in 2026 to about $3.86 billion . But that decline is not a demand problem — it is the mechanical roll-off of two specific, disclosed sources of churn. The first is the Sprint cancellations tied to the old T-Mobile/Sprint consolidation; the second, larger one is the DISH terminations, after Crown Castle sent DISH a notice of default in January 2026 . Together these represent roughly $240 million of full-year 2026 headwind, of which about $220 million is DISH . Strip them out, and the underlying business is growing: management guides full-year 2026 organic contribution to site rental billings, excluding Sprint and DISH, to approximately $135 million, or 3.4% , and in the second quarter that underlying organic growth ran 3.9% .
The reason to believe 2026 is the floor rather than a way station down is that the floor is largely already signed. By the end of the second quarter, more than 90% of full-year 2026 organic growth was contracted, up from roughly 80% at the start of the year , and management has explicitly and repeatedly called 2026 the low point for organic growth . This is the hinge of the whole thesis. "Contracted" is a far stronger word than "expected" — it means the escalators and amendments that produce next year's billings are under executed agreements, not in a sales pipeline. The escalator component alone is guided at roughly $100 million for 2026 , and escalators are contractual price increases that do not depend on carrier capex budgets. Layer in the medium-term demand drivers management points to — AT&T's 600 MHz spectrum, the 800 MHz auctions the FCC has slated to begin in 2027, and Ericsson's projection that U.S. mobile data per smartphone roughly doubles over five years — and you do not need any of them to hit for the trough thesis to work. You need only the DISH and Sprint churn to finish rolling through, which by definition it will.
The 10% AFFO jump is half financing — and that's fine
Here the argument has to be honest about a real weakness, because it is the one the sell-side seized on. Second-quarter AFFO grew 10% year over year to $488 million, or $1.13 per share . But that headline is flattered by the balance-sheet work, not by towers: AFFO in the quarter benefited from a $35 million year-over-year drop in interest expense and a $14 million rise in interest income from parking sale proceeds, and management said plainly the high interest income will not recur in the second half . Adjusted EBITDA actually fell 4% to $675 million . So the durable growth rate underneath is the ~3.4–3.9% organic figure, not the 10% print — and I model it that way. Anyone extrapolating 10% AFFO growth is going to be disappointed; the interest-expense lever is largely spent now that the deleveraging is done.
What that leaves is a clean, covered mid-single-digit compounder. Full-year 2026 AFFO is guided to $1.95–2.00 billion, or $4.53–4.65 per share, up from $4.36 in 2025 . The dividend is $1.0625 per quarter, or $4.25 annualized, most recently declared August 5 and payable September 30 — comfortably inside the AFFO guide, a payout in the low-90s percent that is covered but leaves little retained cash. That coverage is the reassuring part; the flat dividend is the tell I take seriously below.
The balance sheet is repaired, and the proof is in the leverage target
The single cleanest piece of evidence that this is no longer a distressed story is what management did with the $8.4 billion. It used the proceeds to repurchase $1 billion of stock — retiring more than 11 million shares at an average of $88.66 and cutting the annual dividend obligation by $47 million — and to repay roughly $7.2 billion of debt, including all outstanding revolver and commercial-paper borrowings . Crown Castle ended the quarter with $18.24 billion of debt and other long-term obligations against $1.25 billion of cash, 100% fixed-rate debt, a roughly seven-year weighted-average maturity, and about $4.5 billion of undrawn revolver . Net leverage now sits inside the company's stated 6.0x–6.5x investment-grade target range, and management describes the capital-allocation order plainly: fund the dividend, fund the high-return capex, delever to that 6.0x–6.5x band, and only then buy back stock .
There is a scar visible on the balance sheet — total equity is a $3.27 billion deficit, and distributions in excess of earnings have accumulated to $20.8 billion . But that is a mechanical artifact of years of REIT distributions exceeding GAAP earnings plus the fiber impairments, not a going-concern signal. On the metric that governs a REIT's borrowing capacity — net debt to EBITDA — the company is now where it said it would be.
A $3.5 billion claim the market is valuing at nothing
The most asymmetric item in the story is the one guidance deliberately excludes. When Crown Castle terminated the DISH master lease in January 2026, it asserted that DISH owes it in excess of $3.5 billion in remaining payments , and none of that sits in the 2026 outlook . DISH's parent EchoStar and DISH itself have since given that claim two potential recovery avenues. On May 12, 2026, as a condition of approving a spectrum-license transaction, the FCC ordered EchoStar to establish and fund a $2.4 billion trust to satisfy certain qualified claims — including wireless-network infrastructure and site-lease obligations — and Crown Castle believes its DISH claims qualify . Then on June 30, 2026, DISH and affiliates filed for Chapter 11, and Crown Castle is now pursuing recovery through both the bankruptcy and the FCC trust . Critically, management's read is that the trust sits outside the normal bankruptcy waterfall — a materially better position than an unsecured claim against a bankrupt estate.
