Fit

The answer

Does not fit the framework (P1 not met); contested: P2.

General Motors is a car company, and the framework excludes car companies by name (X1). Underneath that exclusion sits the reason it matters: the pure year-10 durability gate (P1) does not hold, and a gate is a gate — nothing offsets it. Confidence is medium: probability spreads were at most 0.25 or one non-load-bearing criterion was contested. One criterion is contested — P2, free-cash-flow consistency — and China carries a sensitivity flag (S1). The name is not routed to watchlist-only; the instrument context (I1) is a fact GM meets, not a reason to exclude.

Here is the decisive point. Every valuation and self-help pillar GM is asked to clear, it clears — a double-digit forward cash yield, a net-cash industrial balance sheet, a share count down 38% in a decade. It fails anyway, because the framework's one binary gate asks whether year-10 revenue and free cash flow will be higher with very high conviction, and for a no-moat, capital-intensive cyclical mid-transition that conviction is not available.

Universe and exclusions — unsoftened

X1 — car company: a direct hit. GM is an auto manufacturer (SEC SIC 3711), the single industry the framework names as an exclusion. Its own 10-K supplies the confirming evidence: GM "operate[s] in a highly competitive industry that has historically had excess manufacturing capacity" [1]. The counter-fact in the same breath: GM is the US share leader at 17.2% and holds roughly a third of the US full-size truck pool, its most profitable franchise [2]. But leadership is precisely what the exclusion discounts: in an oversupplied industry, being first among near-equals does not confer pricing power. This is the "cheap car company" the framework has paid for before.

S1 — China: a sensitivity, not an exclusion. GM's China business runs through equity-method JVs, so it is off-balance-sheet operating exposure, not a Chinese listing — the Chinese-ADR exclusion does not apply. The exposure is real and deteriorating: JV volume fell to 1.88m units (7.1% share, from 8.4% in 2023), and 2025 carried further restructuring in a "market with significant excess capacity" [3].

The other exclusion screens are clean. GM is not a promotional-CEO case (X2): management broke two expensive strategic promises — the EV-profitability ramp and Cruise robotaxi — but delivered its core-auto guidance and named the problems before analysts did, so the pattern is mixed execution, not quarterly-EPS management. It is not a mechanical structural-decline case (X3): revenue shows a single −2.1% year in FY2025, not three consecutive high-single-digit declines. And it is not a consensus darling (X4): GM trades near 0.5x sales and ~6x forward earnings — the exclusion fails in the cheap direction, not the expensive one.

Universe — passed. Geography: a US-primary NYSE listing, Delaware-incorporated, ticker GM — not a foreign or Chinese ADR [4]. Market cap: ~$87.9B (973M shares × $90.30 on 2026-07-28), roughly 8.8x the framework's $10B floor.

Market cap is derived: 973M diluted shares (FY2025) × $90.30 close (2026-07-28); share count from company filings [5].

Pattern match

GM fits none of the framework's four recognition setups cleanly. It is not a large bank at a cyclical bottom (pattern 1 is banks only). It is not a high-dividend-plus-high-FCF-yield case — the dividend yields under 0.8% and no part of the return leans on it (pattern 2). It is not a healthcare/insurance forecasting error (pattern 3). And it is not a quality tech monopoly on a fear dip (pattern 4).

The one genuine resemblance is structural, not categorical: the 2025 dislocation was a whole-industry tariff shock, the kind of industry-wide repricing the framework prizes because mean reversion is structural rather than company-specific. That shape is what makes the dislocation and yield pillars pass on arithmetic. It does not rescue the fit, because the car-company exclusion and the year-10 gate govern, and both cut the other way.

The pillar ledger

Source: deterministic fit tally; verdicts and aggregates as recorded — reference lines, not scores.

