Fit

The answer

Does not fit the framework (P1 not met); contested: X1, X2, X3, X4, S1

P1 is the gate: year-10 revenue and year-10 adjusted free cash flow higher, with very high conviction. All four jury seats returned not met, at a 0.66 probability with a 0.10 spread, and nothing offsets a gate. Confidence is low — the tally's basis is that name-mask divergence or load-bearing probability divergence exceeded 0.20. No exclusion hit and no sensitivity flag were recorded; the watchlist-only overlay does not apply; prior-driven risk is flagged.

Source: the run's deterministic fit tally — trimmed-mean jury probabilities for P1 and P3d, and the adversarial trial's carried probability for P5.

KKR is a durable, growing, cash-generative franchise that has fallen 49.8% from peak to trough. The framework still says no, and it says no on durability conviction rather than on business quality: the measure the system uses to underwrite year-10 cash flow cannot be computed for this company at all, and the conviction sources that would substitute for it grade out as partial or absent. Everything below is the arithmetic behind that.

Universe and exclusions

Both universe tests are met, on three seats each with no dissent.

U1 — listing. KKR is a Delaware corporation whose common stock trades on the New York Stock Exchange under the symbol KKR, Commission File Number 001-34820 [1]. Not an ADR, not a foreign private issuer, not a Chinese issuer. The only counter-facts are historical: the listed predecessor traded on Euronext Amsterdam until the July 2010 US listing, and the firm was a Delaware limited partnership until 1 July 2018. Neither bears on the instrument screened.

U2 — scale. 891,550,894 shares outstanding at 24 February 2026 [2] at the 28 July 2026 close of $102.66 gives $91.5 billion; independent market data for the same session gives $92.18 billion. Either figure is 9.2 times the $10 billion floor. The counter-fact sits inside the same finding: fit_features.market_cap is null with the reason "no positive annual period-end or outstanding share count", so the number is reconstructed from the filed share count and the price feed rather than read off the feature file. The 10-K's own cover cross-check — $91.1 billion of common stock held by non-affiliates at 30 June 2025 — corroborates it, and a 9.2x margin absorbs any share-basis difference.

Exclusions: no hit recorded. The tally's exclusion_hits array is empty and sensitivity_flags is empty. What was checked, and what each check found:

• X1, auto OEM. KKR manufactures nothing and is classified under SIC 6282, Investment Advice. Against that: the Strategic Holdings segment held ownership stakes in 19 operating companies at 31 December 2025 in a segment the filing says is not limited to acquiring companies in specific industries [3], so an automotive holding is not structurally precluded — but the segment is 3.9% of FY2025 segment earnings [4][5] and the word "automotive" appears once in the 333-page 10-K, in a director biography [6].
• X2, promotional CEO. The test needs both prongs cited. The ownership prong fails outright: directors and executive officers hold 206,873,438 shares, 23.20% of the common stock, worth $21.2 billion at $102.66 [7], and five insiders including both co-CEOs bought $50.9 million personally between 9 February and 4 March 2026. The delivery prong is genuinely mixed — two of the three April 2024 Investor Day per-share targets have moved, the total-operating-earnings target retired in November 2025 and the $7-plus adjusted-net-income target walked down on 5 May 2026 [8] — but one prong is expressly not a hit. Full treatment in Self-Help.
• X3, structural decline. revenue_trajectory.three_year_hsd_decline is false and consecutive_decline_years is zero. Extending the same fee line through FY2025 from the filings gives a 15.6% compound rate over FY2016–FY2025 with two down years, deepest −2.8%, never consecutive. Against that: GAAP total revenues fell 11.0% in FY2025, from $21,878.7 million to $19,464.3 million, as insurance net premiums dropped from $7,898.8 million to $3,397.2 million — one year, on a line that has swung between $4.2 billion and $21.9 billion in six years. Full treatment in Durability.
• X4, consensus-saturated story. The test is conjunctive: extreme multiple-to-sales and a saturated narrative. The valuation prong fails — 3.6x trailing sales against a 6.2x peer median and 35.7x trailing earnings against a 45x peer median, after a 49.8% drawdown. The narrative prong does not: 18 of 21 published ratings are buy or strong buy, none are sell, and the $123.48 mean target sits 20.3% above the price, with no target below it. Full treatment in Business and Clock.
• S1, China dependence (a sensitivity flag, never disqualifying). China investments were under 2% of total firm AUM at 31 December 2023, the only precise disclosure in the corpus [9], with Greater China at 14% of Asia-Pacific capital invested at fair value against Japan's 39%. Greater China headcount is 47 of roughly 900 Asia-Pacific employees, down from 64. The counter-fact is the vintage: that percentage is 31 months old and has not been restated in the FY2025 10-K or any later deck, and Asia-Pacific as a whole is 23.0% of asset-management revenue.