The honest way to carry this is exactly how the company carries it: on the balance sheet, the net position tied to the terminated DISH agreements is only about $165 million, which management expects to recover but flags as subject to uncertainty . So the base case should assume nothing. The point is that the market, having watched DISH deteriorate for two years, is almost certainly assuming nothing too — which turns any recovery into a free option struck at zero. A partial allowance against a $2.4 billion trust that ranks ahead of general creditors is not a fantasy; it is a plausible, if unquantifiable, source of cash the current price ignores entirely.
The one thing the bull case cannot fully explain
The counter-argument worth taking seriously is not the coverage arithmetic and it is not the fiber history — both are known and priced. It is the shape of the customer base and one specific management decision. Crown Castle's revenue comes overwhelmingly from three carriers, and DISH just demonstrated exactly what a fourth-carrier ambition collapsing does to a tower landlord: a $220 million hole in a single year . Carrier consolidation is a structural feature of this industry, and a genuinely bearish reading is that three tenants is a machine for occasionally manufacturing large non-renewals, so 2026's "trough" could simply be a landing before the next leg down whenever the next tenant rationalizes its footprint. There is no disclosed 2027 leasing backlog to refute that; the >90%-contracted figure protects 2026, not the year after.
The sharper tell sits in the dividend. Crown Castle held the payout flat at $1.0625 per quarter even after repaying more than $7 billion of debt and hitting its leverage target. A 5.5%-yielding REIT with $4.5 billion of revolver headroom and a covered payout that was genuinely comfortable about 2027 would ordinarily have signaled a return to dividend growth. Holding it flat is management pricing its own uncertainty about next year's leasing — and it is the one fact the optimistic reading cannot fully absorb. I discount it, but not to zero: I read it as caution from a leadership team in its first standalone year, running a business whose forward organic curve depends on carrier capex it does not control, rather than as a coverage confession, because the AFFO plainly covers the dividend and the leverage target has been met. But a fourth consecutive flat declaration would harden the bearish read, and it is the main reason this is a moderate-conviction long rather than a table-pounder.
What it's worth, and why the discount is the opportunity
Value a tower REIT the way its owners do — on price to AFFO per share — and the picture is not that Crown Castle is cheap against its own peer, but that it is priced identically to a far more diversified, less-levered franchise despite carrying a disclosed, finite, rolling-off drag that its peer does not. At $74.90 on 2026 guided AFFO of $4.59 per share , Crown Castle trades at about 16.3x. American Tower, at $173.72 against its raised 2026 guide of $11.00–$11.17 per share , trades at roughly 15.7x — and American Tower carries net leverage near 4.9x versus Crown Castle in the 6.0x–6.5x band . On a normalized forward basis, blending 2026 and 2027 estimated AFFO of roughly $4.72 per share, Crown Castle sits near 16x — well below the low-to-mid-20s multiples U.S. towers commanded before the rate cycle and the fiber debacle re-rated the group.
The re-rating case is not that Crown Castle deserves a premium; a low-90s payout that leaves little cash to reinvest and a flat dividend argue against that. It is that a business printing its disclosed organic trough, with next year's floor already contracted and the churn that caused the decline finishing its roll-off, should not trade at a discount to its own recent range while a diversified peer with lower leverage trades at a similar multiple. A modest re-rate toward 17x on the ~$4.72 normalized AFFO figure supports a price around $80, roughly 5–6% above the current level, on top of a 5.5% dividend yield — a low-double-digit total return before any credit for the DISH claim. The bull outcome, a re-rate toward 19–20x as a clean third-quarter organic print converts the contracted backlog into visible re-acceleration, gets you into the mid-$90s; the bear, a fresh carrier non-renewal that reopens the terminal-growth question and compresses the multiple into the low teens, takes you toward $60.
The catalyst that broke this stock — the fiber write-downs, the dividend cut, the DISH default, the leadership churn — is behind it and in the price. The catalyst the market is under-weighting is the second derivative on organic billings: a steepening curve off a de-risked trough, with a bankrupt-tenant claim thrown in for free. The debate resolves at the next print, expected in late October, when the third-quarter organic figure will either confirm the contracted backlog is converting or reveal fresh non-renewals. On the evidence available today — more than 90% of 2026 contracted, a repaired balance sheet, a covered dividend, and a churn drag that is finite by disclosure — the weight sits with the trough thesis. The market is valuing a repaired U.S. tower franchise as if it were still the broken fiber company it used to be, and that gap is the case for owning it.
Disclaimer: Tenzing Analytics, Inc. is not a registered investment adviser, broker-dealer, or financial institution, and all content provided by Tenzing Memo is AI-generated for informational purposes only. It does not constitute investment advice, a recommendation, or an offer or solicitation to buy, sell, or hold any security or financial instrument. Content is general in nature, not tailored to any individual's financial situation or objectives, and may contain errors, omissions, or inaccuracies; it should not be relied upon as a primary basis for any investment decision. Any forward-looking statements are subject to risks and uncertainties and may not reflect actual future results. Users should conduct their own research and consult qualified financial, legal, or tax advisors before making investment decisions, and by accessing this content, you acknowledge and accept these limitations.