Fit

Does Verizon fit Ruchir's system?

Does not fit the framework (P4b hard fail: rising share count); contested: P1, P2

Verizon is a large, essential, U.S.-listed oligopolist that clears the universe and trips no exclusion, but it does not fit the framework: the capital-allocation pillar (P4b) hard-fails because the diluted share count rose for a decade on stock-based compensation while buybacks ran at zero, and the framework treats a rising-share-count profile as a disqualifier regardless of the rest. Confidence is low: the two model families split on the year-10 gate (P1) and on FCF consistency (P2), and the name-mask probe flipped P1, so a prior_driven_risk flag stands. No exclusion hit; no watchlist_only flag.

Universe and exclusions — unsoftened

Nothing here is soft. Verizon passes every universe test and triggers no exclusion, and each check is cited.

• Geography (U1) — met. Common stock of a Delaware-incorporated, New York-headquartered company, traded directly on the NYSE/Nasdaq under VZ [1]. No ADR or offshore-domicile wrinkle; not a Chinese issuer.

• Market cap (U2) — met. About $196.2B (a $46.38 close on 24 Jul 2026 over 4.231B shares) — roughly 19x the $10B line.

  Source: derived from fit_features.market_cap (price 2026-07-24; shares FY2025).

• Car company (X1) — not triggered. An integrated telecom with two reportable segments (Consumer $106.8B, Business $29.1B) and zero vehicle manufacturing [2].

• Market darling (X4) — not triggered. ~11x earnings, a ~6% dividend yield, flat revenue, and a Hold-heavy sell side — the inverse of the multiple-to-sales darling the exclusion targets. The counter-fact sits in the same breath: a cheap, unloved multiple can equally mark a value trap, which the Yield and Damage Math tabs adjudicate, not this screen.

• China dependence (S1) — not material. 89% U.S. workforce [3]; disclosed translation exposure limited to Pound, Euro, Australian Dollar and Krona, with no renminbi named [4]; no restricted-vendor network equipment. The residual is a small, unquantified international enterprise-services slice.

• Promotional CEO (X2) — not triggered. 2024 guidance was delivered and the fixed-wireless target beaten early; officers and directors hold well under 1% with no open-market buying. The counter-fact: the buyback repeatedly promised at ~2.25x leverage was deferred for years until a new CEO acted in 2026, and the new CEO's register is more promotional — treated in Self-Help.

• Structural decline (X3) — checked, present only in the legacy tail. Business wireline operating income fell 39% since FY2018 and carries a $5.8B accumulated goodwill impairment, and copper carries a lead-cable liability [5] — but it is a shrinking share of a growing whole, so the company-level structural-decline flag is not met.

Pattern match

Of the reader contract's four setups, Verizon is closest to #2 — high dividend yield plus high FCF yield, business not going away (a ~5.9% dividend, a ~9-10% adjusted FCF yield, and an essential connectivity franchise). It is not #4 (a quality tech monopoly on a fear dip) because the drawdown is shallow and un-panicked, and it is not #3 (a healthcare/insurance forecasting error) because there was no one-year guidance cut — guidance was raised twice through the fall. The problem for the #2 read is that the pattern turns on the dividend not being cut, and the dividend is well covered (1.75x), but the entry the pattern also needs — a distressed price forced down by capitulation — is absent (see Dislocation). So the setup is recognizable but incomplete: the income leg holds, the fear leg does not.

The pillar ledger

Reference lines, not grades. Each pillar carries its deciding arithmetic and its strongest surviving counter-fact.

Year-10 durability gate (P1) — contested

The gate holds on structure but the jury split on it. Verizon sits in a three-carrier national oligopoly whose entry is gated by FCC spectrum on 10-year, routinely-renewed terms [6] [7], on top of a $157.0B indefinite-lived license base [8], and revenue rose from $126.0B (FY2016) to $138.2B (FY2025) with zero consecutive-decline years — the three-year high-single-digit-decline disqualifier is not tripped. The strongest counter-fact, in the same treatment: consolidated postpaid net adds collapsed 63% in 2025 (to 861K from 2,355K), and T-Mobile's CFO states Verizon "lost postpaid net accounts and had ARPU declines" [9]. That volume erosion is exactly why the jury contested the gate — Claude seats (a, b) read met, Codex seats (c, d) read not_met; trimmed-mean probability 0.73, spread 0.12, cross-family disagreement. Full treatment in Durability and Business.

