Fit

Fit

Does not fit the framework (P1 not met)

The year-10 durability gate is not met, four votes to nil, at a trimmed-mean probability of 0.395 and a spread of 0.13. Confidence is high on the tally's own basis — two model families agreed, the trial was order-stable, and load-bearing spreads were at most 0.15. No exclusion hit, no China sensitivity flag, no name-mask divergence on a gate criterion, and the watchlist-only overlay is not set. One criterion, P2, is recorded cannot-determine.

The order of operations is what produced that answer: P1 not_met -> does_not_fit (gate; nothing offsets it). Six other criteria are met, including the diagnosis that the 2024-25 damage was mostly temporary. None of them can offset a gate, and the rest of this tab reports them anyway, because a reader is owed the full ledger and not just the binding line.

Universe and exclusions

The screen is clean on every check, and it is clean by a wide margin rather than narrowly.

U1, listing — met, four votes to nil. Devon Energy Corporation is a Delaware registrant, commission file 001-32318, whose common stock of $0.10 par value trades on the New York Stock Exchange under DVN [1]. The Coterra merger that closed on 7 May 2026 changed the asset base and the share count but left the listed security, the file number and the ticker unchanged. The Chinese-ADR branch of the test is not engaged.

U2, scale — met, four votes to nil, on either share count. The primary record shows 1,153,403,107 shares outstanding as of the 18 May 2026 record date [2]; at the $44.405 close of 29 July 2026 that is $51.2 billion, 5.1 times the $10 billion line. The deterministic feature file reaches $28.1 billion instead, because it multiplies the same price by the pre-merger FY2025 count of 633 million — 2.8 times the line. Devon's own June 2026 statement that an $8 billion repurchase authorisation represents "15% of our market value" implies roughly $53.3 billion [3]. Every basis clears. The feature file's staleness is reported in full under Data gaps because it contaminates yields, not the size screen.

Sources: derived — 1,153,403,107 shares outstanding at the 18 May 2026 record date [4] and 633 million FY2025 shares from the deterministic feature file, both at the 29 July 2026 close of $44.405; the third bar solves $8bn / 0.15 from management's own framing [5].

X1, car companies — not hit. Devon aggregates its US operating segments into one reporting segment and describes itself as engaged primarily in exploration, development and production of oil, natural gas and NGLs, with activities "solely focused in the U.S." [6]. All $16,786 million of FY2025 revenue from contracts with customers is oil, gas, NGL and marketing and midstream [7]. There is no manufacturing line. The counter-fact belongs in the same breath: the substantive shape the exclusion was written against — a price-taking, capital-heavy business that screens cheap on free-cash-flow yield — is present here, and it is carried into the year-10 gate below rather than smuggled into an exclusion whose written test it does not meet.

X2, promotional management — not hit; the test needs both prongs and only one is partly evidenced. Prong one, promise versus delivery, is mixed and is stated plainly: the variable dividend, described in May 2023 as paying as much as half of excess cash flow, was dropped on the Q3 2024 call and has not returned [8], and the $2.5 billion debt-reduction programme announced in August 2024 [9] has retired under $500 million of gross debt through March 2026 with the numeric target absent from the last two calls. Against that, 2025 capital came in $408 million below the low end of the initial $4.0 to $4.2 billion 2025 guide [10], and the $1.0 billion optimisation target was declared achieved on the Q1 2026 call, ahead of its end-2026 deadline [11]. Prong two fails outright: the chairman held 2,408,753 shares and the chief executive 941,724 at the 18 May 2026 record date [12], against guidelines requiring six times base salary for the chief executive and three times for other named officers [13]. The counter-fact: directors and executive officers as a group hold 0.46% of the shares outstanding, and no insider has bought on the open market since 4 March 2024 — including through the $26.80 low.

X3, structural decline — not hit. Revenue from contracts with customers ran $15,140 million, $15,919 million and $16,786 million across FY2023 to FY2025 [14]. The longest consecutive decline run in the visible history is two years, FY2019 and FY2020; the current streak is zero; the three-consecutive-year high-single-digit test is false. The trial's temporary probability of 0.71 sits far above the strongly-permanent trigger. The counter-fact: the FY2023 fall of 23.6% happened while production rose, so reported revenue is a weak proxy for structural health in this business in either direction.

X4, market darling — not hit. Devon carries a forward price-to-earnings ratio of 8.65, second-lowest in a seven-name US independent peer set and 21% below that set's median, and a price-to-sales ratio of about 3.3 times, fourth of seven — and the sales multiple is overstated, because trailing revenue holds only about two months of Coterra. The ten-year price chart is a series of round trips, not a compounding line. The counter-fact in the same breath: sell-side positioning is one-sided, with roughly 21 buy and 4 outperform ratings against 2 holds and no sells and a mean target near $59 against the $44.41 close, so the informational edge is thinner than the multiple suggests. The multiple and coverage figures come from dated web lookups on 2026-08-02, not from a filing page, and are recorded as such.

