Built by Grok 4.6 HighCrown Castle Inc. reports Wednesday, October 21, 2026. Prepared Thursday, September 24, 2026. This is the 2026Q3 operations quarter, ending September 30, and it has not printed. The date is estimated from FMP and the historical Wednesday pattern. It is not posted on company IR, and the time of day is not confirmed. Last reported quarter is 2026Q2, released after the close on July 22, 2026. Crown Castle is a calendar-year tower REIT. The homepage calendar labels October 21 as 2026Q4. That is the announcement-month quarter, one quarter ahead. This page is 2026Q3.
CCI | Earnings Preview
| FY2026 site rental | $3,833–$3,878 million · mid $3,855.5 million · −4.8% YoY | FY2026 AFFO | $1,950–$2,000 million · mid $1,975 million · +3.7% YoY · per share $4.59 |
| Organic ex-Sprint/DISH | 3.4% · more than 90% contracted at June 30 · core leasing $60–70 million | Q3-specific swing | $20 million services cut, mainly this quarter · EBITDA held flat |
Crown Castle reports an unconfirmed Wednesday, October 21 print of a quarter that is not guided. The live bar is the full-year 2026 outlook from July 22, and nothing since then has moved it. July added only $5 million to the site-rental midpoint and $5 million to the spoken AFFO bridge. Adjusted EBITDA stayed at $2,665 million–$2,715 million because a $15 million cost save was absorbed by a $20 million drop in services contribution, aimed primarily at the third quarter. Organic growth excluding Sprint and DISH was nudged 10 basis points to 3.4%. More than 90% of that growth was already contracted at June 30. That is a floor on the leasing book, not a sandbag, and it is not a quarterly range the company can miss.
Growth trajectory. Reported revenue is still falling, and the fall is not turning. Towers-basis net revenue went from −4.3% in 2025Q3 and −4.2% in 2025Q4 to −4.8% in 2026Q1 and −4.9% in 2026Q2 ($1,008 million versus $1,060 million). FMP’s $1,010.6 million for this quarter is −5.7% versus 2025Q3 $1,072 million. The path is −4.3%, −4.2%, −4.8%, −4.9%, then a Street −5.7%. Still negative, and a bit worse. Site rental, the repeatable line, improved at the margin: 2026Q1 was −4.9% and 2026Q2 was −4.1% ($967 million versus $1,008 million). Services did the opposite, from about −2% in Q1 to −21.2% in Q2 ($41 million versus $52 million), and pulled the total back down.
The line management wants judged moved the other way inside the year and is still down versus 2025. Organic contribution excluding Sprint was 5.2% in 2025Q3, 4.7% in 2025Q4, 3.1% in 2026Q1, and 3.9% in 2026Q2. Q2 is −80 basis points versus the year-ago 4.7%. The full-year point of 3.4% sits under the Q2 print, so the second half does not have to reaccelerate for the year to land. Core leasing, the cash piece inside that rate, is stuck: $15 million for a second quarter, −46% versus $28 million. Full-year 2025 sprint-adjusted organic was 4.9%. The 2026 guide is about 150 basis points below that. The deceleration is a 2026 event, not a five-year fade. 2022–2025 sprint-adjusted organic was 5.1%, 4.0%, 4.5%, then 4.9%.
AFFO is the line that inflected, and it inflected on capital structure. 2026Q1 AFFO was $446 million, −6.9% versus $479 million. 2026Q2 was $488 million, +9.9% versus $444 million. Adjusted EBITDA did not follow: $675 million, −4.3% versus $705 million, margin about 67.0% versus 66.5%, +45 basis points, and flat versus 2025Q3’s 67.0%. The call tied the Q2 dollar increase to a $35 million year-over-year decline in interest expense and $14 million of interest income on the May 1 sale proceeds, and said the interest income does not recur in the second half. Do not carry +9.9% into this print.
What consensus is underwriting. FMP, pulled September 24, is the only street snapshot. Visible Alpha and Bloomberg are not connected. The October 21 row, actuals null, is revenue $1,010.6 million and EPS $0.685. There is no FMP line for site rental, organic growth, services, or AFFO. FY2026 revenue of $4,039.4 million is −5.3% versus FY2025 net revenues of $4,264 million. FY2027 revenue of $4,089.6 million is only +1.2% versus that FY2026 average. Management has said, in February, April, July, and again on September 10, that 2026 is the organic low point. The published revenue path does not pay for the rebound. FMP’s $0.685 is closest to continuing-operations diluted EPS. 2025Q3 continuing was $0.64 and total diluted was $0.74. It is not AFFO per share, which was $1.12 in 2025Q3 and $1.13 in 2026Q2. FMP FY2026 EBITDA of $2,549 million sits below the company’s entire adjusted-EBITDA range. Those are different objects. Do not call a 5% miss versus the guide.
What to watch on October 21
- Services, the one item they already placed in this quarter. The full-year services gross-margin range was cut $20 million to $70 million–$100 million, midpoint $85 million, versus FY2025 $107 million, −20.6%. First-half gross margin was $24 million plus $23 million, or $47 million. Hitting $85 million leaves about $38 million for the second half, versus $59 million a year ago. A result inside that cut supports the unchanged EBITDA guide. A further miss is how EBITDA goes through $2,665 million.