FCF consistency (P2) — contested

Reported free cash flow held in an $18.7B–$23.6B band across FY2020–FY2025 except a dated 2022 dip to $14.1B on the C-Band capex peak ($23.1B) that recovered to $20.1B by 2025 [10] [11] [12]. So consistency holds on the filed record. The counter-fact that split the jury: the framework's own adjusted-FCF stability series (FCF minus SBC minus five-year acquisition average, rolling five years) is not_computable because the numeric feed lacks capex, SBC and acquisitions — so two Codex seats could not determine the pillar (a, b met; c, d cannot_determine). Treated in Durability and Yield.

Dislocation and yield (P3) — not met

• Trigger (P3a) — not met. The March–July 2026 fall of 18.3% peak-to-trough carries no dated adverse event: the 8-Ks are a routine segment-reporting change and an immaterial $625M BT joint venture ($625M / ~$196B = 0.3% of cap) [13], and both quarterly prints in the window beat and raised guidance [14]. Counter-fact: the sharpest sessions (29 Jun –5.2%, 30 Jun –4.0%) overlapped the BT announcement, but a $625M item cannot plausibly drive a 9% two-day move.

• Capitulation (P3b) — not met. The volume gauge reads only 1.22x the pre-peak median on a 20-day basis — a firm three weeks, not panic — against a history that holds 6x–20x sessions, none of which fall in this decline. Counter-fact in the same breath: individual trough days did reach ~2.1–2.3x and clustered at the low, the directional signature the framework wants, but 2.3x is not capitulation.

  Source: derived from fit_features.capitulation_gauge.volume_spike.

• Adjusted yield (P3c) — not met. On the most generous (business-acquisition-only) treatment the FY2025 adjusted FCF yield is 9.45% on the $196.2B market cap — ~55 bps under the 10% moderate-balance-sheet bar — and any charge for lumpy spectrum widens the gap to 8.2% (mid-cycle spectrum) or 3.5% (trailing C-Band-inflated acquisition average) [15] [16]. Counter-fact: on Verizon-defined unadjusted FCF the yield is 10.26% today — the shortfall is created by the framework's SBC-and-acquisition adjustment, not by weak cash generation.

Adjusted FCF = reported FCF − after-tax SBC − 5-yr avg acquisition spend; the 10% line is the framework's moderate-balance-sheet bar. Derived from FY2025 10-K cash-flow figures [17].

• Forward path (P3d) — not met. Consensus clears 10% on unadjusted FCF now (10.13% FY2025 → 15.12% FY2029), but on the normalized-adjusted basis the yield reaches 10% only around FY2028, and the crossing depends on a quiet spectrum cycle and an accretive Frontier ($22.3B, closed Jan 2026) [18] [19] — a genuinely balanced call: probability 0.54, spread 0.03, cross-family disagreement (one Claude seat met, three not_met). Full workings in Yield.

Balance sheet and self-help (P4) — the decisive fail

• Outlast capacity (P4a) — not met as a repurchase priority. Verizon can comfortably outlast its problem — the 2026 maturity of $17.3B is covered by $20.1B of annual FCF with a never-drawn $12B revolver behind it [20] — but management still ranks buybacks fourth, behind network, dividend, and deleveraging, with net-unsecured leverage at 2.5x and the target reached only in 2027 [21]. So repurchase headroom is the residual, not the priority.
• Repurchase engine (P4b) — not met. This is the hard fail that decides the framework. Here is the decisive point: the diluted share count rose every year for a decade, from 4,086M (FY2016) to 4,231M (FY2025), on stock-based compensation while buybacks ran at $0 in FY2024 and FY2025 [22] [23]. The framework treats a rising-share-count profile as an outright disqualifier — the buyback flywheel that the whole dislocation case is built on runs backwards. The strongest counter-fact, in the same breath: a $25B program authorized in January 2026 is executing — $3.5B repurchased by mid-year against a full-year target of up to $4.5B [24] [25] — but at ~$4.5B (about 2.3% of the $196.2B cap) it only starts to offset SBC, short of a demonstrated multi-year decline.

