TNYSEThe short version
AT&T Inc.
AT&T is the US wireless and fiber carrier — 74.2 million postpaid phone lines and $125.6B of FY2025 revenue — trading at $22.96 after a 31% fall from its September 2025 high.
The shares peaked at $44.67 in July 1999 and have traded around a $23.38 median since 2016; today's $22.96 is 22.5% below the $29.62 close of 2025-09-15.
$22.96
Share price, 2026-07-23
$164.8B
Market cap
5.0%
Adjusted FCF yield
−30.9%
Peak-to-trough drawdown
SwipeScroll▾
As reported
The statements before the argument: $125.6B of revenue, $16.6B of free cash flow
- Scale. FY2025 revenue $125.6B, operating income $24.2B, net income $22.0B, about 133,000 employees.
- Capital base. Net property and equipment $131.6B plus net spectrum licenses $128.1B, against total debt of $136.1B and $18.2B of cash — net debt near $118B.
- Cash. Management-reported free cash flow was $16.6B in 2025, inside a $14.1–17.6B band held every year since 2022.
AT&T publishes no free-cash-flow or consolidated adjusted-EBITDA line in its 10-K; those figures come from its earnings releases and calls.
The fit
Does not fit the framework (P1 not met); contested: P4c, P5
The pillar ledger — reference lines, not grades
| Pillar | Reference line | Where it lands |
|---|---|---|
| Year-10 gate (P1) | Binary — any proper doubt fails | Not met; 3 of 4 jurors, probability 0.675 |
| Adjusted yield (P3c) | ≥10% on a moderate balance sheet | 5.0% — about half the bar |
| Diagnosis (P5) | Temporary or permanent damage | Contested; p(temporary) 0.64, spread 0.25 |
| Dividend safety (P4c) | Covered and committed through stress | Contested; 2.0x cover, 47% rebase in 2022 |
| Dislocation, fear (P3a, P3b) | Real drawdown with volume capitulation | Met; −30.9% and a 2.37x volume spike |
| Confidence | Name-masked re-run agrees | Low — the masked run flipped P1 to met |
- The gate. Revenue higher in a decade is a high-conviction call. Adjusted free cash flow materially higher is not: reported FCF has run flat since 2022 while capex takes 16.6% of revenue. A binary gate does not bend.
- The strongest counter-fact. The name-masked jury — same evidence, ticker hidden — returned the gate met, and the probability gap crossed 0.20. The run carries a prior-driven-risk flag and low confidence.
- The yield arithmetic. $16,600M of FCF less $536M of stock compensation less a $7,871M five-year acquisition average is $8,193M — 5.0% on a $164.8B market cap, against a 10% bar.
The business
Wireless is the business; the wireline legacy is the runoff
FY2025 revenue by segment
Mobility$89.5B71%
Business Wireline$17.2B14%
Consumer Wireline$14.2B11%
Latin America (Mexico)$4.4B4%
- Mobility carries it. $89.5B of FY2025 revenue and 74.2M postpaid phone lines inside a three-carrier, FCC-licensed oligopoly; postpaid phone churn ran 0.90%.
- The runoff. Business Wireline revenue fell 8.4% to $17.2B in 2025 and now runs an operating loss; the copper network is being powered down across the large majority of the footprint by the end of 2029.
- Where it earns. Revenue comes from the US and Mexico, with Mexico about 3% of segment revenue and no customer above 10%.
The dislocation
Down 30.9% from September 2025, in two legs with a full recovery between them
Daily closes since 1990; the current episode ran 289 days from peak to trough.
- Leg one. $29.62 on 2025-09-15 to $24.62 by late October, −16.9%, around an in-line Q3 FY2025 quarter — the only company-documented trigger in the drawdown.
- Leg two. The shares recovered to $29.10 on 2026-03-27, then fell 29.6% to $20.48 on 2026-07-01 on a satellite-competition scare, a downgrade and index exclusion — press and sell-side drivers, not a filed adverse event.
- Now. $22.96 at the 2026-07-23 close: 12.1% above the trough, 22.5% below the peak.
The fear gauge
Volume surged into the low, short of this name's own panic days
2.37x
Peak volume vs pre-fall medianbelow the 4–5x of past panics
85M
Shares a day, July 2026from a ~32M pre-peak median
289
Days from peak to trough
- Capitulation, muted. The heaviest 20-day volume in the fall measured 2.37 times the 180-day pre-peak median — real selling, below the 4–5x days AT&T has printed in earlier panics.
- Mechanical more than emotional. Turnover ran about 85M shares a day in July 2026 against a ~32M pre-peak median, with Russell index exclusion among the named drivers.
- What cannot be measured. FINRA returned no short-interest rows for AT&T in this run, so the short side of the seller mix is unquantified.
Damage math
The price removed $47.8B of market value; the estimates went up
Consensus and price over the trailing 180 days
| Metric | 2026-01-24 | 2026-07-23 | Change |
|---|---|---|---|
| FY2027 EPS | $2.453 | $2.567 | +4.6% |
| FY2028 EPS | $2.575 | $2.906 | +12.8% |
| FY2027 revenue | $129.3B | $132.0B | +2.1% |
| Share price | $23.59 | $22.96 | −2.7% |
- No earnings cut to anchor to. Across the drawdown, consensus FY2027 EPS rose 4.6% and FY2028 rose 12.8%, and AT&T beat consensus EPS in each of the last four quarters.
- What repriced. Market cap fell from $212.6B at the September 2025 peak to $147.0B at the trough and $164.8B now — a $47.8B equity decline against rising forward estimates.