- Organic near 3.4%, not another 3.9%. Q1 was $30 million and Q2 was $38 million of a $135 million year, so the second half is about $67 million, closer to Q1 than to Q2. A quarter above 3.6% excluding the other-billings bump would be the first evidence the trough is behind them. A sub-3% quarter would break the “more than 90% contracted” message.
- Core leasing versus $60–70 million. First half is $30 million. The low end needs no acceleration. The midpoint needs a second half larger than the first. On the July call the book was described as trending toward $60 million, and the CFO deferred the rest of that conversation to this print.
- DISH phasing, not the claim. First-half terminations were about $98 million of a −$220 million year. About $122 million sits in the second half, and management said that was expected and back-end loaded. Reported site rental can step down versus the first half without a guide cut. The $3.5 billion claim and the funded $2.4 billion trust are not in the quarter.
- AFFO without the Q2 interest income. The year midpoint of $1,975 million minus first-half AFFO of $934 million leaves about $1,041 million for the second half, +6.3% versus second-half 2025. An even split is about $520 million, not a repeat of +9.9%. The only clean beat path versus the company’s own history of finishing at the high end is the top of the $2,000 million band, not another $50 million.
American Tower and SBA Communications do not report this quarter before Crown Castle. Third-party calendars put AMT around October 27 and SBAC around November 2, and neither date is on company IR. There is no fresh tower print to hide behind. Internal models, SharePoint, OneNote, Outlook, and Excel were unavailable.
Crown Castle guides the year, not the quarter. There is no Q3 range for revenue, EPS, margin, leasing, or AFFO. The only quarter-specific number on the July 22 call is directional: the $20 million full-year reduction in services contribution is driven by lower activity, primarily in the third quarter. Classification of the live year: in-line, and tight. Not conservative. Not aggressive. Not a sandbag.
FY2025 finished near the high end of site rental and through the high end of adjusted EBITDA and AFFO. That habit is real, and it is the wrong template for this band. Versus February, the AFFO midpoint is up about $55 million. About $50 million of that was the May 1 early close, which was interest, already in the May 1 outlook. July’s operating raise was $5 million of other billings on site rental and $5 million of lower interest expense on AFFO. AFFO per share was held at $4.59 versus May 1. EBITDA has not moved since the initial guide, because the cost save was spent on the services offset. More than 90% of ex-Sprint/DISH organic growth is contracted, the leasing range was not raised, and first-half core leasing of $30 million has the year tracking the low end of $60–70 million. A guide with that much of the year already booked, and with the slack inside EBITDA already used, is not a lowball.
The illustrative second half is arithmetic, not a guide. First-half site rental was $961 million plus $967 million, or $1,928 million. The $3,855.5 million midpoint leaves $1,927.5 million for the second half, −5.1% versus second-half 2025 site rental of $2,031 million. First-half site rental was −4.5% year over year. The year still requires a second half that is slightly worse, because the DISH and Sprint remainder is back-end loaded and straight-line revenue steps down again. First-half adjusted EBITDA of $1,350 million leaves $1,340 million for the second half, −6.7% versus second-half 2025. That is the services cut, not a margin-expansion year. Management’s “couple of hundred basis points over the next year” is a 2027 direction. It is not in the 2026 EBITDA midpoint.
| FY2026 guide | FY2025 | Low | High | Mid | YoY | Read |
|---|---|---|---|---|---|---|
| Site rental revenue | $4,049 | $3,833 | $3,878 | $3,855.5 | −4.8% | July was +$5 million of other billings. Leasing range was not raised. |
| Adjusted EBITDA | $2,863 | $2,665 | $2,715 | $2,690 | −6.0% | Unchanged since February. FMP EBITDA of $2,549 million is a different metric. |
| AFFO | $1,904 | $1,950 | $2,000 | $1,975 | +3.7% | The only guided P&L line that grows. High end is the bull case. |
| AFFO per share | $4.36 | $4.53 | $4.65 | $4.59 | +5.3% | Unchanged versus the May 1 outlook. The step-up was the buyback, not July. |
| Organic ex-Sprint/DISH | 4.9% | 3.4% | 3.4% | 3.4% | −150 bps | Point, not a range. 3.6% if DISH is also out of the prior-year base. |
| Core leasing | $118 | $60 | $70 | $65 | −45% | Range is the call. Daloopa stores the midpoint. First half is $30 million. |
| Services gross margin | $107 | $70 | $100 | $85 | −20.6% | Cut $20 million versus April. The hole is aimed at Q3. |
Versus the street, where the definitions match, the year is also in-line. FMP does not publish site rental or AFFO, so those lines cannot be scored as a gap. FMP FY2026 revenue of $4,039.4 million leaves about $2,021 million of total revenue for the second half after a $2,018 million first half. That is roughly the site-rental midpoint plus about $93 million of second-half services revenue, versus $113 million in second-half 2025. Same direction as the services cut. FMP Q3 revenue of $1,010.6 million matches an even split of that year. It is mildly full if the services air pocket hits reported revenue and not only gross margin. It is not a model of a leasing miss. The leasing book is mostly contracted.