All five of these criteria are marked contested by the tally, for a reason that is about vote labels rather than about the evidence. That is set out in Contested and undetermined below.

Pattern match

KKR fits none of the four setups in the reader's pattern library cleanly.

The closest is the fourth — a quality franchise on a fear dip. The shape matches: a dated, sector-wide repricing on 3 February 2026, when the stock fell 9.69% on 4.02x median volume with no KKR filing or release on the tape; a 49.8% peak-to-trough decline from $167.07 on 31 January 2025 to $83.88 on 12 March 2026; and a fear that is specific and testable, applied to alternative managers through the marks on their software holdings. The pattern's own check is where it parts company. That precedent rests on monopoly or duopoly market structure. Alternative asset management is not one: the six largest listed managers held at least $5.08 trillion of AUM at 31 December 2025, KKR's $743.9 billion is 14.6% of that top-six pool and about 5% of the roughly $15 trillion industry pool KKR itself sizes, and KKR's own Item 1 calls the industry "intensely competitive", names no competitor and discloses no market share [10]. The trigger is also weaker than the pattern wants: the skeptic pass weakened that finding because no corpus document names a single KKR-specific headline that started the 3 February move. Full treatment in Dislocation.

The second pattern — high dividend yield plus high FCF yield — fails on both numbers: a 0.76% dividend yield and a 2.88% framework-basis adjusted yield. The first, cyclicals at the bottom, applies to large banks and does not reach here. The third, an insurance forecasting error, is the interesting near-miss: KKR owns Global Atlantic and Global Atlantic's spread economics did compress, but the market did not anchor the stock one-for-one to a guidance cut. Delivered adjusted net income per adjusted share rose from $4.70 to $4.87 across the drawdown, and the entire 38.6% peak-to-current move is multiple compression, from 35.5x trailing to 16.8x forward. Full treatment in Damage Math.

The pillar ledger

Source: the run's deterministic fit tally — per-criterion verdicts, trimmed-mean probabilities, spreads and vote splits, reproduced without adjustment.

Year-10 durability — the gate

P1: not met. Probability 0.66, spread 0.10, all four seats agreeing, the name-masked seat agreeing. The revenue leg is not the problem — year-10 revenue higher is a high-confidence call. The cash-flow leg is. The framework's adjusted FCF is free cash flow minus stock-based compensation minus the trailing five-year average of acquisition spend, and it is not_computable for KKR in every year: stock-based compensation is missing for FY2016–FY2019, and there is no complete consecutive five-year acquisition window with SBC. GAAP free cash flow cannot stand in, because KKR's consolidated operating cash flow of $477.8 million in FY2025 is struck after $(42,904.1) million of investment purchases and $33,698.2 million of proceeds belonging to its funds [11].

With the measure unavailable, the gate falls to the five conviction sources, and they grade out as follows. Market structure: a six-firm oligopoly, KKR at 14.6% of the top-six AUM pool, the largest member at 25.6% — concentrated, not monopoly or duopoly. Regulatory entry barriers: real, but they sit in insurance, which produced $1,109.4 million of $4,985.8 million of total operating earnings, 22.2% [12]; the 77% that comes from asset management rests on Investment Advisers Act registration, a compliance regime rather than a licence a regulator withholds. Capital intensity: absent from the fee engine — FY2025 purchases of fixed assets were $160.8 million [13] against management fees of $4,100.8 million [14], 3.9%. Operating history: 50 years at the firm level, but Global Atlantic has been consolidated only since February 2021 and Strategic Holdings has been a reported segment only since the first quarter of 2024, and together they generated 60% of FY2025 revenue.