S1, China dependence — not flagged. China revenue is $0 of $16,786 million, China assets $0 of $31,599 million, and China employees zero of roughly 2,200 [15] [16]. The counter-fact: Devon's own forward-looking-statement language names China explicitly as a source of oil, gas and NGL price volatility through trade policy and tariffs [17], so the price Devon realises carries China exposure even though its revenue and asset base do not.

One labelling note, because it looks like disagreement and is not. The tally records no cross-family agreement on X1, X2, X3, X4 and S1: one seat from the second model family wrote "not hit" where the other three seats wrote "not met". The evidence and the finding are identical on both sides, and the name-masked seat matched.

Pattern match

Devon fits none of the reader contract's four setups. That is framing rather than verdict; it does not move a single criterion.

Cyclicals at the bottom. The shape is right and the rule is not: the pattern is written for large banks and explicitly excludes the adjacent cases, and this is a commodity producer whose output price is set outside the firm. The entry moment is also gone. The drawdown ran 43.0% from $47.03 on 31 July 2024 to $26.80 on 8 April 2025 and has since repaired to within 5.6% of the old peak.

High dividend yield plus high free-cash-flow yield. The dividend does not carry the case. At $1.28 a year on the post-merger rate against the $44.405 close, the yield is 2.88%, below the roughly 4% level at which the framework treats the dividend as a material part of the return — the rate is the $0.24 quarterly fixed dividend of 2025 [18] raised 33% within weeks of the merger closing [19], against a disclosed framework that sends roughly 70% of returns to dividends and buybacks [20]. The tally records P4c as not applicable for exactly that reason.

Healthcare and insurance forecasting errors. There is no repricing mechanism of the pattern's kind. An insurer refiles rates and the book readjusts on a regulated calendar; Devon's revenue reprices continuously with the commodity, with no contract-renewal, premium-reset or regulated-rate step to correct an industry-wide forecasting error. Devon also never cut its own numbers during the fall — it raised production guidance twice and lowered capital guidance three times.

Quality tech monopolies and duopolies on a fear dip. The market structure is the opposite of the pattern's. Devon is one of twenty operators charted in its own core basin and holds roughly 9% of that group's inventory locations, and it states plainly that certain competitors "have resources substantially greater than ours" [21].

The full anatomy of the drawdown sits in Dislocation and the recovery calendar in Clock.

The pillar ledger

Source: the run's deterministic fit tally, as recorded. Probabilities and spreads are shown only where the tally records them.

Year-10 gate (P1) — not met

Four votes to nil, at 0.395 trimmed mean across seat probabilities of 0.34, 0.36, 0.43 and 0.47 — a spread of 0.13. The name-masked seat, shown the same dockets with the company's name removed, returned 0.40 and the same verdict.

Here is the decisive point. The gate does not ask whether Devon is a good company or a cheap one; it asks whether year-10 revenue and adjusted free cash flow will both be higher than today's, with very high conviction, and it resolves any proper doubt downward. Four of the framework's five conviction sources return the wrong answer on the primary record, and the fifth cuts both ways.

• Price is set outside the firm. As of January 2026 Devon's oil was sold 71% short-term variable and 29% long-term variable — 100% variable, nothing fixed — and the filing states that "the vast majority of our production is sold at variable, or market-sensitive, prices" [22].
• Regulation is a cost, not an entry barrier. Devon does not expect compliance costs to affect its operations "materially differently than other similarly situated companies", while noting those costs have increased over the years and will likely continue to increase [23]. A cost floor that rises for everyone is not the licensing regime the framework's regulated-entry source describes.
• Capital intensity maintains a depleting base rather than fencing anyone out. Rigs, crews, materials and services are bid for in open markets against competitors with "resources substantially greater than ours" [24]. Capital expenditure plus acquisitions ran 25.6%, 56.0% and 23.3% of revenue across FY2023 to FY2025 [25].
• The reserve base is consumed inside the horizon. Proved reserves of 2,428 MMBoe at 31 December 2025 against 307 MMBoe produced that year is 7.9 years of life, and the ratio has sat between 6.2 and 8.1 years for the whole visible history [26]. Everything sold in year 10 has to be found or bought first.

Source: derived — year-end proved reserves divided by that year's production, from the FY2025 supplemental oil and gas disclosures [27] and the FY2021 [28] and FY2023 [29] equivalents for the earlier years.