FMP Q3 EPS of $0.685 is not the beat bar. The company GAAP diluted range of $1.70–$2.35 is $0.65 wide because discontinued operations are still in the bridge. FY2026 FMP EPS of $1.92 sits inside that range and does not match one GAAP series. FY2025 total diluted was $1.01. Continuing diluted was $2.52. FY2027 FMP EPS of $2.813 is +46% versus FY2026e EPS on +1.2% revenue. That is not an earnings inflection you can underwrite from the revenue path.
Guidance credibility is informative. They are not chronic missers. FY2024, still a consolidated company, finished inside the last pre-year-end band, beating the midpoint on site rental and AFFO and missing the EBITDA midpoint by $7 million. FY2025, tower-only, beat the last EBITDA high by $3 million ($2,863 million versus a high of $2,860 million) and the last AFFO high by $9 million ($1,904 million versus $1,895 million). They will cut a line when the fact pattern changes. They cut FY2024 site rental at the Q2 update, and they cut FY2026 services gross margin by $20 million rather than defend it. What they have not done in 2026 is the FY2025 walk-up. Applying “they always beat the high” to this year’s AFFO double-counts the May interest reset.
Business type is a pure-play U.S. tower REIT after the small-cell and fiber sale closed May 1, 2026. The stack above total revenue is site rental (billings, straight-line, organic contribution, core leasing) and services. Gross margin, operating margin, and a quarterly EBITDA margin are not guided. Internal is n/a on every line. Adjusted EBITDA and AFFO before 2025 in the table below are the pre-sale company. They are shown so the break is visible. They are not the year-ago base. YoY is the same quarter a year earlier.
Eight reported quarters
| Metric | Q3 24 | Q4 24 | Q1 25 | Q2 25 | Q3 25 | Q4 25 | Q1 26 | Q2 26 |
|---|---|---|---|---|---|---|---|---|
| Site rental ($M) | $1,066 | $1,070 | $1,011 | $1,008 | $1,012 | $1,019 | $961 | $967 |
| Site rental YoY | — | — | — | — | −5.1% | −4.8% | −4.9% | −4.1% |
| Services revenue ($M) | $54 | $49 | $50 | $52 | $60 | $53 | $49 | $41 |
| Services revenue YoY | — | — | — | — | +11.1% | +8.2% | −2.0% | −21.2% |
| Total revenue ($M) | $1,120 | $1,119 | $1,061 | $1,060 | $1,072 | $1,072 | $1,010 | $1,008 |
| Total revenue YoY | — | — | — | — | −4.3% | −4.2% | −4.8% | −4.9% |
| Organic ex-Sprint | 5.2% | 4.6% | 5.1% | 4.7% | 5.2% | 4.7% | 3.1% | 3.9% |
| Organic rate vs year ago | — | — | — | — | 0 bps | +10 bps | −200 bps | −80 bps |
| Core leasing ($M) | — | — | $28 | $28 | $33 | $29 | $15 | $15 |
| Core leasing YoY | — | — | — | — | — | — | −46% | −46% |
| Adj. EBITDA ($M) | $1,075 | $1,044 | $722 | $705 | $718 | $718 | $675 | $675 |
| EBITDA YoY, tower base | — | — | — | — | — | — | −6.5% | −4.3% |
| AFFO ($M) | $801 | $785 | $479 | $444 | $490 | $489 | $446 | $488 |
| AFFO YoY, tower base | — | — | — | — | — | — | −6.9% | +9.9% |
| AFFO per share | $1.84 | $1.80 | $1.10 | $1.02 | $1.12 | $1.12 | $1.02 | $1.13 |
| AFFO per share YoY, tower base | — | — | — | — | — | — | −7.3% | +10.8% |
| Interest expense ($M) | $236 | $240 | $236 | $243 | $247 | $246 | $242 | $208 |
| Interest expense YoY | — | — | — | — | +4.7% | +2.5% | +2.5% | −14.4% |
Site-rental YoY improved by about 80 basis points from Q1 to Q2 and is still negative. Total-revenue YoY worsened by about 10 basis points over the same comparison, because services collapsed. EBITDA YoY is less negative (−6.5% to −4.3%) and still negative. AFFO flipped from −6.9% to +9.9% in one quarter, in the same quarter interest expense fell from $243 million to $208 million, −14.4%. In Q1, before the May 1 close was in the quarter, interest expense was still higher year over year and AFFO was down. Margin is not on a clean expansion path into this print. Q2 EBITDA margin of about 67.0% matched 2025Q3 and was only +45 basis points versus 2025Q2. Q1 margin was about 66.8% versus 68.0% a year earlier, −122 basis points.