Here is the decisive point, and its counter-fact belongs in the same breath. The strongest year-10 evidence KKR has is contractual, not competitive: approximately 92% of AUM has a duration of at least eight years at inception or is perpetual [15], which on $743,858 million of AUM is $684.4 billion locked for eight years or more, with a further $118,433 million of uncalled commitments not yet paying fees [16]. Management fees rose from $1,248.5 million in FY2019 to $4,100.8 million in FY2025 and fee-related earnings from $1,080.3 million to $3,714.3 million with no down year. An eight-year lock is a floor under the first eight years of a ten-year question; year nine and year ten depend on re-upping in an industry the company itself calls intensely competitive. That is the gap between "very likely" and "very high conviction", and the gate resolves genuine doubt to not met by construction. The tally records the applied rule as "P1 not_met -> does_not_fit (gate; nothing offsets it)", and it is the only rule the combining logic needed to reach the verdict — though P4b's rising share count is separately marked a hard fail in the spec and would have carried the same result on its own. Full treatment in Durability.

FCF consistency

P2: cannot be determined. All four seats, unanimously. fit_features.fcf_stability.rolling_5y_avg is an empty array, cv_of_rolling_avg is null, and adjusted FCF is null in every year. Only four FCF observations exist in the feature file at all — FY2016 through FY2019, every one of them negative — and they are negative for the same structural reason: consolidated-fund deployment runs through operating cash flow. Reported free cash flow swings from negative $7.28 billion in FY2021 to positive $6.51 billion in FY2024 to positive $0.32 billion in FY2025 purely on that mechanism.

The nearest defensible substitute points the other way, and it should be stated as plainly as the gap. KKR's own after-tax distributable earnings, renamed Adjusted Net Income from FY2024, run $1,613.1m, $1,768.8m, $3,916.1m, $3,512.3m, $3,040.1m, $4,202.3m and $4,377.5m for FY2019–FY2025 — no negative year, and rolling five-year averages rising monotonically from $2,770.1m to $3,287.9m to $3,809.6m. The counter-facts: the measure is management-defined, it is struck before the framework's stock-based-compensation deduction of $722.1 million [17] and before acquisition spend, and the underlying annual series still fell 22.4% from FY2021 to FY2023 as realizations slowed. A substitute is not the test, and the tally records the test as unrun rather than passed.

Dislocation and yield

P3a, identifiable event: met. Four seats. The adverse window is dated by the 3 February to 12 March 2026 price and volume break — 3 February 2026 closed at $103.28, down 9.69% on 11,502,705 shares, 4.02x the 2,863,900-share pre-peak median — and by contemporaneous peer discussion of AI and software pressure. The skeptic weakened the claim to exactly that formulation: the corpus does not identify a single KKR-specific headline that started the move, and KKR's own guidance walk-down came later, on 5 May 2026.

P3b, capitulation: met. Four seats. The measured volume spike is 3.93x the trailing median against the framework's 2x reference line — the highest 20-session average volume inside the fall leg, 11,243,613 shares in the window ending 3 March 2026, against a 2,863,900-share median for the 180 days before the peak. The counter-fact is duration: that window ends nine days before the closing low, monthly volume fell back to 1.66x by April and 1.14x by July 2026, and a second 16.0% decline from 22 April to 29 June 2026 came on 1.6x volume — drift, not a second capitulation. Full treatment in Dislocation.

P3c, yield versus the bar: not met. Four seats. Net recourse debt of $4,445.4 million at 0.75x Adjusted EBITDA (1.41x including Global Atlantic's notes) places KKR in the moderate class, which selects the 10% reference line rather than the 8–9% fortress line or the 25% levered line.

Source: derived from the filed segment measures — Adjusted Net Income less equity-based compensation less the five-year average of acquisition spend, over market value at the 28 July 2026 close; components from the FY2025 10-K [18] and the Q1 FY2026 10-Q [19].