The fifth source, long operating history, is present and is not decisive on its own. Devon was founded in 1971 and has been publicly held since 1988 [30] — 55 years, above the framework's 30-to-50-year band — but the asset base behind that history rotates continuously: Barnett out in 2020, Grayson Mill in for roughly $5.0 billion in 2024, Coterra in May 2026, and a portfolio review opened in June 2026.

The strongest surviving counter-fact, in the same treatment. Devon has replaced what it produces without buying it. Organic additions — extensions, discoveries and all revisions — totalled 2,083 MMBoe against 1,490 MMBoe produced over FY2019 to FY2025, or 140% — read across the FY2021 [31] and FY2023 [32] reserve rollforwards for the earlier years — reaching 188% in FY2025 alone [33]. The combined company claims more than ten years of highly competitive Delaware inventory at the current pace of development, on 746,000 net acres producing 863,000 Boe/d [34], and 2026 guidance of 1.355 to 1.405 MMBoe/d is roughly 65% above the 836 MBoe/d Devon produced standalone in 2025 [35]. Volume is not the doubt. Price is: the jury's deciding sensitivity, carried from Durability, holds 2025 volumes flat and reprices them at Devon's 2020 realisations, which cuts upstream sales 39.5%.

Two honesty notes the skeptic pass forces. The claim behind the price-taker finding was weakened: Devon's production guidance, its 100%-variable oil contract mix and the absence of any customer above 10% of sales all verified, but the 3.7% US, 0.5% world and 5.8% Permian share percentages could not be checked against the corpus and are not relied on here. The year-10 gate claim itself was also weakened, on a citation rather than on arithmetic — the transcript page behind management's inventory qualifier did not resolve, so that qualifier is dropped and only the February 2026 disclosure is used. Neither weakening touched the reserve-life, contract-mix or regulation arithmetic, which is what the four seats voted on. Full treatment in Durability and Business.

Consistency (P2) — cannot determine

Three seats recorded cannot-determine and one recorded not met; the name-masked seat recorded met, which is the widest divergence anywhere in the run and is why nothing here is promoted. The framework's measure is a rolling five-year average of adjusted free cash flow, and that series does not exist for Devon in this run. The named missing datapoints are set out under Contested and undetermined.

What can be computed is reported free cash flow — operating cash flow less capital expenditure — from the filed statements, and on that basis the rolling five-year average has risen through each of the only three windows available.

Source: derived — operating cash flow less capital expenditures from the filed Consolidated Statements of Cash Flows, FY2019 to FY2020 [36], FY2021 to FY2022 [37] and FY2023 to FY2025 [38]. These are reported, not adjusted, figures.

The three windows average $2,964 million with a coefficient of variation of 15.5%, and no year in the covered period is negative — including 2020, when Devon still generated $311 million at a WTI index of $39.59. The counter-fact sits in the same numbers. Single-year reported free cash flow ranged from $133 million in FY2019 to $5,988 million in FY2022, a 45-fold spread inside four years, and all three rolling windows share the FY2022 peak, so they are not three independent observations. On the three adjusted years that can be built, the series moves the other way: $2,027 million, $1,560 million, $1,661 million. Full working in Yield.

Dislocation and yield (P3a, P3b, P3c, P3d)

P3a — met, four votes to nil. A dated adverse event exists. Devon's own filing names the trigger: commodity prices in 2025 fell "driven primarily by economic uncertainty in global trade arising from geopolitical events and shifting trade policies, such as the imposition of tariffs by the U.S. and planned oil output increases by OPEC+" [39]. The stock fell 29.3% across four sessions from 2 to 8 April 2025 on 2.85 times median volume. The counter-fact: 45% of the peak-to-trough dollar loss was already embedded before 2 April, so the named event is the sharp end of a slide that began without one — and the whole episode has since closed, with the shares at $44.41 against the $47.03 peak.

P3b — not met, three votes to one. The capitulation gauge reads 1.92 times the trailing median, below the 2 times reference line, and it is located in the December 2024 leg rather than at the April 2025 low, where the twenty-day average reached only 1.45 times. The heaviest single session of the entire fall closed up 0.8%, and Devon traded more heavily on merger news in March 2026, at 2.87 times, than in any month of the drawdown. The dissenting seat's reasoning is the counter-fact and is a fair one: read daily rather than smoothed, the four April sessions averaged 2.85 times and peaked at 3.45 times, a compressed panic a twenty-day window cannot register.

P3c — not met, four votes to nil, and this is the arithmetic the framework cares most about. The balance sheet selects the bar: net debt of $7,005 million against EBITDA of $7,516 million is 0.93 times [40] [41], which is moderate, so the reference line is 10% rather than the 8-to-9% fortress line. Adjusted free cash flow for FY2025 is reported free cash flow of $3,119 million less $99 million of share-based compensation less the $1,359 million five-year average of cash acquisitions, or $1,661 million [42]. On the feature file's $28,108 million market capitalisation that is 5.91%, 409 basis points short of the line; the three-year average of 6.22% is 378 basis points short. Correcting the denominator to the post-merger $51,217 million widens the miss rather than closing it.