Share count rose ~3.5% over the decade; buybacks were zero in FY2024–25. Source: FY2025 10-K income statement [26] and Note 8 [27]; 2026 program per Q2 2026 call [28].

The framework's levered exception cannot rescue it: that path needs a ~25%+ adjusted yield, a demonstrated multi-year share-count reduction, and non-deteriorating FCF — Verizon meets only the third (FCF grew to $20.1B, first-half 2026 up 16%). Two of three legs fail. The absurdity check also does not fire: at the current price it takes ~9.8 years of reported FCF to retire the whole float — a normal high-single-digit multiple, not the ~3-year figure that flags an impossible price.

• Dividend safety (P4c) — met. The ~5.9% dividend was covered 1.75x by FY2025 FCF (a 57% payout) and 2025 was the nineteenth consecutive year of increases, a streak spanning 2008–09 and the pandemic [29] [30]. Counter-fact: growth is glacial (~1.8%/yr) and rests on FCF that had declined ~1%/yr over the prior five years, so the payout is secure rather than compounding. Full treatment in Self-Help.

Diagnosis (P5) — met (temporary lean)

The temporary-versus-permanent question was tried by two opposing cited briefs and three blind judges, who put the probability the impairment is temporary at 0.66 (individual seats 0.54 / 0.67 / 0.66, mean 0.62, spread 0.13, not flagged contested). At the July trough of $41.99 the price matched a fully permanent scenario even though the panel assigns only ~0.34 to permanence, leaving a price-versus-value gap of ~$54B at the trough and ~$36B (~15%) at $46.38 — but that gap has roughly halved as the stock re-rated from $42 to $46, so the current price already pays for close to the standalone-permanent case [31]. Counter-fact: the gap is entirely assumption-driven (r = 9.0%, base FCF $21.5B), and if permanent decline runs faster than −3% it disappears. Full arithmetic in Damage Math.

Instrument context (I1) — not verifiable

Whether qualifying long-dated options exist and at what implied volatility could not be verified from the corpus: the only sources for a January 2028 expiry and a 25.58% 30-day implied vol are web references absent from the corpus (local data carries realized, not implied, volatility). Stated as a framework fact, not advice; see Clock. This is the run's single skeptic-unverifiable item.

What a 3x in three years would require

The framework's target test is re-rating math unavailable because the applicable bar or normalized adjusted FCF is missing — fit_features.adjusted_fcf and adjusted_fcf_yield are not_computable, so the tally records no bar-yield price, no implied market cap, and no upside figure. What can be stated, from the surviving Yield claims, is the reference arithmetic: clearing the 10% moderate bar at today's price needs adjusted FCF of about $19,623M against the current normalized ~$16,042M — roughly an 18% lift — which consensus reaches only around FY2028. And the base-rate context from Clock argues against a fast re-rating: Verizon's own 2019–2023 de-rating fell 50.6% over ~46 months and remains ~25% below its 2019 peak, and its 2000–2008 drawdown took ~18 years to round-trip; the one fast leg on record was +67.5% off the October 2023 low over ~29 months. There is no Meta-style vertical snap-back in this ~0.2-beta name. A 3x in three years is not what the arithmetic or the history supports.

Contested and undetermined

• P1 (year-10 gate) — contested. Claude seats (a, b) read met on structure; Codex seats (c, d) read not_met on the 63% collapse in postpaid net adds. Trimmed-mean probability 0.73, spread 0.12, cross-family disagreement. The name-mask probe read P1 as met, i.e. the mask flipped the gate — the basis of the prior_driven_risk flag.
• P2 (FCF consistency) — contested. Claude seats (a, b) read met on the filed FCF band; Codex seats (c, d) returned cannot_determine because the framework's adjusted-FCF stability series is not_computable from the numeric feed. No probability is recorded.
• Nothing else was undetermined: every other criterion carried a decisive verdict (no criterion resolved to cannot_determine at the tally level).

Provenance

Item	Value
Jury families / seats	Claude (seats a, b) and Codex (seats c, d); mask seat = Claude
Cross-family agreement	Agreed on all criteria except P1 and P2 (both contested)
Order stability	temporary-first mean 0.66, permanent-first mean 0.605, gap 0.055
Name-mask probe	Gate criterion differed on P1; max probability gap 0.0 → prior_driven_risk = true
Skeptic counts	16 fully checked: 15 survived, 0 weakened, 0 refuted, 1 unverifiable (25 triaged-only)

Source: fit_tally.json provenance and refutations.json.