The arithmetic: FY2025 Adjusted Net Income of $4,377.5 million less equity-based compensation of $722.1 million less a five-year average acquisition spend of $986.6 million gives $2,668.8 million, which over a $92,544.1 million market value is 2.88%. On the twelve months to 31 March 2026 the same construction gives 3.13%; on the three-year average, 2.58%. The shortfalls against the 10% line are 712, 687 and 742 basis points. Normalizing the cyclical block — realized carried interest and realized investment income, whose five-year average sits $522.2 million above the FY2025 figure — lifts the yield only to a range of 2.96% to 3.33%, because the recurring block is at a record rather than depressed. Nor is this the fortress signature of a yield that has suddenly jumped: KKR's own six-year framework-basis history runs 4.06%, 5.01%, 5.68%, 2.70%, 1.89%, 2.33%, a 3.38% median, and the current 3.13% is 0.93x that median against a 2x jump threshold.

The counter-fact is that the denominator, not the numerator, is what makes the yield small. Q1 2026 fee-related earnings of $1,016.4 million were up 23.6% on Q1 2025's $822.6 million and total operating earnings up 19.1% [20]. The yield is low because the price is high relative to earnings, not because earnings fell. Full treatment in Yield.

P3d, forward path: not met. Probability 0.125, spread 0.12, four seats. Consensus normalized EPS of $6.12 for FY2026, $7.35 for FY2027 and $8.60 for FY2028 implies framework-basis yields at $102.66 of 4.11%, 5.31% and 6.53% after deducting $1.900 per share of equity-based compensation and acquisition spend. Consensus does not clear the bar on any forecast year available, so the framework's consensus-backed route is closed and the case would have to underwrite mean reversion explicitly. The counter-fact is a price statement rather than an earnings one: at the 12 March 2026 trough of $83.88, unadjusted FY2028 consensus EPS of $8.60 was a 10.25% yield, and the whole gap between that and 6.53% is the $1.90 of deductions the framework makes and the street does not, plus the 22.4% recovery since the trough.

Balance sheet and self-help

P4a, balance sheet: met. Four seats. Nothing in KKR's $13,145.0 million of corporate debt principal matures inside twelve months, 78.3% falls due after five years [21], and cash and short-term investments of $4,789.7 million [22] plus $4,241.6 million of undrawn revolvers [23] give $9,031.3 million of liquidity. Year-one interest of $673.6 million is covered 5.5 times by fee-related earnings alone. The counter-fact is what the firm did with the drawdown year: it raised capital rather than returning it, issuing $2,543.4 million of Series D mandatory convertible preferred in March 2025 [24] and carrying $10,482.2 million of unfunded commitments to its own funds [25].

P4b, repurchase engine: not met, and this is the criterion the spec marks a hard fail rather than a scaled shortfall. Basic weighted-average shares rose from 545,096,999 in 2019 [26] to 890,342,060 in 2025, 8.5% a year; diluted from 557,687,512 to 955,756,926, 9.4% a year [27]. Executed repurchases across the eight years 2018–2025 total $1,401 million of cash, 1.5% of today's market value, or 0.19% a year, with FY2024 at nil and FY2025 at $3.4 million [28]. The absurdity check the framework applies runs the wrong way: at $102.66, retiring the entire share count takes 25.0 years of adjusted cash earnings ($91,527 million over $3,660.5 million), 20.9 years on unadjusted Adjusted Net Income, and 96 years at the fastest repurchase pace KKR has ever run — against a roughly three-year reference for a price that cannot survive. The levered exception is unavailable on all three of its legs: at 1.40x net debt to Adjusted EBITDA on the corporate-debt perimeter, or 1.41x on the carrying-value basis used in the yield computation, the balance sheet is moderate rather than levered; the implied adjusted yield is about 4.0% against the 25% the exception requires, and the share count is rising rather than falling.

The counter-fact is a genuine change of behaviour, not a rounding error. In Q1 2026 KKR repurchased 2,173,970 shares against 578 equity awards retired — the first quarter in the record where open-market buying dominates award retirement — and had spent $317 million through 1 May 2026 [29] at an average of about $91, some 11.4% below the 28 July close. The board added a $500 million automatic increment, leaving approximately $122 million remaining at 1 May 2026 [30]. Even so, weighted-average basic shares still rose 0.33% year-over-year in Q1 2026, and $122 million of remaining authorization is 0.13% of market value against a $500 million top-up worth 0.55%. Full treatment in Self-Help.

