Fit

Fit — Lyft (LYFT)

Outside the framework's universe (U2 not met)

Lyft clears the geography screen — a primary US Nasdaq listing [1] — but fails the size screen: at roughly $5.9B of market value it sits about 41% below the framework's $10B floor, and a universe miss ends the fit question before any pillar is weighed. Confidence is medium. No exclusion was triggered, nothing was contested, and neither jury family split on the universe call.

Universe and exclusions

Here is the decisive point. The framework's universe is US-listed equities (no Chinese ADRs) with a market capitalization above $10B. Lyft passes the first test and misses the second.

The screen that closes the question is size, not scale. Lyft is not a small business: FY2025 revenue was $6,316.3M and gross bookings were $18.5B [2]. What falls short is the equity value: 417.659M diluted weighted-average shares at the $14.20 close (2026-07-24) is a $5.93B market capitalization — $4.07B, about 41%, below the $10B bar. The counter-fact sits in the same breath: on revenue and volume this is a large, duopoly operating business; the framework's floor is a claim about equity value, and on that measure Lyft does not clear it.

Source: market cap derived from 417.659M shares × $14.20 (2026-07-24); share and revenue base per FY2025 10-K, Consolidated Statements of Operations [3].

None of the framework's hard exclusions is triggered — a clean screen, checked item by item.

Sources: geography and listing per FY2025 10-K cover [4]; geographic revenue split per FY2025 10-K [5]; guidance-delivery record per the Q4 FY2024 call [6]; the full treatment of each exclusion is in the Business, Durability and Self-Help tabs.

Two exclusion items are worth naming plainly even though neither triggers. Insider economic ownership is thin — all ten officers and directors hold under 1% of the shares, and CEO Risher owns about 0.27% [7] — but the promotional-CEO exclusion turns on repeated missed promises, and the delivery record is clean, so the exclusion fails on its own terms. And the FY2025 headline net income of $2,844.0M is a deferred-tax valuation-allowance release over a $188.4M operating loss [8], not cash earnings — the cash case rests on the $1,115.6M of free cash flow instead [9].

Pattern match

Of the framework's four recognition setups, Lyft most resembles the fourth — a quality tech duopoly on a fear dip, where the fear is specific and testable (here, autonomous-vehicle displacement rather than TikTok or a China quarter). It is not the first setup (no bank), the second (no dividend), or the third (no insurer repricing a mis-forecast year). The resemblance to the fourth is partial: the framework's precedents are monopolies and dominant platforms, whereas Lyft is the roughly 24% No. 2 in a two-player US market with low switching costs [10]. This is framing, not a verdict — the pattern is a lens on the setup, and the setup is moot once the universe screen fails.

The pillar ledger

The pillars are recorded here for completeness; the universe miss already closes the fit question, so none of what follows can reopen it. Each criterion is stated against its reference line, not scored.

Source: deterministic fit tally; per-criterion arithmetic traces to the FY2025 10-K [11] and the linked evidence tabs below.

Year-10 gate (P1)

The binary gate does not hold. Very-high-conviction year-10 durability of adjusted free cash flow cannot be established for a No. 2 in a low-switching-cost duopoly that owns no autonomous-vehicle technology (Level 5 was sold in 2021) and depends on AV partners who are also better-capitalized competitors — a variable Lyft's own 10-K calls a "new and evolving market, which makes it difficult to predict its acceptance, its growth" [12]. The jury's mean probability that the gate holds was 0.485, spread 0.12. The strongest surviving counter-fact: early AV adoption in San Francisco has so far been additive to ridesharing volume, and Lyft's FlexDrive fleet-operations role offers a path to participate rather than be displaced — but an additive-so-far reading is not the very-high conviction the gate demands. Full treatment in Durability.

FCF consistency (P2)

Not met — unproven rather than disproven. The deterministic 5-year rolling adjusted-FCF series is empty because there are fewer than five consecutive positive adjusted-FCF years: adjusted FCF was negative in four of the last five years, and reported free cash flow turned positive only in FY2024 [13]. The counter-fact: the turn is recent and steep — reported FCF went from -$248.1M (FY2023) to +$766.3M (FY2024) to +$1,115.6M (FY2025) [14] — so the instability is a ramp out of losses, not mature-business volatility. Either way a stable five-year average cannot yet be computed. See Yield and Durability.

