First towers-only quarter after the $8.4B fiber/small-cells sale: site rental revenue fell 4.1% on DISH and Sprint churn, but AFFO per share rose 10.8% to $1.13 as $7B+ of debt paydown cut interest; GAAP net income fell 68% on a $280M disposal loss.
Revenue
$1.0B
-4.9% YoY
Net income
$94M
-67.7% YoY
Diluted EPS
$0.22
-67.2% YoY
Operating margin
46.6%
Overview
Q2 2026 was Crown Castle's first quarter as a towers-only company. On May 1, 2026 it closed the sale of its fiber business (to Zayo) and its small-cells business (to EQT Infrastructure) for $8.4 billion of net cash. It used that cash to pay down about $7.2 billion of debt and buy back $1.0 billion of its own shares. The quarter's numbers point in different directions depending on which line you read:
Tower rental revenue fell 4.1% to $967 million. Almost all of the drop came from two customers leaving: DISH, which stopped paying and whose leases Crown Castle terminated in January, and the long-running cancellation of old Sprint equipment now that Sprint is part of T-Mobile.
Profit from the tower business (continuing operations) rose 12.8% to $299 million. The increase came from lower interest costs and interest earned on the sale cash, not from the towers themselves.
Total GAAP net income fell 67.7% to $94 million, because of a $280 million accounting loss on the business that was sold.
AFFO, the cash-earnings measure tower REITs are valued on, rose 10% to $488 million ($1.13 per share), and management raised its full-year AFFO guidance slightly.
All operating figures below are for continuing operations (towers). The fiber and small-cells business is shown separately as "discontinued operations" through April 30, 2026, the day before the sale closed.
Why AFFO matters more than net income for a tower REIT
A REIT (real estate investment trust) must pay out most of its taxable income as dividends. In return it pays almost no corporate income tax; Crown Castle's tax provision was $4 million this quarter. Investors therefore care most about how much cash is left over each year to pay that dividend. GAAP net income answers that question poorly for a tower owner, for three reasons:
Depreciation is large and non-cash. Accounting rules spread a tower's cost over its useful life and subtract a slice of it every quarter ($171 million of depreciation, amortization and accretion this quarter). Well-kept towers do not actually lose value that fast, and no cash leaves the company for it.
Straight-line accounting. Leases with built-in annual rent increases are booked as if the average rent were paid evenly over the lease. That makes reported revenue differ from the cash billed.
One-offs. This quarter includes a $280 million loss on the fiber sale and a $24 million gain from buying back bonds below face value. Neither says anything about how much cash the towers produce.
Read 0 community reports on Crown Castle, or write your own.Write a report
AFFO (adjusted funds from operations) starts from net income, adds back real-estate depreciation, strips out straight-line effects, stock-based pay and one-off gains and losses, and subtracts the modest capital spending needed to maintain the towers ($7 million this quarter). The result is roughly the recurring cash available to shareholders. AFFO is a company-defined measure that is not audited; the figures below come from Crown Castle's Q2 2026 earnings release (8-K Exhibit 99.1, filed July 22, 2026).
Key metrics (Q2 2026, three months ended June 30)
Metric
Q2 2026
Q2 2025
YoY Change
Net revenues (continuing)
$1,008M
$1,060M
-4.9%
Site rental revenues (continuing)
$967M
$1,008M
-4.1%
Operating income (continuing)
$470M
$506M
-7.1%
Operating margin (continuing)
46.6%
47.7%
-1.1 pts
Income from continuing operations
$299M
$265M
+12.8%
Net income (total, incl. discontinued)
$94M
$291M
-67.7%
Diluted EPS, continuing operations
$0.69
$0.61
+13.1%
Diluted EPS (total)
$0.22
$0.67
-67.2%
Adjusted EBITDA*
$675M
$705M
-4.3%
AFFO*
$488M
$444M
+9.9%
AFFO per share*
$1.13
$1.02
+10.8%
Organic site rental billings growth, ex-DISH & Sprint*
4.2%
3.7%
+0.5 pts
Dividend declared per share
$1.0625
$1.0625
flat
*Non-GAAP measures from the company's Q2 2026 earnings release. Operating margin is operating income as a share of net revenues, i.e. the share of revenue left after running the business, before interest and tax.