P4c, dividend cover: not applicable. Four seats. At $0.78 a share against the $102.66 close the yield is 0.76%, so the dividend carries no weight in the return case either way. The coverage arithmetic is recorded regardless: $649.9 million paid in FY2025 [31] is 17.8% of adjusted cash earnings, on a policy raised in each of the last five years, from the $0.62 a share announced in February 2022 [32] to $0.78 for 2026 [33].

Diagnosis

P5: met, at a probability of 0.76 with a spread of 0.10, carried from the adversarial trial and not re-elicited by the jury. Three order-randomised judges returned 0.76, 0.66 and 0.76; the mean is 0.727, the temporary-first seat averaged 0.76 and the permanent-first seats 0.71, an order gap of 0.05, and the trial records the ruling as not contested. The deciding evidence was the FY2025 segment table, where recurring operating earnings rose while the entire deceleration sat in the investing line: fee-related earnings $3,267,796 thousand to $3,714,313 thousand, up 13.7%, and total operating earnings $4,358,553 thousand to $4,985,804 thousand, up 14.4% [34].

Two facts belong beside that ruling. First, the permanent side's best-verified evidence — Global Atlantic's admitted return-on-equity compression, in a segment worth 22% of operating earnings — is why one seat recorded the probability stopping well short of 0.9. Second, the trial's own quote-check found failures on both sides: the permanent brief's Level III exhibit verified on figures but not on its attributed quote and omitted offsetting context on the same page, its rebuttal exhibit was discredited at two seats because pre-tax income rose 21.1% to $7,099,160 thousand on flat tax expense while net income fell on noncontrolling-interest attribution [35], and the temporary brief's forward-monetization quotes were cited to page 3 when they sit on page 4 [36]. Full treatment in Damage Math.

A temporary diagnosis does not offset the gate. It is the reason this name is worth understanding rather than dismissing, and it changes nothing about P1.

Instrument context

I1: not verifiable. Four seats. The tab-level finding is that listed KKR options extended to 21 January 2028 (541 days, 17.8 months) and 17 December 2027 at the 28 July 2026 close, with 30-day mean implied volatility of 41.5% — inside the framework's up-to-50-to-55 reference band and short of the 18-month target by seven days. The skeptic could not verify any of it: every evidence reference is a web source and this phase runs corpus-only, with no local options chain, long-dated open-interest or implied-volatility file. The jury therefore recorded not_verifiable rather than exists, and because the framework's watchlist-only overlay attaches only to a fits or leans-fit verdict, it does not apply here — watchlist_only is false in the tally. Full treatment in Clock.

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

The tally records the re-rating arithmetic as unavailable. Its re_rating_math block carries a null bar, null normalized adjusted FCF, null implied market cap at bar and null upside, with the note: "Re-rating math unavailable because the applicable bar or normalized adjusted FCF is missing." That is a direct consequence of fit_features.adjusted_fcf, adjusted_fcf_yield, yield_baseline, balance_sheet_class and market_cap all returning not_computable.

What the surviving claims do carry is the price side of the same test, computed from the filed segment measures rather than from the feature file.

Source: derived from the filed segment measures and CapIQ consensus normalized EPS (vintage 29 July 2026); earnings components from the FY2025 10-K [37].

On the mid-cycle normalization the 10% line sits at $34.23 or $30.42 a share, 67% to 70% below the current price and 59% to 64% below the 12 March 2026 trough. On FY2028 consensus the framework-basis 10% price is $67.00, 34.7% below today and 20.1% below the trough; on the unadjusted headline it is $86.00. Reaching 10% at $102.66 instead of by a lower price would take FY2028 framework EPS of $10.27, equivalent to headline EPS of $12.17 — 41.5% above the $8.60 consensus mean and 32.2% above the $9.20 consensus high.