The counter-fact, in the same treatment. Devon's reported free cash flow of $3,119 million — the $3.1 billion the company itself markets [43] — is 11.1% on the same market capitalisation, above the bar. The adjustment, not the operations, is what puts the name below the line, and the whole of the adjustment's step-down traces to the $3,808 million Grayson Mill purchase entering the five-year window.

P3d — met, four votes to nil, at 0.615 with a spread of 0.01 — the tightest agreement in the run. Consensus forward free cash flow clears the bar on the corrected market capitalisation: 10% of $51,217 million is $5,122 million, and FY2027 consensus free cash flow of $7,417 million less $200 million of assumed combined share-based compensation less the $1,359 million trailing acquisition charge is $5,858 million, an 11.44% yield with a $736 million cushion. The counter-fact is the width of that assumption, not the commodity: scale the acquisition charge to combined revenue instead of carrying Devon's trailing dollar figure and the same consensus gives 9.88%, below the bar; double the trailing charge and it gives 8.78%. FY2027 consensus revenue was also cut 7.9% and earnings per share 4.4% in the thirty days before the quote date.

Source: derived — adjusted free cash flow is reported free cash flow less share-based compensation less the five-year average of cash acquisitions, built from the filed Consolidated Statements of Cash Flows [44]; consensus free cash flow is the vendor mean at the 30 July 2026 vintage. The two denominators are labelled on each bar and are not interchangeable.

Balance sheet and self-help (P4a, P4b, P4c)

P4a — met, four votes to nil. Devon can outlast a long weak-price stretch without being forced anywhere. The single material covenant caps total funded debt at 65% of total capitalisation; the actual ratio was 24.8% at 31 December 2025, and the $3.0 billion revolver was undrawn and extended to March 2030 [45]. Liquidity at the end of 2025 was $4.4 billion, $1.4 billion of cash plus the undrawn facility [46]. No maturity year through 2030 exceeds $1.0 billion on Devon's own schedule, and $1.21 billion once the legacy Coterra series are added [47]. Roughly 30% of returns go to the balance sheet by disclosed policy [48], and management describes the split as optimisable quarter by quarter — "different quarters can present different opportunities that we want to be nimble around" [49]. The counter-fact: the combined maturity schedule is this run's construction, not a filed statement — the first post-merger consolidated balance sheet arrives with Q2 2026 results on 4 August 2026, and it excludes roughly $2.6 billion of cash paid for Delaware acreage in May 2026.

P4b — not met, four votes to nil, and it is the framework's stated hard fail rather than a shortfall against a reference line. Period-end shares went from 382 million at year-end 2020 [50] to 1,153,403,107 at the 18 May 2026 record date [51], a rise of 202% driven by three equity-funded acquisitions: WPX in 2021, 37.3 million shares for Grayson Mill in 2024 [52], and Coterra in 2026. Share-based compensation is not the cause: it ran $99 million against $6,711 million of operating cash flow [53].

Sources: Consolidated Statements of Equity, FY2018 to FY2020 [54] and FY2022 to FY2025 [55]; the 2026 bar is the 1,153,403,107 shares outstanding at the 18 May 2026 record date [56].

The counter-fact, in the same treatment, and it is a real one. Between deals the buyback demonstrably shrinks the count. Devon retired 99,798 thousand shares for $4,393 million at an average of $44.02 under the $5.0 billion programme, taking period-end shares from 663 million at year-end 2021 to 622 million at year-end 2025, a 6.2% reduction [57], and the heaviest year by share count was 2025, at $34.07. Management has since upsized the authorisation to $8 billion and said on the Q1 2026 call that it was "positioned to increase repurchases activity beyond our legacy level, and capitalize on any discount to our intrinsic and relative value" [58]. The framework's answer is nevertheless mechanical: the mergers were all-stock, the count is rising, and a rising count from serial acquisition is a stated fail regardless of what happens between deals. The disclosed base repurchase of $1.0 to $1.5 billion a year is 2.0% to 2.9% of the corrected market capitalisation, roughly a quarter to a third of the flywheel the framework assumes at a 10% adjusted yield. Full treatment in Self-Help.

P4c — not applicable. The criterion activates on a dividend yield near 4% or a case that leans on the dividend; Devon's is 2.88%. Recorded regardless: cover is 4.3 to 5.0 times on consensus free cash flow, the fixed layer was paid straight through 2020's $2,671 million loss year, and the variable layer has not been paid since the third quarter of 2024, taking total dividends paid from $1,858 million in 2023 to $619 million in 2025 [59].