Reported organic contribution, which still includes Sprint and DISH, was −1.8% in Q2 versus −0.6% a year earlier, −120 basis points. That is the line the tape should not confuse with the 3.9% ex-churn rate. Escalators are the stable piece, about $25 million a quarter, and the $100 million full-year line is that run-rate. Most of the “more than 90% contracted” comment is escalators, not new leasing.
Five-year organic rate
Sprint-adjusted organic contribution to site rental billings is the series that survives the fiber sale. Tower count did not grow. The 2026 step-down in tenants per tower is DISH coming off the book in Q1 2026, to 2.0, and it was still 2.0 in Q2. It is not a slow leak across the prior five years.
| Metric | 2021 | 2022 | 2023 | 2024 | 2025 | 2026 guide |
|---|---|---|---|---|---|---|
| Sprint-adjusted organic | — | 5.1% | 4.0% | 4.5% | 4.9% | 3.4% |
| Change in the rate | — | — | −110 bps | +50 bps | +40 bps | −150 bps |
| Towers | 40k | 40k | 40k | 40k | 40k | flat |
| Tenants per tower | 2.3 | 2.4 | 2.5 | 2.4 | 2.4 | 2.0 in Q2 |
| Site rental, tower basis ($M) | — | — | — | $4,268 | $4,049 | $3,855.5 |
| Site rental YoY, tower basis | — | — | — | — | −5.1% | −4.8% |
Current quarter — 2026Q3
No company midpoint, so % Diff versus a guide is n/a. The two consensus cells are FMP versus the year-ago quarter.
| Metric | Q3 2025 | Q2 2026 | Q3 guide | Consensus | Internal | What the bar is |
|---|---|---|---|---|---|---|
| Total revenue | $1,072 | $1,008 | Not Guided | $1,010.6 | n/a | −5.7% YoY. A further deceleration from −4.9%. |
| Site rental | $1,012 | $967 | Not Guided | n/a | n/a | Even split of the $1,927.5 million second-half remainder is about $964 million, −4.8% versus Q3 2025. The −5.1% is the half versus second-half 2025, not this cell. Not a guide. |
| Services revenue | $60 | $41 | Not Guided | n/a | n/a | Q2 was −21.2% YoY. The new weakness is aimed here. |
| Services gross margin | $31 | $23 | Not Guided | n/a | n/a | FY range cut $20 million, primarily this quarter. Q2 was −11.5% versus year-ago $26 million. |
| Core leasing | $33 | $15 | Not Guided | n/a | n/a | A third $15 million quarter puts the year on the low end. |
| Organic ex-Sprint/DISH | 5.2% | 3.9% | Not Guided | n/a | n/a | FY point is 3.4%. The year does not need a 3.9% repeat. |
| Adjusted EBITDA | $718 | $675 | Not Guided | n/a | n/a | Even split of the year is about $670 million, −6.7%. Not a guide. |
| AFFO | $490 | $488 | Not Guided | n/a | n/a | Even split about $520 million, +6%. Q2’s +9.9% does not recur. |
| AFFO per share | $1.12 | $1.13 | Not Guided | n/a | n/a | FMP $0.685 is not this line. Illustrative even split is about $1.22. |
| FMP EPS | $0.64 cont. | $0.69 cont. | Not Guided | $0.685 | n/a | +7.0% vs continuing Q3 2025. −7.4% vs total GAAP $0.74. Check the definition on the night. |
FQ+1 — 2026Q4, not guided
Crown Castle will not issue a Q4 range on this call either. FMP’s quarterly file in the workspace has no 2026Q4 revenue or EPS estimate. The items that still cover “the rest of the year” are the FY ranges, the statement that the services hole is primarily Q3 so Q4 is the lesser piece, the straight-line step-down in the second half, and the reiterated $2.1 billion midpoint for second-half 2026 plus first-half 2027 AFFO. That $2.1 billion bridge is not an FY2027 range and not a Q4 number. It dates to the February cut after DISH was removed.
| Metric | Q4 2025 | Q4 2026 guide | Consensus | Internal | Read |
|---|---|---|---|---|---|
| Total revenue | $1,072 | Not Guided | n/a | n/a | No FMP Q4 row in the workspace file. |
| Site rental | $1,019 | Not Guided | n/a | n/a | Second-half residual is the FY midpoint minus the first half, not a Q4 guide. |
| Services revenue | $53 | Not Guided | n/a | n/a | Call put the new weakness mainly in Q3, not Q4. |
| Core leasing | $29 | Not Guided | n/a | n/a | $30–40 million still needed across Q3 and Q4 together. |
| AFFO | $489 | Not Guided | n/a | n/a | H2 2026 plus H1 2027 midpoint remains $2.1 billion. Not a Q4 figure. |
FY guide in force, and FY2027 not guided
FY2026 is the guide already on the books. FY2027 is Not Guided. FMP’s +1.2% revenue step and +46% EPS step are street math, not a company outlook.
