Yield

Yield

On the framework's adjusted-FCF basis, Molina's yield is not a clean number: FY2025 reported free cash flow was negative (−$636M), so current-year adjusted FCF is roughly −$854M and the spot yield is meaningless. The case rests on normalization. Consensus forward FCF averages ~$1.26B over FY2026–FY2028 — about 10.6% of today's $11.9B market cap raw, and ~8% after the framework's SBC and acquisition haircut. That sits at the bottom of the 8–9% fortress bar (Molina carries net cash) and roughly 190 bps under the 10% default bar.

The deterministic feature file could not compute adjusted FCF (stock-based compensation is absent from the structured feed, and acquisitions were defaulted to zero — false for a serial acquirer like Molina). Every figure below is reconstructed directly from the filed cash-flow statements and cited to the page; the gap is recorded in the ledger.

The adjustment, line by line

The framework's yield basis is adjusted FCF = reported FCF − stock-based compensation − the trailing 5-year average of acquisition spend. Molina's own cash-flow statements supply all three lines. Stock-based comp is modest ($47M–$116M). Acquisitions are not: "Net cash paid in business combinations" ran from $3M to $755M a year as Molina bought Medicaid and Marketplace books (Magellan Complete Care in 2020, ConnectiCare for $350M in 2025 [1]), so the 5-year-average acquisition line is the material part of the haircut.

Adjusted FCF = reported FCF − SBC − trailing 5-yr average acquisitions; shown only for FY2023–FY2025, where a full five-year acquisition window (FY2019 onward) is sourced. Derived from company filings. Sources: FY2025 10-K, Consolidated Statements of Cash Flows [2]; FY2022 10-K [3]; FY2021 10-K [4].

The 5-year-average acquisition line, terminal year inclusive: FY2023 uses FY2019–FY2023 — $(0+755+129+134+3)/5 = $204M; FY2024 uses FY2020–FY2024 — $273M; FY2025 uses FY2021–FY2025 — $171M. So FY2025 adjusted FCF is −636 − 47 − 171 = −$854M, and FY2023 — the last clean year — is 1,578 − 115 − 204 = $1,259M. The adjustment removes roughly $220M–$390M a year; on Molina it is the acquisition average, not SBC, that does the work.

The reported line is worth seeing plainly, because it is what makes single-year yield unusable here.

Reported FCF = operating cash flow − capex. Source: FY2019–FY2025 10-Ks, Consolidated Statements of Cash Flows [5].

Free cash flow swung from +$2.0B (FY2021) to −$636M (FY2025) with no comparable swing in the underlying business — net income stayed positive at $472M even in FY2025 [6]. The swing is working capital: in FY2025, "amounts due government agencies" alone consumed $591M, with receivables and medical-claims payable another ~$277M [7]. For a government-payer insurer, a single year's FCF is a timing artifact, not the earning power.

The yield, three ways

Because current-year FCF is negative, the three framework cuts all read low, and each is stated on today's $11.9B market cap ($224.82 × 52.9M shares, 16 July 2026).

Adjusted FCF ÷ market cap of $11.9B. Current = −$854M; trailing 3-yr avg = $187M (FY2023–FY2025 adjusted, averaged); through-cycle = $652M (FY2020–FY2025 reported-FCF average $1,006M − avg SBC $85M − avg acquisitions $268M). Derived from company filings [8].

The current-year cut is negative (−7.2%). The trailing three-year average is 1.6% — that window happens to capture the entire deterioration, since FY2024 was already weak and FY2025 turned negative. The fairest own-history figure is the six-year through-cycle average adjusted FCF of about $652M, which is 5.5% on today's price.

On the fortress "jump" signature — a stable low yield that suddenly vaults toward the bar — Molina shows a muted version. That same $652M of through-cycle adjusted FCF was about 3.1% at the September 2024 peak ($360.77 on ~57.7M shares, ~$20.8B market cap) and is 5.5% now: the 66% drawdown (Dislocation) roughly doubled the yield. But it is a partial jump, not the Microsoft-style 4%→9% snap, because the numerator collapsed at the same time the price did.

Which bar applies

The balance-sheet class selects the reference line. The deterministic feature returned unknown (it needs an EBITDA field the income feed does not carry), so the computation is done here from the filed balance sheet.

At FY2025 year-end, long-term debt was $3,766M against cash and equivalents of $4,248M and a further $4,008M of short-term investments [9]. Net debt is therefore −$482M against cash alone, and −$4.5B including short-term investments — a net-cash position. Consensus net-debt estimates confirm it and see it deepening: −$0.8B (FY2025) widening to −$2.5B (FY2028). EBITDA (operating income $781M + depreciation and amortization $195M ≈ $976M in FY2025) is positive, so net-debt/EBITDA is negative. By the framework rule (net debt ≤ 0 → fortress), Molina is fortress, and the applicable line is 8–9%.

One honest qualifier: an insurer's cash and investments are largely regulatory capital held at licensed subsidiaries, not freely deployable like a tech balance sheet's net cash. On a stricter reading Molina behaves like a moderate balance sheet, which selects the 10% default bar. The name sits short of both:

Normalized adjusted FCF yield ≈ 8.1% — at the floor of the 8–9% fortress bar, and about 190 bps short of the 10% default bar.