The target test, stated as arithmetic rather than as a recommendation: three times $102.66 is $307.98 within three years. At the FY2028 consensus framework EPS of $6.700 that is 46.0x; on the unadjusted headline EPS of $8.60 it is 35.8x, which is approximately the 35.5x trailing multiple the stock carried at its January 2025 peak. So a 3x in three years is a claim that the January 2025 multiple returns in full and consensus is delivered on top of it.

The base rates come from this name's own record. Five completed NYSE-era drawdowns of 25% or more ran 33.9% to 57.7% deep, with a median of 47.7%, a median 14.5 months from trough back to the prior peak, and a median 22.3-month round trip; the slowest, 2014–16, took 28.7 months from trough to recovery and 53.3 months peak to peak. The current episode is the second deepest at 49.8% and took 13.3 months to reach its low. Those are round-trip base rates: regaining the $167.07 peak from $102.66 is +62.7%, not +200%. Full treatment in Clock.

Contested and undetermined

Contested: X1, X2, X3, X4 and S1 — all five exclusion criteria, and no pillar criterion. Three seats voted on each, and the split is identical across all five: seat A returned not_met, seat B returned no_hit, seat C returned met. Cross-family agreement is recorded as false on each, and the name-masked seat differed from the named seats on X1 through X4, which is what carries the prior_driven_risk flag and the low confidence tier.

Both readings, from the dockets. The first reading is that the exclusion test is not satisfied — seat A's not_met and seat B's no_hit say the same thing in different words. The second reading is seat C's met, which its written basis uses to mean the screen is passed, not that the exclusion fires: on X4 it records "the extreme-multiple prong is not met"; on X2, "the weak-ownership prong is absent"; on X1, "the theoretical Strategic Holdings route names no auto OEM and is only 3.9% of segment earnings". Read as prose, the three seats reach the same conclusion in three vocabularies. Read as verdict labels, they disagree, and the deterministic tally compares labels. The tally's own exclusion_hits array is empty; the contested marking and the confidence downgrade it drives stand as recorded, and neither is adjusted here.

Undetermined: P2, unanimously, with the missing datapoint named by every seat. In each juror's own words:

Seat A — fit_features.fcf_stability.rolling_5y_avg — empty; requires stock-based compensation for FY2016-FY2019 and a complete consecutive five-year acquisition window to build the adjusted-FCF series at all.

Seat B — a consecutive five-year adjusted-FCF series — fit_features.fcf_stability.rolling_5y_avg is empty and adjusted_fcf is null in every year (SBC missing FY2016-FY2019; no complete five-year acquisition window with SBC)

Seat C — rolling 5-year adjusted FCF stability series, including complete SBC and acquisition inputs

Seat D — five consecutive years of framework adjusted FCF and rolling-average volatility

Masked seat — Five consecutive fiscal years of adjusted FCF (FCF minus SBC minus trailing 5-year average acquisitions) — SBC for FY2016-FY2019 and a complete consecutive five-year acquisition window are absent, so fit_features.fcf_stability has an empty rolling_5y_avg and a null CV.

I1 is recorded as not_verifiable rather than undetermined, for the reason given above: a citable in-corpus options source giving listed expiries of at least 12 months, contract-level open interest and spreads, and a dated implied volatility does not exist in this run's corpus.

Provenance

Source: the run's deterministic fit tally (jury composition, agreement, spreads, mask divergence, skeptic counts) and the adversarial trial tally (order stability).

Two model families sat on the jury and agreed on every criterion that decides the verdict, including the gate; they disagreed only on how to label five exclusion criteria whose written reasoning matches. The verdict was pressed harder than the confidence tier suggests — but the tier is computed from label agreement, not from reasoning, and a masked seat that diverges on four criteria is precisely what the framework treats as a signal that the answer may be running on priors rather than on evidence. Two specific findings did not survive intact: the 3 February 2026 trigger was weakened to a dated price-and-volume window with an unsourced proximate headline, and the entire instrument-context finding was recorded unverifiable because its sources are external to the corpus.

The falsifier ledger

These are the standing conditions under which the run's reading would have to change. Thresholds, direction and window are as recorded.

Name-specific conditions, with their thresholds and windows.