| Metric | FY2025 | FY2026 mid | FY2026 YoY | FY2027 | Street, FMP |
|---|---|---|---|---|---|
| Site rental | $4,049 | $3,855.5 | −4.8% | Not Guided | FMP does not publish site rental. |
| Total revenue | $4,264 | Not Guided | — | Not Guided | FY2026e $4,039 million, −5.3%. FY2027e $4,090 million, +1.2%. |
| Adjusted EBITDA | $2,863 | $2,690 | −6.0% | Not Guided | FMP $2,549 million in 2026 and $3,027 million in 2027. Different definition. |
| AFFO | $1,904 | $1,975 | +3.7% | Not Guided | No AFFO consensus. The $2.1 billion bridge is H2 2026 plus H1 2027, not FY2027. |
| AFFO per share | $4.36 | $4.59 | +5.3% | Not Guided | Dividend of $4.25 is about 93% of the midpoint. Target payout is 75–80% by growing AFFO, not by cutting. |
| Organic ex-Sprint/DISH | 4.9% | 3.4% | −150 bps | Not Guided | Barclays sketches about 4.2% next year. JPMorgan sketches about 3%. FMP revenue does not pick a side. |
The tone arc is not a steady improvement in confidence. It is a regime change in February, a hold in April, a financing update in May, and a small operating mix shift in July that took the cost-save upside away.
Q3 2025, Christian Hillabrant’s first call as CEO, raised the FY2025 midpoints and described demand as still strong. Services were a positive. DISH was not yet a default. Q4 2025 delivered the year above the midpoint and through the high end on EBITDA and AFFO, then set 2026 as the low point: zero DISH revenue, organic 3.3%, about 80% already contracted, and a post-close 12-month AFFO bridge cut to a $2.1 billion midpoint. Q1 2026 reiterated every major line and refused to annualize a seasonal cost underspend. May 1 closed the fiber sale two months early. Operating lines did not change. The AFFO midpoint rose $50 million on interest. July 22, versus that May 1 outlook, added $5 million of site rental, held EBITDA, cut services contribution by $20 million mostly in Q3, cut costs by $15 million, and added $5 million of AFFO. Organic went to 3.4%, or 3.6% if DISH is also out of the prior-year base. The contracted share went from about 80% to more than 90%.
What got more confident: the sale is behind them, the organic book is mostly contracted, cost reduction is described as durable, and the DISH process is a bankruptcy with a committee seat rather than a standalone lawsuit. What got less confident: services, which flipped from a small guide tailwind in February (a $5 million contribution in the April AFFO walk) to a −$15 million drag, dated to this quarter; and any suggestion that Q2’s AFFO print is the new base. The 2027 growth language is the same speech as February. It has not been upgraded to a number. Hillabrant also said they are not ready for multi-year guidance. That is a team that knows the street does not yet have the proof.
The July bridge, in the year-over-year AFFO walk versus 2025, is why EBITDA did not move. Organic contribution to site rental billings went from −$110 million to −$105 million. The change in expenses went from +$25 million to +$40 million. The change in services contribution went from +$5 million to −$15 million. Interest did the rest. AFFO growth in that walk is +$70 million, which lands on the $1,975 million midpoint against FY2025 $1,904 million. Patel’s answer to “why isn’t the beat flowing through” was this table. He did not invite a high-end EBITDA outcome.
Assumptions that have to stay true for the guide to hold:
- The uncontracted slice of organic growth shows up at least at the low end of $60–70 million of core leasing. Activity had “played out as expected.”
- The $20 million services hole is the whole hole, and it is mostly Q3. There is, in their words, no straight line from services to leasing.
- DISH stays at −$220 million and Sprint at −$20 million. A bankruptcy settlement would be upside outside the guide.
- The $15 million cost reduction is durable. The Q1 repair, maintenance, and sustaining-capex underspend was seasonal and was expected to reverse in the same half as the services cut.
- Interest expense stays inside $787 million–$832 million. The elevated interest income is gone.
- Straight-line revenue stays around a negative $60 million midpoint, worse in the second half. That is a different figure from the −$70 million year-over-year change in the straight-line adjustment. GAAP site rental can look worse than billings even when organic is on the guide.
A “solid” Q3, in their own framework, can have site rental down again year over year and down versus Q2. That is the back-end-loaded churn plus straight-line, not a miss against their language. Organic should land near 3.4%, not repeat 3.9%, unless leasing or other billings run ahead. Services should be the weak line. AFFO should be judged against the $2.1 billion bridge after interest income drops out, not against $488 million. Any dollar of DISH escrow cash in 2026 AFFO would be new. Through September 24 they have not put one in, and the September 10 conference repeated no DISH revenue in 2026.
Since July 22 the guide has not been revised. September 10, Kris Hinson restated organic growth of 3.6% excluding Sprint attrition as the low point and said 2026 activity was somewhat back-end loaded. That 3.6% is the existing presentation with DISH removed from the prior-year base. It is not a raise. The September 29 RBC conference is still ahead of this preview. Nothing from it is in the guide. The CFO and CEO pair on the last three calls is the pair for this one. Patel stays through the FY2026 10-K. The operating-chief seat is already empty as of September 23. That is a stability point on the person who will defend the numbers, and a vacancy on the person who runs the business. It is not a guidance event.