Diagnosis (P5) — met

Four votes to nil at 0.71, carried from the adversarial trial and not re-derived by the jury. Three blind judges reading opposing cited briefs returned 0.71, 0.68 and 0.78 for a temporary rather than permanent impairment; the mean is 0.723, the spread 0.10, and the reading-order gap 0.02 — temporary-first seats averaged 0.71 and permanent-first seats 0.73, which is what "order-stable" means here.

The arithmetic behind it survived the skeptic pass intact. At the 8 April 2025 trough the equity carried $12.66 billion of damage — 626 million shares at the peak times the $20.23 fall. The problem itself, taken from Devon's own net-earnings bridge, was an approximately $1.3 billion pre-tax realised-price effect [60], or $1,033 million after tax. Discounted at 10%, two years of that is $1.79 billion and a perpetuity of it is $10.33 billion; probability-weighted at 0.71 temporary, $4.27 billion. The gap at the trough was $8.39 billion, 66% of the fall. Devon's own filed 10%-discounted standardised measure of proved reserves fell only $1.0 billion, from $19,770 million to $18,765 million, over the same year [61].

The counter-facts sit inside the same treatment, and one of them is why the setup is gone even though the diagnosis holds. The skeptic weakened the current-price claim and its weakened form is what stands: at the $44.405 close the remaining damage against the July 2024 peak is $1.64 billion, below even the two-year temporary figure, so no positive gap survives at this price on any of the three scenarios — and on the last full session close of $42.66 the residual is $2.74 billion, leaving a $0.94 billion gap against the temporary case only. Separately, a per-unit margin gap survives at flat benchmark prices: Q1 2026 field-level cash margin was $27.78 per Boe against $30.16 a year earlier [62] even though WTI rose 1% and Henry Hub rose 38%, because gas realised 33% of Henry Hub against 70% a year earlier and NGLs 25% of WTI against 31% [63]. On the annual series the same measure fell from $29.63 per Boe in 2024 to $24.97 in 2025 [64]. Whether that gap closes is the first falsifier in the ledger below. Both cases are argued in full in Damage Math.

Instrument context (I1) — not verifiable

Stated as fact, not as advice, and recorded as not verifiable by all four seats because no options or volatility source sits in the document corpus. Listed DVN options exist beyond twelve months: from a Cboe delayed-quote file timestamped 1 August 2026, the 21 January 2028 expiry (17.6 months out) carried 20,715 contracts of open interest and the 15 December 2028 expiry (28.4 months) 8,595 — together 5.7% of the 517,330 contracts outstanding across all expiries. At-the-money implied volatility read 0.376 and 0.383 at those two expiries, below the framework's roughly 0.50 reference level; AlphaQuery's 180-day implied-volatility mean was 0.384 as of 31 July 2026. Bid-ask spreads and executable size at the 2028 expiries were not obtainable from any free source. Because none of this resolves to a filing page, the tally records the criterion not verifiable rather than met, and it never blocks or unblocks a pillar.

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

The tally computes no re-rating arithmetic for this run. Its own note, recorded verbatim: "Re-rating math unavailable because the applicable bar or normalized adjusted FCF is missing." The applicable bar is undetermined in the feature file, because balance_sheet_class returns unknown for a missing FY2025 EBITDA, and normalised adjusted free cash flow is not_computable for the reasons set out below. No substitute figure is offered in the tally's place.

What the surviving ledger does carry, labelled as consensus arithmetic rather than as the framework's target test: 10% of the corrected $51,217 million market capitalisation is $5,122 million a year of adjusted free cash flow, and FY2027 consensus reaches $5,858 million on Devon's trailing acquisition charge or $5,061 million on a revenue-scaled one — clearing the bar on the first convention and missing it on the second. The float retires in 6.9 years of FY2027 consensus free cash flow, against a framework reference of roughly three years, and in 34 to 51 years at the disclosed base repurchase rate.

The base rates come from Devon's own price history rather than from a projection. Across fifteen completed drawdowns of 35% or deeper since 1990, the median depth was 47.2% and the median time from peak to trough 4.9 months; thirteen of the fifteen regained the prior peak, at a median of 29.3 months, and only four of those thirteen did it inside eighteen months. Two never recovered at all.

Source: derived — a 30% reversal zigzag over 9,211 daily closes from 2 January 1990 to 29 July 2026, as computed for Clock; the price series excludes dividends, so total-return drawdowns are shallower than shown.