The 8.1% is the normalized figure derived next; on trailing figures the gap is far wider.

Normalized mid-cycle yield

Molina is not a commodity cyclical, but it is mid-cycle-sensitive in a specific way: managed-care margins move with the medical cost trend, and FY2025–FY2026 sit at a cost-trend trough. Consensus normalized EPS falls from $13.93 (FY2025) to $5.16 (FY2026) — the guidance-cut year — then recovers to $9.29, $12.93 and $20.30 through FY2029 as Medicaid rates reprice and the Marketplace/Medicare books readjust. Normalization has to look through the trough.

Two independent routes converge on roughly the same place:

From consensus forward FCF (next section): the FY2026–FY2028 average raw FCF is ~$1,259M; subtracting a normalized $100M of SBC and a $200M ongoing acquisition run-rate (a modest round-up of the $171M trailing 5-yr average, given Molina's continued M&A cadence — it closed the $350M ConnectiCare acquisition on 1 February 2025 [10]) leaves ~$959M adjusted → 8.1%.

From mid-cycle earnings power: consensus mid-cycle net income lands near $1.0–1.2B (EPS ~$13–20 on ~53M shares). Adding depreciation and amortization (~$200M) less capex (~$130M), with through-cycle working capital near zero, gives ~$1.07–1.27B of FCF; less the same $300M of SBC and acquisitions leaves ~$770–970M adjusted → 6.5%–8.2%.

The assumptions a skeptic should test: (1) medical cost trend normalizes by FY2027 so the FY2026 EPS trough is genuinely a trough; (2) acquisition spend stays near its ~$200M/yr trailing pace rather than stepping up; (3) working capital neither drains nor releases cash on average through the cycle. Loosen (1) — cost trend stays elevated into FY2028 — and the normalized yield falls toward 6%; tighten acquisitions to zero and it rises toward 9–10%.

The consensus check

Consensus forward FCF (CapIQ, via data/sp/estimates.json, generated 22 July 2026) is the anchor, because the framework's own forward feature reads directly off it. The near-term FCF profile is lumpy — FY2026 is inflated by the reversal of the FY2025 working-capital drain, FY2027 gives some of it back — so the multi-year average matters more than any single year.

Raw yield = consensus mean FCF ÷ $11.9B market cap (matches fit_features.consensus_forward_yield). Adjusted yield subtracts $100M normalized SBC + $200M acquisition run-rate before dividing. Source: CapIQ consensus, data/sp/estimates.json; derived.

Raw consensus FCF clears the 10% bar in three of the four forward years and averages 10.6% over FY2026–FY2028 ($1,259M ÷ $11.9B). After the framework's SBC and acquisition haircut it averages ~8.1% — the fortress bar is essentially met on consensus's own numbers, the 10% default bar is not.

That places Molina between the two canonical readings. The setup is not "consensus disagrees" — the sell side already models FCF back above the fortress line by FY2026. It is closer to fear: the equity fell 66% while consensus carries $1.3B+ of normalized FCF. But it is not the clean Centene case where consensus forward FCF clears the default bar outright; here the adjusted number lands ~190 bps under it, so the reversion underwrite must be explicit rather than assumed.

The mean-reversion path. The mechanism is the managed-care repricing cycle. Molina's FY2025–FY2026 margin compression is a cost-trend forecasting miss — the same category of error the framework hunts in insurers — and Medicaid rates reset actuarially on a 6–12 month lag, with Marketplace and Medicare bids repriced annually. Consensus embeds exactly this: normalized EPS recovering from the $5.16 FY2026 trough back above the FY2025 $13.93 level by FY2029. My estimate: roughly a 70% probability the normalized adjusted FCF yield holds at or above the 8–9% fortress bar within one to three years — this is consensus-backed, not a stretch — but only about a 45% probability it clears the stricter 10% default bar sustainably, which needs either the acquisition drag to shrink or margins to run above mid-cycle. What consensus would have to concede for the underwrite to fail: medical cost trend staying elevated past FY2026 so rates never catch up, or a collapse in Marketplace enrollment if the enhanced ACA premium tax credits lapse — a real, dated risk to Molina's Marketplace book that the Damage Math trial weighs.

FCF-to-revenue conversion

The framework treats a deteriorating FCF/revenue trend as a strike against the flywheel, so it deserves a straight look.

Reported FCF ÷ revenue. Source: FY2019–FY2025 10-Ks [11]; derived.

Conversion is low, volatile, and ends the period falling — 9.4% in FY2020 to −1.4% in FY2025, with a cumulative FCF/revenue of just 3.0% across the six years. Read literally, that undercuts the buyback flywheel: a business that converts 3% of revenue to cash cannot fund large repurchases from cash generation alone.

The counter-fact sits in the same statements. Revenue grew from $16.8B to $45.4B over the period, and for a government-payer insurer rapid premium growth consumes working capital — new members and new state contracts arrive with receivables and rate-settlement timing before the cash catches up. Net income never went negative, and Molina still bought back $1.0B of stock in each of FY2024 and FY2025 [12] — funded partly off the balance sheet, which is how the Self-Help buyback capacity has to be read. The conversion trend is the genuine risk to the yield case: it is the falsifier to watch, and if FCF/revenue does not recover as the FY2025 working-capital drain reverses, the normalization above does not hold.