Year to date (six months ended June 30, continuing unless noted): net revenues of $2,018M (-4.9% from $2,121M); site rental revenues of $1,928M (-4.5% from $2,019M); income from continuing operations of $520M (-5.3% from $549M); total net income of $245M versus a $(173)M loss in H1 2025; AFFO of $934M (+1.2% from $923M), or $2.14 per share versus $2.11.
Site rental revenue: the base grew, but DISH and Sprint took it away
The $41 million revenue decline is the net result of several moving parts. The figures come from the earnings release; the 10-Q confirms the DISH and Sprint amounts.
Organic growth ("organic contribution to site rental billings") was +$38 million, or 4.2% measured against prior-year billings with DISH excluded. That is up from 3.7% a year earlier on the same basis. It comes from:
new tenants and upgrades on existing towers (+$15M)
contractual rent escalators, the fixed annual increases built into leases (+$25M)
other billings (+$5M)
offset by ordinary churn, meaning customers not renewing (-$7M)
DISH terminations took away $49 million. Crown Castle stopped recognizing any DISH revenue from January 1, 2026.
Sprint cancellations took away $5 million. This is T-Mobile removing duplicate Sprint equipment under the previously agreed network consolidation. Year to date it totals $10 million, against $98 million for DISH.
Accounting items took away $26 million. Straight-lined revenue fell by $23 million and amortization of prepaid rent (upfront payments recognized gradually) fell by $3 million. Neither affects cash billed in the quarter.
The underlying tower business is therefore still growing at about 4% a year. The headline decline comes from one customer that stopped paying plus accounting effects that shrink reported revenue more than cash.
Customer concentration remains high: about 93% of H1 site rental revenue came from T-Mobile, AT&T and Verizon (10-Q). Services revenue (site development work for carriers, not under long-term contract) fell to $41 million from $52 million.
DISH: a $3.5 billion claim, and a bankruptcy
Crown Castle sent DISH a notice of default and termination on January 12, 2026. It claims DISH owes all remaining payments under the master lease, which total more than $3.5 billion. Two developments since then (10-Q, note 9):
On June 30, 2026, DISH and certain affiliates filed for Chapter 11 bankruptcy.
As a condition of approving a spectrum transaction by DISH's parent EchoStar, the FCC required EchoStar to fund a $2.4 billion trust for qualifying claims, including tower lease obligations. The trust was established on July 28, 2026. Crown Castle believes its claims qualify, subject to meeting the requirements and to how much money is available in the trust.
Crown Castle carries a net balance-sheet position of about $165 million related to DISH and says it expects that amount to be recoverable. The $3.5 billion claim is far larger than that and than the $2.4 billion trust, which must also cover other claimants, so any recovery beyond the carried amount is uncertain. The 2026 outlook assumes nothing from DISH.
The fiber and small-cells sale: how it shows up in the numbers
Terms: gross price of $8.5 billion, less $124 million of preliminary price adjustments (still subject to a post-closing settlement), for $8.4 billion net. Zayo bought fiber solutions and EQT bought small cells.
Presentation: the business is reported as discontinued operations for every period shown, through April 30, 2026. That is why Q2 2026 includes only one month of fiber operating income ($75 million, compared with $278 million in Q2 2025).
Loss on disposal: $280 million in Q2 ($625 million year to date; $1,082 million in H1 2025). This is the excess of the business's book value over the sale price, less costs to sell. The Q2 increase reflects continued investment in the fiber business until closing and the price adjustments. Because Crown Castle is a REIT, the loss produces no tax benefit.
Net result: the discontinued segment contributed a $205 million loss in Q2, which is why total net income ($94M) is far below continuing income ($299M).
Where the $8.4 billion went: debt and buybacks
Debt. Under a board authorization of up to $7.2 billion, Crown Castle took the following steps (10-Q, note 5):
repaid all of its old 2016 bank credit facility ($3,276M) and all of its commercial paper (short-term corporate IOUs)
repaid $750M of 3.700% notes at maturity in June
bought back $530M face value of bonds in the open market for about $500M, booking the $24M net gain noted above
repaid $1.0 billion of 1.050% notes at maturity in July, after quarter end
It also repaid $900M of 4.450% notes in February, before the sale closed.