Year-10 durability — the gate that fails (P1: not met; probability 0.495, spread 0.10)

The gate asks whether year-10 revenue and free cash flow will be higher than today, with very high conviction. Revenue higher in a decade is plausible — GM grew from $109B (2020) to $168B (2025) and gains US share. Adjusted free cash flow higher with very high conviction is not available, because the variables that decide it sit largely outside GM's control and currently cut against it: EV-transition timing (GM took $7.9B of GMNA realignment charges in 2025 after adoption came in slower than planned), a recurring tariff drag of $3.1 billion of EBIT-adjusted in 2025 with $3–4B estimated for 2026 [6], a China equity engine that swung from roughly +$1B to multi-billion losses with the JV agreement expiring in 2027, and a moat confined to one truck franchise. The four jurors agreed (cross-family) at a mean probability of 0.495 — a coin toss where the gate demands near-certainty. Any proper doubt resolves the binary gate to does not fit. This is the reference line GM cannot reach, and it decides the report.

The strongest surviving counter-fact: GM has produced positive free cash flow every year from 2016 through 2025, is retiring float aggressively, and guides to 2026 net income of $10.3–11.7 billion and EBIT-adjusted of $13.0–15.0 billion — normalized earning power well above the charge-depressed 2025 result [7]. Earning power is not the question; ten-year conviction is, and it is not there. Full treatment on the Durability tab.

Consistency — contested (P2)

Reported free cash flow was positive in all ten years FY2016–FY2025, ranging $6.5B to $17.6B, with the rolling five-year average rising from ~$8.5B to ~$10.3B — volatile year to year but not unpredictable [8]. On that reading, the consistency bar is met. The contest is over the basis: the framework's yield is adjusted FCF (after SBC and acquisitions), and the deterministic stability feature returned not_computable because the pipeline flagged share-based compensation as missing for every year. SBC is in fact disclosed in the statements of equity ($531M/$543M/$253M for 2025/2024/2023) [9], so the series can be reconstructed — but the source-of-record feature does not compute it. The jury split cleanly across families on exactly this point (see Contested and undetermined below). Full treatment on the Yield and Durability tabs.

Dislocation and yield — the pillars GM passes (P3a met; P3b not met; P3c met; P3d met, probability 0.855, spread 0.01)

The dislocation was real and dated: GM fell 29.4% from a $60.20 close (25 Nov 2024) to $42.48 (8 Apr 2025), triggered by the 25% imported-vehicle tariff, which GM matched with an EBIT-adjusted guidance cut to $10.0–12.5B from $13.7–15.7B [10]. So P3a is met. But the fear gauge is not: traded volume through the fall spiked only ~1.6x the trailing median — orderly repricing, not the 60–70% forced-selling capitulation the framework requires at a peak-fear moment. So P3b is not met, and the entry trigger is no longer live: at $90.30 GM sits ~50% above the pre-fall peak, the drawdown fully reversed.

On valuation the arithmetic passes. GM's adjusted-FCF yield clears the 10% default bar on every reasonable basis — ~13.5% on three-year-average adjusted FCF, ~12.1% on FY2025 automotive FCF, and ~11.0% on the most conservative anchor, consensus forward FCF — so P3c is met. The counter-fact travels with it: the ~19% headline yield on reported FCF overstates the case, because 2025's $17.6B was lifted by a ~$9.1B working-capital release and blends in GM Financial's cash flows [11]. Forward, consensus FCF clears the bar every year FY2025–FY2028 (~11.0% rising to ~13.6%), so no mean-reversion underwrite is needed; the jury put this at probability 0.855 with a spread of just 0.01. The framework's own caution applies squarely here: a car company screening cheap on FCF yield is its canonical value trap, so a passing yield is the signature it distrusts, not a fit. Full treatment on the Dislocation and Yield tabs.

Source: consensus FCF (CapIQ) on the $87.9B market cap; the 10% line is the framework's default balance-sheet bar. Derived from company filings and consensus estimates [12].

Balance sheet and self-help — passed (P4a met; P4b met; P4c not applicable)

On an industrial basis GM is net cash: automotive liquidity of ~$35.7B against ~$16.2B of automotive debt, with only $663M due in 2026 — capital allocation is not forced toward debt paydown, so P4a is met [13]. The counter-fact sits alongside it: consolidated debt is ~$130B, of which $114B is GM Financial's captive-finance book, market-funded and matched against ~$123B of finance receivables and leases — non-recourse to automotive, but it does make a consolidated leverage figure hard to read [14].