Dislocation and yield (P3)

The dislocation is real but the fear gauge is muted, and the yield clears its bar on one favorable year. LYFT fell 48.5% from a $24.57 peak (2025-11-12) to a $12.65 trough (2026-03-30) — a drawdown derived from daily price data — P3a met. But sustained selling volume ran only 1.35x its trailing median and the trough printed on below-median volume, so the capitulation P3b looks for is absent; the counter-fact is a single 74.2M-share panic session on the February 11 earnings gap. On yield, FY2025 adjusted FCF of about $702.9M is an 11.9% yield on the $5.93B market cap, clearing the roughly 8-9% fortress bar — P3c met — but only on the single best year, against a three-year-average adjusted yield of 1.3%. Consensus forward FCF is a ~12-15% adjusted yield, above the bar, so the sell side already underwrites it (P3d met; jury probability 0.675, spread 0.05). Full anatomy in Dislocation and Yield.

Balance sheet and self-help (P4)

The balance sheet passes; capital allocation fails the framework's hard rule. Lyft is net cash — net debt of -$79.0M, $1,837.2M of liquidity against $1,053.0M of near-zero-coupon convertibles with nothing material due before 2029 [15] — a fortress that can outlast the dislocation without ever prioritizing debt paydown (P4a met). But the framework fails outright any company whose share count keeps rising, and Lyft's diluted weighted-average shares grew 84% from 227.5M (2019) to 417.7M (2025), a 6.0% five-year CAGR driven by stock-based compensation [16] — P4b not met. The counter-fact is a genuine FY2025 inflection: point-in-time Class A shares issued fell from 409,474K to 400,856K after $500M of buybacks [17], the first annual decline — but the multi-year trend still reads rising, and P4c (dividend cover) is not applicable because Lyft pays no dividend. Full treatment in Self-Help.

Diagnosis (P5)

The adversarial trial put the probability the impairment is temporary at 0.62 — a lean toward temporary that this report carries and cannot override. The three blind judges landed at 0.62, 0.43 and 0.64 (mean 0.563, spread 0.21); the harness does not flag the ruling as contested. The dated near-term hit was a single non-recurring item: a $210M legal, tax and regulatory reserve, $168M of it reducing Q4 2025 revenue, on a quarter that still set records for profit and cash and reaffirmed the 2027 targets [18]. The counter-fact sits inside the same ruling: one seat at 0.43 leaned permanent, so meaningful one-in-three weight rests on the possibility that rising, non-discretionary insurance cost per mile is a structural margin drag rather than a passing item. Both cases are laid out in Damage Math.

Instrument context (I1)

Resolves not verifiable. Long-dated listed options exist on LYFT — expirations to December 2027 and January 2028, deep open interest of roughly 656,000 contracts — so the framework's watchlist-only-by-instrument consequence is not triggered; and 30-day implied volatility of about 76% sits above the framework's 60-70 reference band. These facts rest on web aggregators (AlphaQuery, Market Chameleon) dated 2026-07-23 with no in-corpus filing to cite, so the criterion is recorded as not verifiable rather than met. Detail in Clock.

Re-rating arithmetic

The framework's target test asks what price a bar-yield on normalized adjusted free cash flow implies, and how far consensus would have to move to get there. The deterministic re-rating math is unavailable here: the applicable fortress bar is 8.5%, but normalized adjusted FCF is not_computable in the feature file (the structured cash-flow feed carried no capex, SBC or acquisition lines), so neither the implied market cap at the bar nor the upside to it can be shown without improvising the numerator. That gap is stated, not filled.

What the tabs can offer is orientation, clearly caveated. On the single most favorable year, adjusted FCF of about $702.9M against the 8.5% bar would support roughly $8.3B of equity value — above today's $5.93B but below the $10B universe floor — and a transparent DCF on consensus FCF puts enterprise value near $16.4B at a 10% discount [19]. Both figures lean on a FY2025 cash base flattered by roughly $0.8B of insurance-reserve working-capital build, which the Yield tab normalizes. For base rates, Lyft's own turnaround-era history shows four drawdowns of 46-55% since 2023, three of which round-tripped to their prior peak within 10-18 months — encouraging, but a small single-regime sample, and the pre-2023 regime shows the opposite: a roughly 90% de-rating from the 2019 IPO that never recovered (Clock).

