Damage Math

Damage Math

Molina's adjusted EPS fell from a $24.50 FY2025 guide to $11.03 actual, and FY2026 is guided to at least $5.00 [1] [2]. The stock fell 66% to its February 2026 trough and sits 38% below its peak. A transparent two-scenario NPV puts plausible value destruction at 18% if the hit is temporary and 52% if it is permanent. The gap that opened at the trough has largely closed at today's $224.82; the trial puts the probability the damage is temporary at 0.37.

The near-term hit — the numerator

The damage began as a series of guidance cuts, not a revenue miss. In April 2025 management reaffirmed full-year 2025 adjusted EPS of "at least $24.50 per share, representing 8% growth over the full year 2024" [3]. The year closed at $11.03 of adjusted EPS and $8.92 GAAP — a 55% shortfall to that guide and a 51% decline from 2024's $22.65 [4]. Reported net income fell from $1,179 million to $472 million [5].

The 2026 guide extends the reset rather than reversing it: adjusted EPS of "at least $5.00 per diluted share," itself burdened by $2.50 of identified drags: $1.50 from implementing the new Florida CMS Medicaid contract and $1.00 from the Medicare (MAPD) product [6]. Consensus agrees: FY2026 normalized EPS sits at $5.16 across 19 analysts (CapIQ, data/sp/estimates.json).

Source: FY2025 guide and 2026 guide, Molina earnings releases [7] [8]; consensus per CapIQ estimates (data/sp/estimates.json, vintage 2026-07-22).

The revision reached the out-years too, which matters because those years stand in for normalized earning power rather than the trough. Over the six months to July 2026, consensus FY2027 normalized EPS was cut from $16.34 to $9.29, and FY2028 from $25.09 to $12.93 — the latter a 48% cut to a year far enough out that it should already be past the rate-trend imbalance.

Source: CapIQ estimate momentum, 180-day vintage vs current (data/sp/estimates.json, momentum block; as_of 2026-01-23 and 2026-07-22).

One line held throughout: revenue. Premium revenue grew to $43,052 million from $38,627 million, up 11.5% [9], total revenue reached $45.4 billion, and consensus has FY2026 roughly flat at $44.3 billion (data/sp/estimates.json). The impairment is entirely a margin event, which the Dislocation anatomy dates to four 2025 guidance cuts and which frames the whole question here: is a margin that compressed a rate-reset away, or a level that has moved.

The price move over the same window

Against a roughly 55%–80% cut to near-term earnings, the equity fell less at its worst and much less today. From the September 2024 peak of $360.77, the shares bottomed at $122.65 in February 2026 (−66%) and have since rallied 83% to $224.82, leaving them 38% below the peak.

Source: daily price series; drawdown per the deterministic capitulation gauge (fit_features.capitulation_gauge.drawdown).

The contrast is the heart of the tab. Near-term (FY2026) consensus EPS is down roughly 80% from the pre-event ~$26 trajectory; the market cap is down 38% from peak, to $11.9 billion on 52.9 million shares. On enterprise value the move is similar — the balance sheet barely changed, with long-term debt rising from $2.9 billion to $3.8 billion and cash-plus-short-term-investments of $8.3 billion, much of it regulated at the health-plan subsidiaries and so not free buyback fuel [10]. This is the reverse of the Centene pattern Ruchir hunts, where the stock fell about one-for-one with the earnings cut and anchored to the trough. Here the market discounted a recovery from the outset: at $224.82 it pays 44x depressed 2026 consensus EPS but only 11x the $20.30 that consensus models for 2029.

The NPV arithmetic, two scenarios

Adjusted EPS is the cleaner owner-earnings proxy for this business than free cash flow: Molina pays no dividend, returns cash through buybacks, and its reported FCF swings on Medicaid working capital — a $535 million operating-cash outflow in 2025 against a $644 million inflow in 2024 [11]. The deterministic adjusted-FCF feature is not_computable here for want of a clean stock-comp series, so no adjusted-FCF yield is carried (see Yield); the model below is built on earnings.

The assumptions are stated so the result recomputes from the page:

• Owner earnings = adjusted EPS.
• Discount rate r = 9.8%, back-solved from the market's own pre-event pricing: $360.77 peak = $24.50 / (r − g) at g = 3%, so r = 24.50/360.77 + 0.03.
• Terminal growth g = 3%; capitalization rate (r − g) = 6.8%.
• Pre-event normalized earning power = $24.50 (the April-2025 FY2025 guide, equal to the 2029 investor-day target of $25).
• Transition years 2026–2028 use current consensus normalized EPS: $5.16 / $9.29 / $12.93.

The baseline, no-impairment value computes to $24.50 / 0.068 = $360/share, which ties to the observed peak — the model is anchored, not free-floating.

Temporary (V-shaped). Earnings sit at consensus through 2028, then earning power is fully restored to $24.50 from 2029, growing 3%. Discounted transition (4.70 + 7.71 + 9.77) = $22.2; terminal ($24.50/0.068 = $360, at end-2028) discounted = $272. Value ≈ $294/share, an 18% haircut to baseline — the cost of the earnings you forgo while recovering, even if nothing is permanently lost.

Permanent (level shift). Earning power resets to a lower plateau. A mild reset holds the plateau at the $20.30 consensus 2029 level: $22.2 + ($20.30/0.068 discounted) = $248/share, a 31% haircut. A harsher reset plateaus at $13 from 2028, with no recovery beyond the near-trough: $171/share, a 52% haircut.

Source: derived DCF-lite; discount rate 9.8% back-solved from the 2024-09-16 peak of $360.77 at g = 3%; transition-year EPS from CapIQ consensus (data/sp/estimates.json).