The buyback engine is executed, not promised: GM spent $11.1B/$7.1B/$6.0B on repurchases in 2023–2025 and retired roughly 35% of its shares since late 2023 at a blended ~$49, well below today's $90.30; the diluted count fell from 1,570M (2016) to 973M (2025), a 38% reduction [15]. The framework's hard-fail — a rising share count from SBC or serial M&A — is absent, so P4b is met; the only offset is that GM paused repurchases in Q1 2025 when tariffs hit, confirming buybacks yield to liquidity under stress [16]. The dividend is immaterial (P4c not applicable): under 0.8% yield, covered many times, and no part of the return leans on it. Full treatment on the Self-Help tab.

Diagnosis — a moderate lean, not a settled call (P5: met, probability 0.63, spread 0.22)

The blind adversarial trial set the probability that the impairment is temporary at 0.63, with a per-judge range of 0.44–0.66 and a spread of 0.22 — a moderate lean toward temporary, reflecting a discrete, fading EV/tariff hit, not a consensus. The dissent matters: one permanent-first judge sat at 0.44, weighing the permanent China and Cruise write-offs, and the report cannot override the ruling. The price-vs-value gap the framework hunts existed only briefly at the April 2025 trough and has since closed and reversed — market cap is ~$88B, roughly $20B above the pre-event peak. Full treatment on the Damage Math tab.

Instrument context — a fact GM meets (I1: not verifiable in-corpus)

Long-dated listed options exist on GM (LEAPS to January 2028) and 30-day implied volatility was ~38% on 2026-07-28, below the framework's ~50–55 reference line — so on instrument grounds the name would not be routed to watchlist-only. The verdict is recorded as not verifiable because these facts are web-sourced; the corpus and local structured feeds carry no citable option expiries, open interest, or implied-volatility series. Stated as a framework fact, never as advice. Full treatment on the Clock tab.

What a 3x-in-3-years would require

The framework's target test — the price at the bar yield on normalized adjusted FCF, and what consensus would have to concede — does not compute here. Re-rating math is unavailable because the applicable bar or normalized adjusted FCF is missing: the deterministic adjusted-FCF series is not_computable (the SBC pipeline gap above), so no normalized adjusted-FCF anchor or implied market cap at the bar can be set without substituting an improvised figure.

What can be said in arithmetic points the other way from a fresh entry. The yield already clears the 10% bar (~11.0–13.6% forward), and the buyback flywheel is live, so the cash-return mechanics are present — but the stock trades at an all-time high with the dislocation fully reversed, so there is no fear discount to re-rate off. For base-rate context, the Clock tab records four GM drawdowns of 35–64% since 2011, each fully recovered, with trough-to-recovery running 10–23 months (centered near 18) and full round trips of 2.9–3.8 years. Those episodes describe what a future trough entry might return; none is available today, and the year-10 gate closes the case regardless.

Contested and undetermined

Contested — P2 (FCF consistency). Both readings are on the table, and the jury split across model families. The two Claude-family seats read P2 as met on reported FCF (positive every year FY2016–FY2025, rising five-year average). The two Codex-family seats read it as cannot determine, because the deterministic adjusted-FCF stability feature is not_computable — adjusted FCF lacks the pipeline SBC needed for a rolling five-year adjusted series. The masked probe resolved this to "met." The disagreement is exactly cross-family (a=met, b=met / c=cannot_determine, d=cannot_determine), which is why the tally records P2 as contested rather than resolved. The honest statement: on reported FCF the consistency bar is met; on the framework's own adjusted basis it cannot be confirmed from the source of record.

Undetermined — I1 (instrument context). Recorded as not verifiable: options and implied-volatility facts exist only in web data, with no corpus or structured-feed source to pin option expiries, open interest, or the exact IV figure. No criterion resolved to cannot determine at the criterion level.

Provenance

Dimension	Result
Jury families	Claude, Codex (two seats each: a, b = Claude; c, d = Codex; masked = Claude)
Cross-family agreement	Every gate criterion agreed across families; P2 is the sole cross-family split
Order stability	temporary-first mean 0.63, permanent-first mean 0.55, gap 0.08
Name-mask probe	max probability gap 0.005; no gate criterion changed under masking (prior_driven_risk: false)
Skeptic checks	22 fully checked — 15 survived, 5 weakened, 1 refuted, 1 unverifiable (plus 22 cheap-triaged)

Source: deterministic fit tally and skeptic ledger.