Consensus is underwriting the 2026 guide. FMP’s −5.7% revenue bar and +1.2% FY2027 path sit on a 3.4% organic year. They do not pay for a post-2026 rebound. The contradiction is that management keeps calling 2026 the low point, and the quarterly cadence does not yet agree: core leasing is stuck at $15 million, and the full-year organic guide is below the Q2 rate. That is not yet a contrarian long. The trough call has been repeated since February and has not been given a 2027 number. The Q3 test is narrower than the narrative.
| Catalyst | Last print and the comp | What the bar assumes | Why it matters |
|---|---|---|---|
| Organic rate, and whether 2026 is the low | Q2 3.9%, −80 bps vs 4.7%. FY guide 3.4%. 2025 was 4.9%. | No organic line in FMP. Revenue of $1,010.6 million fits a quarter inside 3.4%. It does not fit the start of a rebound. FY2027 revenue +1.2% does not either. | A print near 3.4% with no raise is what the revenue bar already assumes. A raise through 3.6% on the definition that keeps DISH in the base would be the first evidence the trough is behind them. |
| Services, placed in Q3 | Revenue $41 million, −21.2%. Gross profit $23 million vs $26 million. Q3 2025 gross profit was $31 million. | No services line in FMP. The AFFO-walk services item swung from +$5 million to −$15 million. | The Q3-specific operating risk. EBITDA did not move because the services hole was paired with about $15 million of costs. A miss beyond that bridge pushes EBITDA through $2,665 million. A normal seasonal dip would mean the cut was too large. |
| Core leasing vs $65 million | $15 million for a second quarter, −46% vs $28 million. First half $30 million. | FMP revenue does not require the high end. Landing at $60 million means the second half stays at the current pace. | Escalators cannot produce the rebound. They are already at the $100 million guide. A third $15 million quarter, with organic left at 3.4%, tells the street the 2027 rebound is still a statement. |
| DISH timing inside the year | −$49 million in Q2, same as Q1 −$49 million. Year is −$220 million. About $122 million remains. | Consensus is not modeling a DISH surprise. The guide says the churn is contracted and back-end loaded. Tenants per tower already stepped to 2.0. | Q3 cannot miss DISH relative to the company’s own math unless the quarterly phasing differs. That moves the quarter, not the year. Sprint is the opposite: the quarterly non-renewal shrank to −$5 million. |
| DISH claim and the trust | Not in revenue. Q2 10-Q: claim above $3.5 billion, revenue recognition stopped January 1, net book position about $165 million that management expects to recover. Trust funded July 28. | No claim or trust proceeds in the FMP revenue or EPS estimates. | Cash in Q3 would be a surprise, and the documents say not to expect it. The July 30 FCC order keeps an EchoStar intercompany loan off the fund. EchoStar has asked for that order to be reversed. Slower is better for the claim. It is not Q3 AFFO. |
| Interest, and why AFFO and EBITDA diverge | Interest $208 million, −14.4%. Net debt $17,099 million, −29.3% versus $24,174 million a year ago. The step is the May 1 close. Q1 net debt was still $24,577 million. | No quarterly AFFO estimate. Guide arithmetic is about +6% AFFO in the second half after the interest income drops out. Q3 is the first full quarter of the lower interest run-rate. | July’s extra $5 million of AFFO was debt-paydown timing, not the quarter’s leasing. Consensus that “AFFO is growing” is consensus that the interest save persists. Leverage of 6.3x is management’s figure and matches net debt over four times Q2 EBITDA. It is inside 6.0–6.5x. It is calculated, not a tagged fundamental after 2024. |
| Spectrum, edge, and the dividend | AT&T closed the EchoStar spectrum purchase on July 28. Not in the Q2 leasing number. Edge is a trial on existing sites. Dividend held at $1.0625. The $1 billion buyback is finished. | None of that is in the $1,010.6 million quarter. FMP’s +1.2% in 2027 does not capitalize a 600 MHz amendment cycle. The dividend is already declared for September 30. | Management is earlier than the model, and has been consistent. 600 MHz is a reason activity improves after 2026, not a this-quarter deployment. Edge can be real in 2027 and irrelevant to whether Q3 lands on the guide. A new buyback would ignore the order he repeated: dividend, then capex, then the leverage band, then buybacks. |
Ground-lease purchases are rising and are not yet a mix change. Q2 land-interest purchases were $36 million, +125% versus $16 million. Ground leased is still 57%, unchanged from a year ago. Hillabrant said Crown Castle is about 11 points behind American Tower and SBA on land ownership and intends to close that gap over a couple of years. Q3 can confirm the spend. It cannot confirm the gap is closing. The $15 million cost save is already inside the unchanged EBITDA guide.
Peer color does not change the CCI number, and none of it prints first. American Tower’s June quarter reiterated a low reported U.S. organic rate with a mid-single-digit rate once DISH is excluded. SBA described carrier activity as steady and is not a September datapoint for this quarter. The industry narrative and Crown Castle’s narrative match. Crown Castle’s published revenue consensus does not pay for the next leg.