Against that, every published self-help mechanism lands inside eighteen months: a roughly $200 million synergy exit rate in December 2026 and a roughly $980 million run rate at the end of 2027 [65], on a management team that reached its standalone $1.0 billion optimisation target [66] ahead of its own end-2026 deadline [67]. None of that changes the gate, and none of it is a recommendation; it is the calendar the arithmetic would have to run on.

Contested and undetermined

Nothing was recorded as contested. The tally's contested list is empty. No criterion carries two live readings that the framework declines to resolve.

One criterion is undetermined: P2, adjusted free-cash-flow consistency. Three seats returned cannot-determine and one returned not met. Each seat that could not determine named the datapoint it lacked, and the three formulations are reproduced exactly as recorded:

1. rolling five-year average of adjusted FCF (fit_features.fcf_stability empty: SBC missing FY2016-17, no complete consecutive five-year acquisition window)
2. complete rolling five-year adjusted FCF stability series with SBC and acquisition adjustments
3. rolling 5-year adjusted FCF stability series with complete SBC and acquisition inputs

They are the same absence under three descriptions. The deterministic derivation needs a consecutive five-year window with both share-based compensation and cash acquisitions present, and the structured cash-flow feed carries no acquisitions line at all and no share-based compensation before FY2018, so no window closes. Reported free cash flow can substitute for direction but not for the framework's measure, which is why three seats declined to call it rather than reading the proxy as an answer.

The name-masked seat recorded P2 as met, the one place in the run where masking changed a label. It did not touch a gate criterion, the tally records no gate difference under masking, and the maximum probability gap between the masked and unmasked reads is 0.035 — which is why prior_driven_risk is not set.

Two further criteria are absent for structural reasons rather than uncertainty. P4c is not applicable, because the dividend yield is 2.88% and the case does not lean on it. I1 is not verifiable, because no option-chain or implied-volatility source sits in the corpus. Both are stated above.

Provenance

Source: the run's deterministic fit tally, trial tally and refutation ledger, as recorded.

Two sentences on what that means for a reader who will not open the machinery. The verdict was reached by four jurors reading the same evidence dockets independently — two from each of two model families — plus a fifth juror shown the same dockets with the company's name stripped out, and the masked juror reached the same gate conclusion at almost the same probability, which is why the run carries no prior-driven-risk flag. Every claim the verdict turns on was handed to a skeptic that recomputed the arithmetic from the cited pages: four claims came back weakened and are reported here in their weakened form, one could not be verified at all and carries no weight, and none was refuted.

The falsifier ledger

These are the standing conditions that would change the read, reproduced exactly as recorded. Several are the same test nominated by more than one tab; the duplication is the ledger's, not this tab's, and nothing has been merged away.

Framework templates. Five generic conditions the system applies to every name.

Condition (as recorded)	Threshold and direction	Window
adjusted FCF or EBITDA declines where flat-or-better was underwritten	Any sustained decline in adjusted FCF or EBITDA; cuts against	Not defined in the ledger
revenue declines for a third consecutive year	Three consecutive declining years; cuts against. Current streak is zero, longest run two years	Rolling, annual
capital allocation pivots to debt paydown over repurchases	Debt paydown displacing repurchases when the discount is widest; cuts against. Policy is roughly 30% to the balance sheet today	Rolling, quarterly
share count inflects upward	Any further rise; cuts against. Already triggered — plus 202% since year-end 2020	Rolling
the industry repricing cycle fails to materialize where industry-wide mean reversion was underwritten	Absence of the assumed sector-wide reversion; cuts against	Not defined in the ledger

Name-specific, with thresholds and windows as recorded.