Now set the price damage beside the value damage. At the February trough of $122.65, the price had destroyed $238/share against the peak — more than even the harsh-permanent read destroys ($189). That is the dislocation: the price overshot the worst defensible value case by roughly $49, and the temporary case by roughly $172. At today's $224.82, the price has destroyed $136/share, which corresponds to a permanent plateau of about $16–18 of earning power — between the mild and harsh permanent scenarios.

Source: derived — price damage vs the $360.77 peak; NPV damage from the two-scenario model above (data/sp/estimates.json; fit_features.capitulation_gauge).

Probability-weighting closes the loop. At the trial's 0.37 weight on temporary, intrinsic value works out to between $217 (0.37 × $294 + 0.63 × $171) and $265 (0.37 × $294 + 0.63 × $248). The current $224.82 sits inside that band; the trough of $122.65 sat far below it. The gap Ruchir hunts was wide at the February trough and has largely closed after the 83% rally — at $225 the market is paying roughly what a permanent-leaning outcome is worth. A higher discount rate would narrow the gap further, not widen it: at r = 11%, the baseline falls to about $306 and the temporary value to about $255, so the temporary case loses its margin of safety fastest. The full snap-back to $24.50 in the temporary scenario is itself generous to the bull, and even on that generous input the price is no longer materially below value.

The trial — temporary or permanent, tried fairly

Whether the impairment is temporary or permanent was decided not here but by the profile's adversarial trial: two opposing, corpus-cited briefs, ruled on by three blind judges. Both cases carry real evidence.

The case for temporary. The 2025 cut is a rate-versus-trend timing gap on a repricing calendar, plus a Marketplace book of about 10% of revenue being deliberately shrunk to breakeven. Of the original $10.50 revision from $24.50, management attributes half to Marketplace and one-third to the Medicaid rate/trend imbalance [12]. The mechanism is quantified and asymmetric: every 100 bps of Medicaid rate improvement adds roughly $4.50 to EPS, 60% of revenue reprices each January at rates management projects "modestly in excess of trend," and second-half-2025 Medicaid already annualizes to a $6.50 "jumping off point" for 2026 [13]. Management counts $8.65 of "embedded earnings" waiting to be harvested and frames the episode as "inclement weather rather than climate change… temporary rather than permanent" [14].

The case for permanent. The damage survived multiple repricing windows and shows up in every segment. Consolidated MCR rose to 91.7% from 89.1%, with Marketplace blowing out to 90.6% from 75.4% and its medical margin falling to $423 million from $617 million despite premium nearly doubling [15]. The remedy for Marketplace is withdrawal, not recovery: membership is guided to 220,000 at year-end 2026 from 655,000 at the end of 2025 [16], with premium down $2.3 billion [17]. Medicare is exiting MAPD with a $93 million impairment [18]. And Q4 carried about $2.00/share of unfavorable retroactive California Medicaid adjustments — the exact delayed-and-retroactive state-pricing risk the model runs on [19]. The bull's own destination concedes the point: the 2029 target of $25 EPS is only April-2025's $24.50 reached four years late, and its bridge leans on future revenue (+$4.25), projected initiatives (+$1.50) and M and A (+$1.00) rather than the existing book simply healing [20].

The ruling. The judges put the probability the impairment is temporary at 0.37 — a lean toward permanent — with individual reads of 0.37, 0.29 and 0.38 (range 0.29–0.38, spread 0.09), and the result was not contested. Reading order barely moved it: judges who read the temporary brief first averaged 0.37, those who read permanent first 0.335, a 0.035 gap. This diagnosis is the report's, and this tab does not override it — the arithmetic above is what the diagnosis is applied to, not a competing verdict.

Source: profile adversarial trial ruling (ruchir/trial/tally.json); panel p_temporary = 0.37.

Which line broke, and whether it self-corrects

Two drivers account for almost all of the hit, and they carry opposite recovery profiles. Marketplace caused about half the original revision, and its correction is structural subtraction: management is cutting membership by two-thirds and pulling $2.3 billion of premium, so the loss is being removed from the P and L rather than repaired — a smaller, breakeven book, not the old growing one [21]. The enhanced ACA subsidies underpinning that book expire at the end of 2025, which is why the retreat is deliberate rather than cyclical.

The Medicaid piece — roughly one-third of the revision — is the part with a genuine self-correcting mechanism. Medicaid is priced in annual state-set per-member premiums that reset against a continuously moving cost trend; in 2025 locked rates lagged utilization, pushing the Medicaid MCR to 91.8% [22]. With 60% of revenue repricing each January and rates management expects modestly above trend, this is the driver that can mean-revert within 12–18 months — the Clock on which re-recognition depends. The risk that keeps the ruling tilted toward permanent is that the reset has already been tested across several windows and the retroactive-recoupment exposure (California, $2.00/share) is a recurring feature of the rate-taker model, not a one-off. What would move the diagnosis toward temporary is concrete and dated: FY2026 adjusted EPS landing well above the $5 guide with consolidated MCR below 90.5%, and January-2027 Medicaid rates exceeding trend by at least 150 bps with pretax margin back above 3%.

Source: FY2025 Form 10-K, segment MCR and medical margin table [23].

The bottom line for this tab is arithmetic, not a verdict on the stock: the near-term earnings hit was large (roughly 55%–80%), the price hit was smaller and front-loaded, and a conservative NPV destroys 18% of value if the impairment is temporary and 52% if permanent. At the February trough the price fell more than any of those cases justified — a real gap. After an 83% rally, at $224.82 the price sits inside the probability-weighted value band, and the trial's 0.37 reading leaves that value tilted toward the permanent end. The dislocation was there; most of it has been paid back.