Window is July 23 through September 24, after the Q2 report and before this print. Nothing in it is a new guide, a completed deal that changes the tower P&L, or a reported Q3 figure. The dividend was held. The DISH recovery vehicle was funded and then contested. A senior-leadership handoff does not change who reports Q3. Conference commentary restated the trough-year view. Sell-side organic paths for the years after 2026 now disagree, which is the setup, not a Q3 number.
| Date | Item | Source | Read for this print |
|---|---|---|---|
| Jul 28 | AT&T closes the EchoStar spectrum purchase. $2.4 billion goes into the Wireless Creditor Trust. | DISH bankruptcy notice of closing; EchoStar disclosure; Crown Castle 10-Q filed Aug 5 | The company says its master-lease claims are covered claims and that payment still depends on the trust’s rules. Net book position on the terminated agreements is about $165 million. The contractual claim remains above $3.5 billion. None of it is Q3 site rental. The close removes the “if it closes” caveat on AT&T spectrum. It does not pull 600 MHz deployment into this quarter. |
| Jul 30 | FCC says EchoStar intercompany repayment claims are not claims on the trust. | FCC DA-26-797 | Keeps an affiliate loan, later described as about $8.8 billion, from competing for the same $2.4 billion. Better for tower creditors. Does not set Crown Castle’s recovery or put cash in the quarter. |
| Aug 5 | Quarterly dividend held at $1.0625, payable Sept 30. | Crown Castle IR; 8-K Item 7.01 | Same rate as May 20, 2026. Annualized $4.25 is about 93% of the $4.59 AFFO-per-share midpoint, above the 75–80% target. The framework is to hold $4.25 until AFFO grows into that payout. Not a raise and not a new cut. |
| Aug 20 | Barclays moves to Overweight. Its own model has organic contribution near 4.2% next year. | Investing.com, citing Brendan Lynch; The Fly | The argument is that satellite fear is overdone and that the services-margin cut is a weak lead indicator for leasing. It is a 2027 debate. It does not change the July 22 range or the Q3 consensus. |
| Aug 31 | EchoStar asks the full FCC to reverse the July 30 trust order. | Unsecured creditors’ committee filing, describing the Aug 31 application | If the clarification is overturned, the intercompany claim can compete again and recovery odds fall. No decision. Recovery stays outside the model. |
| Sep 10 | Hinson calls 2026 the organic low point. AI is not in the 2026 numbers. | Bank of America conference; RCR Wireless write-up of the same remarks | 3.6% excluding Sprint attrition is the existing guide with DISH out of the base, not a new range. Activity was described as back-end loaded, which supports 2H leasing inside $60–70 million and does not raise the 3.4% year. The edge trial uses existing sites. He said the company is not spending meaningful capex on it. |
| Sep 14 | JPMorgan moves to Underweight. Its model is about 3% organic for the next few years. | The Fly via TipRanks; Investing.com | The stated reason is leasing, not the executive changes. About 3.0% organic from 1.8% new leasing, 1.4% churn, and 2.6% escalators. That is below the company’s 3.4% and below Barclays’ 4.2% for 2027. The useful tension is activity: management says back-end loaded, and this note says activity is not turning. |
| Sep 14 | Patel retires March 31, 2027. Piche leaves the COO seat September 23. Hinson is to become CFO April 1, 2027. | Press release and 8-K, Items 5.02 and 7.01 | Patel is the CFO on this call and on the year-end print. The handoff is internal and dated. The operating seat is already vacant, and the commercial role Hinson leaves in April 2027 has no named successor. Not a guidance event. Two senior exits before a print is still a flag. |
| Sep 29 | RBC conference, not yet held as of this preview. | Company IR calendar | The one scheduled appearance before October 21. Nothing from it can be in the guide yet. Citi and Goldman on September 9 did not produce an open-source number that moved the range. |
Left out on purpose: the July 22 release itself, target cuts that kept ratings in the earnings-reaction window, conference-appearance notices, the September 15 ex-dividend mechanics, and a September 17 Form 4 that withheld shares for taxes on vested equity. Those do not change the outlook. Price is not used as a thesis input.
What matters for October 21 is qualitative. The FMP bar and the July 22 range are untouched. The dividend was held at a payout that still sits above the target. The trust is funded, currently shielded from the intercompany claim, and that shield is under review. Hinson told investors 2026 is the trough and that AI is not in the 2026 forecast. Barclays and JPMorgan now bracket the post-2026 organic rate at roughly 4% versus roughly 3%. Patel will still deliver the quarter. The operating role is already vacant.
Consistent revenue beater. Mixed, and lately deteriorating, on GAAP diluted EPS. Over the 12 reported quarters from Q3 2023 through Q2 2026, Crown Castle beat FMP revenue consensus 11 times and missed once (91.7%). The last four quarters are 4 for 4, but the beat has stopped widening: +1.5%, +1.1%, +1.5%, +1.3%. The latest print is $1,008 million versus a $995.1 million estimate, +$12.9 million. GAAP diluted EPS beat in 7 of 12 (58.3%) and in only 2 of the last 4. The path on that line over the last four quarters is +41.2%, +21.4%, −9.7%, −43.6%. Q3 2026 has not printed.