Condition (as recorded)	Threshold and direction	Window
FY2026/FY2027 total field-level cash margin fails to recover toward the 2024 $29.63/Boe level even at WTI above $70, i.e. the Q1 2026 $27.78/Boe gap persists - confirming per-unit margin loss rather than price [68] [69]	$29.63/Boe reference, WTI above $70; cuts against temporary	FY2026 and FY2027
Permian gas and NGL realizations stay near 33% of Henry Hub and 25% of WTI through 2027 rather than normalizing as takeaway is added - making the residual-stream discount structural, not transient [70]	33% of Henry Hub, 25% of WTI; cuts against temporary	Through 2027
YE2026 proved reserves fall on non-price revisions, or the standardized measure drops materially at a flat SEC price deck - the one outcome that converts price mechanics into genuine asset impairment [71]	Non-price reserve revisions negative at a flat price deck; cuts against temporary	Year-end 2026
2027 exit run-rate synergies land far below the ~$980m target, or the $8bn repurchase goes largely undrawn - showing the post-Coterra per-share recovery depended on a promise rather than delivered cash [72] [73]	$980m run rate, $8bn authorisation; cuts against	Year-end 2027
FY2026 or FY2027 proved reserves and standardized measure fall materially on non-price revisions under a flat SEC price deck.	Non-price revisions at a flat deck; cuts against temporary	FY2026 and FY2027
Total field-level cash margin stays below $25/Boe, or Delaware stays below $26/Boe, despite WTI averaging at least $65.	$25/Boe company, $26/Boe Delaware, WTI at least $65; cuts against	Not defined in the ledger
The verified $1B optimization and Coterra synergy targets fail to convert into at least $4B annual free cash flow at mid-$60s WTI.	$4bn annual free cash flow at mid-$60s WTI; cuts against	Not defined in the ledger
Net leverage rises above 1.5x while dividend/buyback commitments are cut or deferred through year-end 2027.	1.5x net leverage; cuts against	Through year-end 2027
YE2026 or YE2027 standardized measure falls materially at a flat or higher SEC price deck, or proved reserves decline on non-price revisions - which would convert the FY2025 dip from price mechanics into genuine asset decay.	Standardized measure at a flat or higher deck; cuts against temporary	Year-end 2026 and 2027
Field-level cash margin fails to recover toward the $29.63/Boe 2024 level through FY2027 even with WTI near $70, confirming that the Permian gas (33% of Henry Hub) and NGL (25% of WTI) realization gap is a permanent structural discount rather than a transient basis problem.	$29.63/Boe, WTI near $70; cuts against temporary	Through FY2027
FY2027 free cash flow lands far below the ~$7.4bn consensus and Q4-2027 run-rate synergies fall well short of ~$980m, so that post-merger cash generation depends on issuing equity rather than on the assets - and the $8bn buyback goes largely undrawn.	$7.4bn consensus free cash flow, $980m synergies; cuts against	FY2027, fourth quarter
Devon adds further debt or equity to sustain volumes (another acquisition financed like Grayson Mill) while per-share FCF stagnates, showing that maintaining production now permanently requires outside capital.	Any further debt- or equity-financed acquisition alongside flat per-share FCF; cuts against	Not defined in the ledger

Source: the run's falsifier ledger, reproduced verbatim; the threshold, direction and window columns parse each condition and add nothing to it. The four entries whose ledger text carried raw internal source references are shown with those references resolved to their filing pages.

The one condition already met is the fourth template: the share count has inflected upward and stayed there, which is the mechanism behind P4b.

Data gaps

The tally records forty-eight data-gap entries, most of them the same limitation reported independently by several tabs. Deduplicated, they are these.

The feature file's market capitalisation is wrong by a factor of 1.82, and it propagates. fit_features.market_cap multiplies the 29 July 2026 close of $44.405 by 633,000,000 shares taken from the FY2025 income statement — a count that predates the 7 May 2026 Coterra merger. The filed count is 1,153,403,107 as of the 18 May 2026 record date, giving $51,217 million rather than $28,108 million. Every yield the file divides by market capitalisation is overstated by the same factor: consensus_forward_yield reads 22.35% for FY2026 against 12.27% corrected, and 26.39% for FY2027 against 14.48%. The universe screen is unaffected on either basis. Both figures are stated wherever they appear on this tab; neither was silently substituted.

The framework's core yield arithmetic could not be recomputed by the deterministic derivation. fit_features.adjusted_fcf, adjusted_fcf_yield, yield_baseline, fcf_stability and float_retirement_years all return not_computable. The structured cash-flow feed carries a free-cash-flow field only for FY2016 and FY2017, no share-based compensation before FY2018 and no acquisitions line at all, so no complete consecutive five-year window exists; that feed also reports FY2017 free cash flow of $2,865 million against $2,909 million of operating cash flow, implying $44 million of capital expenditure, which is not credible and was not used. Every adjusted figure on this tab was rebuilt from the filed Consolidated Statements of Cash Flows in the FY2021, FY2023 and FY2025 Forms 10-K and is labelled as a derivation. Because the corpus holds no Form 10-K covering fiscal 2017 or 2018, the five-year acquisition average is computable only from FY2023 onward, so the yield baseline rests on three years rather than the five to seven the brief asks for. fit_features.balance_sheet_class returns unknown because FY2025 EBITDA is missing from the feed; it was computed here from the filed income statement at 0.93 times and cross-checked against the vendor actual at 0.95 times, both moderate. fit_features.share_count_trend uses weighted-average diluted shares rather than period-end shares and carries no 2026 observation, so it understates both the 2025 reduction and the entire Coterra dilution; this tab shows the period-end series. fit_features.capitulation_gauge measures the drawdown from a 31 July 2024 peak of $47.03 rather than the 7 June 2022 cycle peak of $78.04, so it reads minus 43.0% where the longer window reads minus 65.7%; both are stated.