The verdict was pressed hard and did not move. Two independent model families ran the checklist blind; they agreed on the P1 gate and every other gate criterion, disagreeing only on how to classify P2's not_computable feature — and a name-masked re-run shifted no gate and moved probabilities by at most half a point, so the result is not an artifact of GM's identity. One skeptic check was refuted outright: a claim that the 2025 EV charge was largely non-cash failed against the Q2 FY2026 10-Q, which shows the $7.9B charge was $3.2B non-cash and $4.7B cash-related — that finding is excluded from the support here.

The falsifier ledger

These are the standing what-would-change-this conditions. The first five are the framework's own templates; the rest are the name-specific thresholds and windows from the diagnosis.

• adjusted FCF or EBITDA declines where flat-or-better was underwritten
• revenue declines for a third consecutive year
• capital allocation pivots to debt paydown over repurchases
• share count inflects upward
• the industry repricing cycle fails to materialize where industry-wide mean reversion was underwritten
• FY2026 EBIT-adjusted lands below ~$13B or auto FCF below ~$9B, showing the core earning-power level itself, not just optics, has stepped down.
• Material new EV/regulatory charges land in 2026 beyond 'substantially completed' (the 10-K already flags 'additional charges'), proving the write-off was a floor not a ceiling.
• GMNA EBIT-adjusted margin stays structurally sub-7% rather than recovering toward 8-9%, making the tariff/mix compression permanent.
• China keeps bleeding with further restructuring and no recovery in equity income above ~$1B.
• FY2026 and FY2027 adjusted EBIT and adjusted automotive free cash flow meet or exceed guidance after special items, with EV charges ending and no new capacity or supply-contract write-downs.
• Annual tariff EBIT impact falls below $1B or is fully offset without price, mix, volume, or footprint impairment.
• China JVs return to sustained profitability and market-share recovery with no further restructuring or impairment charges.
• GM discloses profitable EV, personal autonomy, or software growth sufficient to replace the abandoned EV capacity and Cruise robotaxi option value.
• FY2026 actuals miss the $13-15B EBIT-adjusted / $9-11B auto FCF guidance materially (results by Jan 2027), showing the rebound was cosmetic.
• North America EBIT-adjusted margin compresses structurally and durably below ~6%, signaling the truck/SUV franchise itself — not just EVs — is eroding.
• New EV or restructuring charges recur beyond the 'substantially completed' figure into 2027, proving the write-off was a floor not a ceiling.
• Net tariff EBIT impact re-expands past $5B on a structural basis despite the IEEPA reversal and refunds.

Data gaps

What the run could not answer:

• Adjusted-FCF series (the framework's yield basis). fit_features.adjusted_fcf and fcf_stability returned not_computable because the pipeline flagged SBC as missing for FY2016–FY2025. SBC is in fact disclosed in the statements of equity and was reconstructed for the reported-basis reads; the feature should not be treated as a genuine absence of the data. Acquisitions are correctly zero (GM discloses no business-acquisition cash line).
• Balance-sheet class. fit_features.balance_sheet_class returned unknown (FY2025 net-debt/EBITDA unresolved); established directly from Note 13 — automotive net cash on an industrial basis, with GM Financial's ~$114B treated as matched-funded captive debt.
• Live consensus positioning. A dated forward-P/E, rating distribution, and price-target refresh could not be pulled from the pipeline; figures rest on a single July-2026 WebSearch snapshot, not a corpus source.
• Fear-gauge and flow detail. Reported short-interest is unavailable (feed status "unavailable," zero history rows), and no holder-by-holder or forced-seller evidence surfaced for the drawdown window.
• Instrument facts. Precise current option open interest/liquidity was not verifiable from a citable dated source; the IV level relies on a single third-party source (AlphaQuery, 2026-07-28) without primary or two-source corroboration.
• Forward path and stress record. Consensus FY2029 FCF is not disclosed by the vendor, truncating the forward-yield path at FY2028; and the current entity (incorporated 2009) has no free-cash-flow record through a credit-driven recession.