• Fee engine. FRE per share falls below $1.13 for two consecutive quarters (Q2'26/Q3'26), or FPAUM declines sequentially from $615bn — the recurring engine, which grew +13.7% in FY2025 and +23% per share in Q1'26, would then be compressing rather than the multiple.
• Embedded gains. Total embedded gains fall below ~$15bn by 4Q'26 results without a commensurate rise in realized carry - meaning the $18.3bn reservoir is being written down at DCF-heavy Level III marks (55% PE / 91% real assets) rather than harvested, which would simultaneously validate the permanent side's valuation-sensitivity exhibit.
• Insurance. Insurance operating earnings decline year-over-year, or Global Atlantic requires a capital raise, ratings action, or material surrender/recapture event - turning the conceded, management-labelled-cyclical spread compression into a realized structural ROE reset in the 22%-of-operating-earnings segment.
• Realizations. FY2027 gross realized carry fails to exceed FY2026's despite the >$1.2bn signed forward pipeline, i.e. 'delayed monetizations… would not be lost… shift to 2027 and beyond' proves untrue - two consecutive realization-drought years with no catch-up.
• Fundraising. New capital raised over the four quarters to 1Q'27 falls below ~$100bn versus the verified $129bn record — the only clean read on whether clients repriced the franchise itself.

The framework's standing templates, carried at the level the system states them, with no name-specific threshold attached:

• 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.

The ledger as recorded. The conditions above are restated in several places across the run's briefs, at different levels of detail. The remaining entries are reproduced verbatim below — including the internal source tags two of them carry — so the ledger the verdict machinery holds and the ledger on this page are the same ledger.

The two tagged entries point at pages this report already links: the Q1 FY2026 call [38], the Level III valuation-weights disclosure [39], the Global Atlantic surrender risk factor [40] and the insurance-subsidiary ratings risk factor [41].

Data gaps

The run could not answer the following, and the verdict was reached in spite of them rather than by working around them.

The framework's own yield machinery does not compute for this company. fit_features returns not_computable for adjusted_fcf, adjusted_fcf_yield, yield_baseline, balance_sheet_class, fcf_stability, float_retirement_years and market_cap. The root cause is structural rather than a data-feed failure: KKR's consolidated cash-flow statement includes its funds' and Global Atlantic's flows, so no meaningful reported free cash flow exists. Stock-based compensation is missing for FY2016–FY2019 and no complete consecutive five-year acquisition window with SBC exists, which is what empties fcf_stability and nulls adjusted_fcf in every year. data/financials/cash_flow.json carries capex and free cash flow only through FY2019 and no SBC field for any year. Every yield, float-retirement and balance-sheet-class figure on this report is therefore derived from the filed segment measures with workings shown, and the adjusted-FCF substitute omits the acquisition deduction, which biases it upward. The 25.0-year float-retirement figure and the 1.40x net-debt-to-Adjusted-EBITDA classification are the analyst's, computed from cited filing pages.

Market capitalisation is reconstructed, and the null cascades. fit_features.market_cap is null with the reason "no positive annual period-end or outstanding share count". The U2 screen, every yield, every float-retirement figure and every consensus_forward_yield entry had to be computed from the 10-K cover-page share count (891,550,894 at 24 February 2026) and the price feed instead. fit_features.share_count_trend.per_year is empty and both CAGRs are null; the share-count series was read directly from the audited statements of operations across three 10-Ks.

Revenue history is stale in the feature file. fit_features.revenue_trajectory stops at FY2020 and tracks the Fees and Other line rather than GAAP total revenues, so consecutive_decline_years = 0 and three_year_hsd_decline = false are computed on a stale window. FY2021–FY2025 revenue was read off the filed statements of operations instead ($19,464.3 million for FY2025); the flag value is unchanged by the extension.