Debt on the balance sheet fell to $18.2 billion at June 30 from $24.3 billion at December 31. All of it is now fixed-rate, with a weighted-average interest rate of 3.7% and roughly seven years to maturity. Q2 interest expense fell 14% to $208 million, and interest income rose to $18 million from $4 million as sale cash sat in interest-bearing accounts. A new $4.5 billion revolving credit line (maturing 2031) was essentially undrawn.
The company's supplement puts net debt at $17.1 billion, or 6.3x last-quarter annualized Adjusted EBITDA. That is still high leverage, but the new credit facility's covenant allows up to 7.0x. Roughly $2.2 billion of debt matures in the 12 months after June 30, including two March 2027 note issues.
Buybacks. Crown Castle completed its full $1.0 billion repurchase program within Q2, buying about 11 million shares at an average of $88.65. Shares outstanding fell to 426 million from 435 million at year-end. The supplement uses a June 30 closing price of $75.73, so the shares were bought well above where they ended the quarter.
Dividend
Crown Castle declared and paid $1.0625 per share in Q2, unchanged from Q2 2025. The dividend was already reset in 2025: dividends declared in H1 2025 totaled $2.628 per share, compared with $2.125 in H1 2026. The 10-Q says it expects to raise the dividend per share as cash flows grow. At $1.13 of AFFO per share, the Q2 dividend used about 94% of AFFO (our calculation), which leaves little room for increases until AFFO grows.
Guidance (updated July 22, 2026, earnings release)
Full-year 2026 outlook
Range
Midpoint vs FY2025
Site rental revenues
$3,833M–$3,878M
$3,855M vs $4,049M (-5%)
Adjusted EBITDA
$2,665M–$2,715M
$2,690M vs $2,863M (-6%)
AFFO
$1,950M–$2,000M
$1,975M vs $1,904M (+4%)
AFFO per share
$4.53–$4.65
$4.59 vs $4.36 (+5%)
Net income (incl. discontinued)
$730M–$1,010M
$870M vs $444M
Discretionary capex
$150M–$250M
—
Compared with the May 1 outlook, the AFFO midpoint rose by $5 million:
site rental revenue +$5M
operating costs lower: site costs -$10M and overhead -$15M
interest expense -$5M
offset by services margin cut by $20M on lower activity
Organic growth excluding DISH and Sprint is guided at about $135 million, or 3.4% (3.6% excluding DISH from the prior-year base). The 10-Q expects about $220 million less site rental revenue from DISH this year. The 2026 restructuring plan, a headcount reduction, is expected to save about $55 million in 2026 and $65 million a year at full run-rate, at a cost of about $25 million in charges ($14 million recorded so far).
Takeaway: The sale turned Crown Castle's per-share cash earnings upward even though the towers themselves earned less. Adjusted EBITDA fell 4.3%, mostly because DISH stopped paying. AFFO per share still rose 10.8% because debt, and with it interest expense, shrank by about $6 billion. That boost comes once. After 2026, growth in AFFO and the dividend depends on the roughly 4% organic tower growth, on churn falling once DISH and Sprint roll off, and on how much of the DISH claim is recovered.
Outlook: our read
Near term: the full-year AFFO guide implies about $1.04 billion in H2 ($1,975M midpoint minus $934M in H1). That is roughly $520 million a quarter, above Q2's $488 million. That is consistent with Q3 and Q4 carrying a full quarter of lower interest cost (plus the July repayment of $1.0 billion of notes), though it also relies on the restructuring savings arriving on schedule. The heavy dividend payout means little cash is left over for new buybacks without taking on more debt.
What to watch next quarter: (1) whether organic growth holds near 4%; (2) any recovery from DISH through the bankruptcy or the FCC trust, which would be pure upside against a guide that assumes zero; (3) the final settlement of the $124 million purchase-price adjustments on the fiber sale; (4) how the March 2027 maturities are handled, repaid from cash or refinanced at today's higher rates versus the 2.9%–4.0% coupons they carry.
Structural risk: with about 93% of revenue from three carriers and leverage above 6x, further carrier consolidation or slower network spending would hit Crown Castle harder than a diversified landlord. Its growth now rests entirely on the US tower business.