Contested and undetermined

Nothing was contested: the tally records no criterion with both readings live, and the trial ruling is not flagged contested. Nothing was marked cannot-determine. One criterion, I1 (instrument context), resolves not verifiable — its facts rest on web aggregators rather than a filing — and P2 (FCF consistency) is best described as unestablished rather than failed, because fewer than five consecutive positive adjusted-FCF years exist to average. Neither reopens the universe result.

Provenance

Source: deterministic fit tally, provenance block (jury composition, order-stability, mask-divergence and skeptic counts).

Two model families, four seats plus a masked probe, agreed unanimously on both universe criteria, so the finding that closes this report was not a close call. The verdict was pressed on the pillars — a skeptic pass that refuted none of the 16 fully-checked claims and a name-mask probe that moved no gate — but because the universe screen is categorical, that pressure changed nothing about the answer.

The falsifier ledger

These are the standing conditions that would change the framework read if the universe question were ever reopened (for example, if the equity re-rated back above the $10B floor). Several restate the same threshold from different tabs; each is recorded verbatim.

• 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
• Gross bookings decelerate below ~10% for two consecutive quarters, converting a priced de-rating into demonstrated demand impairment.
• Insurance cost-per-mile keeps compressing revenue margin AND adjusted EBITDA margin stalls or falls below FY25's 2.9%, breaking the 4%-by-2027 path.
• Documented U.S. rideshare share loss to Uber/AV entrants, contradicting the 'network partner' framing.
• 2027 targets ($25B GB, 4% EBITDA, >$1B FCF) missed or withdrawn, or FCF ex-insurance-float turns down.
• Through Q2-Q4 2026, insurance cost per ride declines outside California and insurance reserves/accruals stop being the primary working-capital driver.
• Through Q2-Q4 2026, high-teens bookings growth persists while incentive-program costs and sales and marketing as a percentage of revenue decline.
• FY2027 targets of $25B gross bookings, 4% adjusted EBITDA margin, and over $1B FCF are achieved with positive GAAP operating income and without material legal, tax, or regulatory add-backs.
• Lyft shows stable or rising share in Uber/AV-overlap markets without revenue-margin compression.
• Bookings/Rides decelerate below ~10% or turn negative for two consecutive quarters (next prints ~Aug and ~Nov 2026), converting a price event into a demand event.
• Insurance reserve build reverses and TTM FCF falls below ~$0.9B, showing the $829M working-capital tailwind was a one-time flatter rather than a recurring float dynamic.
• Documented U.S. rideshare share loss to AV/Uber, or GAAP operating loss keeps widening ex-legal/regulatory charges (insurance/incentives absorbing scale structurally).
• 2027 targets ($25B GB, 4% EBITDA margin, >$1B FCF) withdrawn or clearly missed at/after FY2026 results.

Data gaps

The run could not answer several things the fit rests near, from the tally's list:

• Framework adjusted FCF, adjusted-FCF yield, the yield baseline and the 5-year FCF-stability series are all not_computable in the feature file — the structured cash-flow feed omitted capex, SBC and acquisitions, so every FCF figure here is reconstructed from the filed consolidated cash-flow statements and should be treated as an estimate.
• Market cap in the feature file uses 417.7M diluted weighted-average shares; period-end shares were 400.9M after FY2025 buyback retirements, so true current market cap is nearer $5.69B and the yields are marginally higher than reported — but still far below the $10B floor.
• Precise, current (July 2026) US rideshare market shares are not in the corpus; the ~76%/~24% split is Bloomberg Second Measure's March 2024 reading.
• The official short-interest feed returned zero rows; the ~24%-of-float short-interest level is from publicly reported bi-monthly data via web, not the corpus.
• Implied-volatility and options open-interest figures are web-sourced (AlphaQuery, Market Chameleon, 2026-07-23); the run's own web-research phase failed on a billing error, so the I1 facts rest on external aggregators rather than a filing.
• No independent market-share time series is in the corpus to test the AV/Uber displacement premium that the residual price-vs-value gap represents; the trial's flip conditions name it as the decisive future evidence.