Consensus vintages are incomplete on both ends. data/sp/estimates.json carries dated revision snapshots only for FY2027 and FY2028. There is no FY2026 vintage at any date, so the FY2026 pre-cut figure used in the NPV model ($6.7531) is imputed by grossing the current $6.11654 up by the 9.43% average out-year cut; if a genuine pre-trigger FY2026 vintage exists above that level, the measured cut is larger and the trigger-window gap smaller. The earliest vintage of any kind is 29 January 2026, so the consensus path across the first year of the drawdown — during which the price fell 31.2% — cannot be reconstructed at all. Driver-level consensus has no historical vintages either: data/sp/va.json is a single 21 July 2026 snapshot. Vendor free-cash-flow consensus is internally inconsistent with the same file's cash-from-operations estimates and publishes no contributor count, so the consensus check uses normalized EPS instead.

The seller cannot be identified. FINRA returned zero reported short-interest rows for KKR, and the short-interest history, short-sale-volume, borrow-pressure, peer-context and public-net-short files are all empty, so short-interest level, change, days-to-cover and borrow cost cannot be stated in either direction. data/governance/beneficial_owners.json parses 51 Schedule 13D/G filings but resolves no owner names and no percentages, so institutional turnover through the drawdown cannot be quantified; the 2026 insider record comes from data/governance/insider_activity.json because the insider-transactions index ends at 2025-05-06.

Competitive structure rests on inference. KKR's own competition disclosure names no competitor and discloses no market share, so the peer AUM figures behind the P1 market-structure finding come from the peers' own FY2025 10-Ks, and the industry pool used as the denominator is a 2022 Preqin-based estimate carried in KKR's April 2024 investor day — the implied shares are upper bounds, not measurements. The Strategic Holdings segment's 19 constituent companies are not individually named in the FY2025 10-K, so the X1 auto-OEM check rests on the absence of any automotive reference in the business and segment disclosure rather than on a published holdings list. Peer valuation multiples, the current market capitalisation and the analyst rating distribution have no filing page behind them.

China exposure is measured once, in 2023. No China revenue line, China asset line or China-specific risk factor exists in the FY2025 10-K. The only quantified figure in the corpus is the April 2024 investor day footnote measured at 31 December 2023, so the S1 quantification is 31 months old, and the FY2025 10-K does not disaggregate the $410.1 billion balance sheet by geography, so asset-side dependence cannot be measured at all.

The trigger has no named headline, and web research was unavailable. No document in this corpus names the specific announcement that caused the market-wide move of 3 February 2026; the mechanism is cited to three peer filings and to KKR's own call, and the date is fixed by the price and volume record. The search tool exits with a missing-dependency error and the research provider returned an insufficient-credit error, so no external confirmation could be obtained for the trigger, the pre-trigger consensus record, current competitive and technology threats, corporate-bond refinancing context, or the options chain behind I1.

Instrument facts are partial. The implied volatility of the 21 January 2028 series specifically was not verifiable from available sources — only the 30-day measure (41.5%, data date 28 July 2026) is dated and citable. Contract-level open interest for the 17 December 2027 and 21 January 2028 expirations was not verifiable from available sources; liquidity is characterised only at session level, 6,668 calls and 1,805 puts on 28 July 2026.

The record stops before the next print. No transcript later than the Q1 2026 call of 5 May 2026 exists in the corpus, and second-quarter 2026 results were scheduled for 30 July 2026, one day after this report was written — so realized carried interest against the $720 million Q1 run-rate, the updated forward monetization figure against $1.2 billion, repurchase activity after 1 May 2026 and the authorization balance are all unobserved. No successor Investor Day or medium-term guidance framework has been announced to replace the 2024–2026 targets that expire 31 December 2026, so the post-2026 guidance-reset catalyst has no date. KKR does not publish a year-by-year debt maturity schedule; the finest grain in the filings is a four-bucket contractual-obligations table. KKR's own corporate credit ratings are not disclosed anywhere in the corpus — the A and A2 ratings in the 10-K are Global Atlantic life-insurance subsidiary financial-strength ratings. The corpus holds no KKR annual report before FY2021, so the multi-decade operating record is asserted from the 2008–2009 IPO disclosures rather than from a continuous filed series, and no FY2019 After-tax Distributable Earnings was located, so the segment-basis series starts at FY2020. Enterprise value is approximate — consolidated debt is dominated by non-recourse fund and CFE obligations, so market value is used for every damage measure — and no KKR filing or presentation in the corpus discusses the share price, the drawdown or the multiple directly.