FMP’s own Q2 2026 actual of $0.69 is not that GAAP total. It equals continuing-operations diluted EPS of $0.69. Total diluted EPS was $0.22. Against the same $0.3904 estimate, FMP’s database prints a +76.7% EPS beat and the consistent GAAP series prints a −43.6% miss. The heatmap uses the consistent series. Revenue actuals equal FMP in all 12 quarters. GAAP diluted EPS equals FMP in 9 of 12, within a penny in two more, and breaks only in Q2 2026. There is no 12-quarter AFFO consensus, so AFFO is not a scored row. Daloopa has no Company Metrics series that stores a CCI surprise.
The −35% as-reported revenue declines through 2025 are the fiber business leaving the print. FMP’s revenue bar stepped down with the base, so the surprise after Q1 2025 is versus a towers-basis estimate. Towers-basis revenue YoY underneath those beats is still −4.3%, −4.2%, −4.8%, −4.9%. The level is decelerating. The surprise is not.
| Surprise vs FMP | Q3 23 | Q4 23 | Q1 24 | Q2 24 | Q3 24 | Q4 24 | Q1 25 | Q2 25 | Q3 25 | Q4 25 | Q1 26 | Q2 26 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Revenue | −1.4% | +1.0% | +0.7% | +0.2% | +0.9% | +0.7% | +2.0% | +2.0% | +1.5% | +1.1% | +1.5% | +1.3% |
| GAAP diluted EPS | −20.4% | +9.9% | +0.4% | +3.9% | +4.3% | loss | loss | +22.7% | +41.2% | +21.4% | −9.7% | −43.6% |
Green is a beat. Solid green is a revenue surprise of about 2% or a GAAP EPS surprise above about 8%. Yellow is a beat inside half a point, too small to call a pattern. Red is a miss. The Q4 2024 and Q1 2025 EPS cells are labeled loss rather than printed as −2,593% and −641%. Those percentages are a goodwill impairment and a discontinued-operations loss, and they are off any linear scale. Q4 2024 GAAP diluted EPS was −$10.97, including goodwill impairment of $4,958 million. Q1 2025 total diluted EPS was −$1.07, which is continuing-operations $0.65 plus discontinued-operations −$1.72. Q2 2026 is red on total GAAP because discontinued operations were −$0.47. Asset write-downs that quarter were only $2 million. The miss is the discontinued line, not a write-down inside the tower business. FMP’s published +76.7% cell for that quarter is not shown.
Read left to right. The revenue row is one miss, then eleven beats. The dark cells are Q1 and Q2 2025 at +2.0%. The row fades and stays green. Dollar beats peaked near $21 million in the first half of 2025. The last four are $16.2 million, $11.5 million, $15.2 million, and $12.9 million. The four-quarter average revenue surprise went from about +0.1% to +1.4% and then stayed at +1.4%. Magnitude improved, then stalled. The EPS row is not a streak. Small beats, two loss quarters, two large percent beats in the second half of 2025, then two misses, and the miss got larger in dollars: +$0.22, +$0.12, −$0.04, −$0.17.
Management does not walk GAAP EPS versus FMP. The variance they explain is versus their own guide. Q3 2025 was a beat-and-raise against the outlook: 5.2% organic, or $52 million, more than offset at site rental by Sprint and non-cash straight-line and prepaid rent, with the EBITDA midpoint up $30 million. Q4 2025 finished near the high end of site rental and through the high end of EBITDA and FFO, and the same remarks took DISH out of 2026. Q1 2026 was called solid and the guide was reiterated. The scored result was a revenue beat and a GAAP EPS miss. Q2 2026 is the quarter an analyst called the AFFO better than expected, and Patel split the answer: cost saves are real, services weakness is why EBITDA did not move, and the extra $5 million of AFFO is interest timing. The call’s claim is an AFFO beat on cost and capital structure, not a leasing beat, and not a beat of FMP’s GAAP line.
Read-through for October 21. The base case from the revenue record is another small beat, on the order of +1.1% to +1.5% of the $1,010.6 million bar, or about $11–15 million. A repeat of the Q2 dollar beat would be roughly $1,024 million, still about −4.5% versus $1,072 million. That extends the towers-basis decline. It does not end it. Services are inside net revenue, so the historical beat can be absorbed by the services gap they already flagged and still leave site rental in line. The base case on GAAP EPS is not a base case until the definition matches. The estimate jumped from $0.3904 for Q2 to $0.685 for Q3 after FMP booked a $0.69 actual. That is the signature of a bar that has moved toward continuing operations. Scoring $0.685 as if it were the old total-GAAP bar would make a clean tower quarter look like a miss. Scoring it as if it were still a soft total-GAAP number would make a normal continuing-operations print look like a beat. On the consistent total-GAAP series, the last two quarters missed, and the miss got larger. On continuing operations, Q2 beat only because the estimate was still $0.39. Check which EPS prints on the night before calling either.