- The limit on that reading. The revision series reaches back only to 2026-01-24, after the peak, so estimate movement across the first leg is not directly observed.
The NPV question
At $22.96 the price capitalizes free cash flow shrinking about 2.4% a year, forever
Implied perpetual FCF growth at each price level
Equity FCF as a growing perpetuity: g = r − FCF ÷ market cap, on $18.0B of guided 2026 FCF and an 8.5% cost of equity.
- The gap. Guidance and consensus put FCF rising from $18B toward $21B by 2028 — the opposite sign to the −2.4% a year the current price capitalizes.
- The permanent case. Justifying the $47.8B equity decline on a flat $18B perpetuity requires sustainable FCF stepping down about $4.1B a year, to roughly $13.9B — 23% below the guide, permanently.
- The ruling. Three blind judges put the probability the damage is temporary at 0.64 — contested, at 0.64, 0.43 and 0.68, with management's own $24.8B and $4.4B wireline write-downs on the permanent side.
Yield
Adjusted free cash flow yields 5.0% against the framework's 10% bar
Free cash flow yield, four ways
Adjusted (5-yr average acquisitions)
5%
Mid-cycle (~$4B spectrum charge)
7.3%
Generous (EchoStar as one-off)
9.1%
Unadjusted reported FCF
10.1%
- What the gap is made of. Stock compensation is trivial at $536M; about 90% of the $7.9B five-year acquisition average is the one-time 2021–22 C-band spectrum super-cycle.
- Which bar applies. Net debt of $117.9B is 2.53x adjusted EBITDA — a moderate balance sheet, so the 10% line governs, not the fortress 8–9% or the levered 25%.
- The reported yield. The 10.1% headline figure ($16.6B ÷ $164.8B) clears the bar only by leaving the spectrum spend out of the calculation.
Durability
The moat is the capital base; the doubt is that the cash holds a band, not a trend
Reported free cash flow
2024 included about $2.3B of since-excluded DirecTV distributions, so the like-for-like 2024 figure is $15.3B.
- Why revenue is the easy half. $131.6B of net plant plus $128.1B of spectrum licenses is a base a competitor cannot cheaply replicate, and the postpaid phone base grew three straight years to 74.2M.
- Why cash is the hard half. Capex runs $20.8B, or 16.6% of revenue, net debt is about $118B, and a ~$23B EchoStar spectrum purchase is pending.
- Share pressure. Comcast and Charter together added about 3.4M wireless lines in 2025 against AT&T's 1.47M, pressing the wireless pricing that funds the enterprise.
Self-help
The share count is falling now, after rising 23% through the acquisition years
Shares outstanding
- The engine restarted. Buybacks ran $4.5B in 2025 after four years near zero, taking the count from 7.587B in 2022 to 7.179B in 2025, a −0.8% five-year rate.
- The counter-fact. The habit is two years old, 2025 repurchases were struck near $27.0 a share against today's $22.96, and the count rose about 23% from 2016 to 2022 on serial acquisitions.
- Dividend, contested. The 4.8% payout is covered 2.0x by FCF and committed through 2028; AT&T also rebased it 47% in 2022, and 2026 dividends plus buybacks are essentially 100% of the FCF guide.
The clock
The path back above the bar runs through 2028, at close to even odds
Consensus free cash flow
S&P Capital IQ consensus; the framework-adjusted yield on this path crosses 10% around 2028.
- The crossing. Adjusted forward yield runs 8.3% in 2026, 8.9% in 2027 and 10.1% in 2028. The jury put the probability of a return above the bar within one to three years at 0.57, spread 0.02.
- The calendar. The EchoStar close lands at the end of July 2026 and steps leverage to about 3.2x; management guides back toward 2.5x within roughly three years, with the copper cost roll-off starting late 2026.
- What breaks the path. A fresh C-band-scale spectrum auction would reset the five-year acquisition average and push the crossing past 2029.
Re-rating math
The target test cannot be computed in this run; the name's own base rates can
Past 30%+ drawdowns and time to reclaim the high
| Episode | Depth | Years to reclaim |
|---|---|---|
| 2016-07 | −37.1% | not reclaimed |
| 2019-11 | −32.4% | not reclaimed |
| 2021-05 | −32.1% | 3.2 |
| 2022-05 | −31.4% | 1.9 |
| 2023-01 | −34.1% | 1.1 |
| 2025-09 (current) | −30.9% | open |
- Why it is blank. The tally records no bar and no normalized adjusted FCF: the adjusted series is not computable here — stock compensation is missing for FY2016–2017 and no complete five-year acquisition window exists.
- What is computable. Closing a 5.0% adjusted yield to the 10% bar is roughly a doubling; the run's own path reaches it around 2028 at 57% probability, contingent on spectrum spend normalizing.
- Both sides of the base rate. The three completed post-2021 drawdowns of this depth reclaimed their highs in 1.1 to 3.2 years on a total-return basis; the 2016 and 2019 episodes never have.
Total-return basis, dividend and spin adjusted, from daily price history through July 2026.
What to watch
A real dislocation in an essential business, on a cash line that has not grown and a gate that does not bend
- 01Q3/Q4 2026 guidance cut: FY FCF below $18B or adjusted EPS below $2.25
- 02Postpaid phone net adds turn negative or churn rises materially
- 03capital allocation pivots to debt paydown over repurchases
- 04share count inflects upward
This distills a fixed fit test built tab by tab — universe, dislocation, damage, yield, durability, self-help and clock — against one investor's stated framework.
Compiled from the full report · 2026-07-24 · For information, not investment advice.