The verdict was pressed hard: fifteen load-bearing claims were independently recomputed and all fifteen survived, none were refuted, and only the options/implied-vol facts (I1) came back unverifiable. The reason confidence is nonetheless low is not weak evidence but disagreement — the two model families genuinely split on the year-10 gate and on FCF consistency, and the name-mask probe flipped P1, so the framework flags that the gate read may be prior-driven rather than fact-driven.

The falsifier ledger

The standing conditions that would change the framework read, carried verbatim from the tally:

1. adjusted FCF or EBITDA declines where flat-or-better was underwritten
2. revenue declines for a third consecutive year
3. capital allocation pivots to debt paydown over repurchases
4. share count inflects upward
5. the industry repricing cycle fails to materialize where industry-wide mean reversion was underwritten
6. Q3/Q4 FY2026 FCF or EPS misses the raised guidance or guidance is walked back toward the original 4-5% EPS / ~7% FCF path, exposing the rebound as lapping/cost-cut/buyback sugar.
7. Consumer postpaid churn re-accelerates above ~95bps or postpaid phone net adds turn negative — the volume-for-price swap failing.
8. Mobility/broadband service revenue fails to reach ~3-4% organic growth in 2H26/2027 once the 180bps lap rolls off — 'volume replaces price' left unproven.
9. Consumer reporting unit moves to a quantitative goodwill test with a narrowing fair-value cushion, or a new Consumer impairment is booked — core wireless earning power actually eroding, not just wireline.
10. Q3 or Q4 FY2026 churn reaccelerates above about 0.95% or postpaid phone net adds turn negative after the outage credit laps.
11. FY2026 adjusted EPS growth is cut below the original 4%-5% guide or free cash flow misses about $21.5 billion.
12. FY2027 organic mobility and broadband service revenue remains below 2% despite normal promotion levels, showing the 180 bp price reset was not offset by volume.
13. A new Business or wireline impairment, or Frontier-related leverage/capex overrun, consumes the verified cash-flow gains.
14. Q3/Q4 FY2026 churn re-accelerates above ~95bps or postpaid phone net adds turn negative, showing the base (not just one quarter's ARPA) was damaged.
15. FY2026 FCF falls short of $21.5B or the two guidance raises reverse, breaking the beat-and-raise cadence.
16. A new Consumer goodwill impairment or a second Business impairment is recorded, or the 2027 deleveraging path stalls.
17. Revenue growth stalls below the 2-3% floor with no offsetting FCF/EPS growth, confirming lost monetizable earning power rather than a mix choice.

Data gaps

• Adjusted FCF (FCF − SBC − 5yr-avg acquisitions), its rolling-5-year stability, adjusted-FCF yield, balance-sheet class and net debt all return not_computable in fit_features because the numeric cash-flow feed omits capex, SBC and acquisition lines; every adjusted figure here was rebuilt from the filed 10-K cash-flow statements and should be reconciled into the feature file.
• The structured feed's net-debt reading of −$0.4B captures only the $18,618M current-portion debt; true total debt is $158,150M and net debt $139,102M per the FY2025 debt note — the feature would otherwise mis-select the fortress bar.
• International/China revenue is not separately quantified in the FY2025 10-K; S1 immateriality rests on the 89% U.S. workforce, disclosed FX exposures, and the no-restricted-vendor statement rather than a percentage.
• Reported short interest is unavailable (data/short_interest empty), so any crowded-short unwind through the fall cannot be quantified; the corpus also lacks a benchmark/sector price series, so the rate-driven mechanism of the fall is inferred from the dividend-yield move.
• No dated source for long-dated (18-month) implied volatility; the 25.58% figure is the 30-day mean IV (2026-07-24), and the pre-run web-research pass was unavailable, so options/IV facts (I1) could not be verified against the corpus.
• Post-Frontier (closed 20 Jan 2026) pro-forma net debt, EBITDA and adjusted FCF are not in the filed FY2025 statements; the ~3.0x leverage and integration-capex estimates are directional.
• Exact credit-rating letters could not be confirmed; the filing states ratings were unchanged in 2025 and 2024 and Verizon remains investment-grade [32].