No post-merger financial statements exist anywhere in the corpus. The latest Form 10-Q covers Q1 FY2026, filed before the 7 May 2026 close, and the latest Form 10-K is FY2025 standalone Devon; the first combined quarter reports on 4 August 2026. The FY2025 pro forma combined statement of operations is referenced in the 7 May 2026 Form 8-K but the exhibit is not indexed. Coterra's standalone financial statements, share-based compensation, acquisition history, reserves and inventory disclosures are absent entirely, so combined revenue, EBITDA, free cash flow, reserve life, organic replacement and any pro forma valuation multiple could not be read from a filing page. Combined debt, cash and liquidity here are assembled from the FY2025 footnote plus the 25 June 2026 Coterra note exchange, and exclude 2026 repayments, the roughly $2.6 billion May 2026 acreage payment and any legacy Coterra revolver. No corpus document states the size of the combined revolving facility or any post-merger covenant. The combined-company share-compensation deduction of $200 million a year is an assumption, not a sourced figure; it moves the FY2027 adjusted yield by roughly 20 basis points either way.

Consensus and market data are thin where the drawdown happened. The run's web-research provider returned HTTP 402 on every phase, so current market capitalisation, peer multiples, sell-side rating counts and production denominators come from direct dated lookups on 2026-08-02 and are recorded in the manifest rather than cited to a page. The CapIQ revision feed carries only trailing-180-day snapshots from 30 January 2026 onward, so no consensus revision path exists across the 2024-25 trigger window at all — the near-term hit is reconstructed from Devon's own dated guidance decks and the FY2025 net-earnings bridge instead. The 2026 snapshots straddle the merger re-basing, which moves FY2027 consensus revenue from $15.8 billion to $26.6 billion, leaving only the 29 June to 29 July 2026 pair directly comparable. The Visible Alpha driver set holds a single 2026-07-29 vintage with no pre-drawdown snapshot. No adjusted-free-cash-flow consensus is published, so the vendor's operating-cash-flow-less-capex mean is used as the proxy and named as such, and the commodity deck underlying it is undisclosed. No peer or sector price series is staged, so the claim that April 2025 repriced the whole industry rests on Devon's filings and contemporaneous reporting rather than measured relative performance. The 29 July 2026 bar used as "current" carries 200 shares of volume and a 4.09% move; the last full session, 28 July 2026, closed at $42.66.

Ownership, flow and instrument data. The short-interest dataset returned zero rows for reported positions, short-sale volume, public net-short disclosures, borrow pressure and peer context alike, so the short-interest figures used are third-party compilations of semi-monthly FINRA reports with approximate settlement dates. The insider record has a February-to-April 2026 gap between two feeds. The disposition of the 37.3 million shares issued to the Grayson Mill seller on 27 September 2024 is not disclosed, so whether that block became supply during the fall cannot be established. Long-dated option open interest and implied volatility come from Cboe and AlphaQuery files dated 1 and 31 July 2026 and carry no filing page; bid-ask spreads and executable size at the 2028 expiries were unobtainable, as Barchart, OptionCharts and MarketChameleon all block automated access.

Market structure and history could not be measured from primary sources. No named market-share statistics for US onshore exploration and production exist anywhere in the corpus; concentration is inferred from Devon's own 2026 estimated Lower-48 peer chart, which names nineteen listed operators and excludes ExxonMobil, Chevron and every private operator, combined with external production aggregates. Devon publishes no location-level inventory count with break-even bands after 2023, so the ten-year inventory assertion could not be reconciled to a well count at a stated oil price. Revenue history before FY2016 exists in neither the feature file nor the corpus, so a 55-year operating history could be charted for ten years only. The corpus holds no FY2018 to FY2020 annual reports, so the $2,956 million of FY2018 repurchases in the numeric feed has no page anchor and the executed-buyback record is fully cited only from FY2019. The corpus holds no forward commodity strip or independent mid-cycle price deck, so the temporary-versus-permanent split rests on the trial's ruling and realised-price history rather than a market-implied curve. Devon's standalone oil sensitivity was read off a machine-extracted bar chart in the 5 May 2026 deck rather than printed text.

Calendar items that are not yet knowable. No company-confirmed report dates exist for the third or fourth quarters of 2026; dates used elsewhere in the report are projected from the 2025 cadence and labelled indicative. The portfolio review announced at merger close has no published timetable — the June 2026 update commits only to updates "at the appropriate time" — so divestiture proceeds, their size, and whether they go to repurchases or debt paydown can be neither dated nor sized.

One retrieval limitation worth naming. The PageIndex substring engine returned zero hits on every query issued for the pillar work, so discovery relied on the semantic engine plus direct reads of the index page arrays, and convergence between the two retrieval engines could not be used as a materiality signal.