Molina Healthcare (MOH, $9.1B cap) is the purest-play Medicaid MCO in the public markets — roughly 95% of revenue from government programs, highest loading in the complex on state Medicaid rate cycles. The stock fell 54% from the 2025 peak through February 2026 on sector rerating, deepened when management cut 2026 guidance to $5+ EPS and amended its credit covenant to a 1.75x interest coverage floor. The Q1 2026 10-Q (filed April 23) resolves the covenant question. It leaves unresolved whether the trend improvement hidden in reserve development ever reaches shareholders.

What the Filing Says

Covenant kill risk eliminated. The $740M in term loans maturing February and August 2027 are gone. Zero term loans outstanding. Credit Agreement extended to November 2030. Revolver $0 drawn on $1.25B available. Compliant with the amended 1.75x ICR. Next maturity June 2028 ($800M senior notes). S&P downgraded to BB- in April post-paydown; Moody's Ba2 unchanged.

Medicaid MCR 92.0% vs 92.9% FY guide. Up 170 bps YoY (from 90.3% Q1 2025). Described as "modestly favorable to expectations." Q1 is seasonally the best quarter; management has separately guided that 2/3 of FY earnings land in H1. The word "trough" — used prominently on the Q4 2025 earnings call — is absent from the 10-Q.

$253M favorable reserve development absorbed by MLR corridors. Prior-period reserve development ran more favorable than Q1 2025 ($186M) in every segment: Medicaid -$153M, Medicare -$71M, Marketplace -$26M. The 10-Q: "The impact of prior year reserve development in the first quarter of 2026 was mostly absorbed by minimum MLRs and medical cost corridors." The improvement did not reach P&L. It was returned to states.

Cross-ticker convergence. ELV's Q1 10-Q (same day) reported $1,124M favorable development vs $1,025M prior year. UNH reported MCR -90 bps qualitatively. Underlying medical cost trend is improving across the complex. Capture diverges by product mix: MOH (≈95% government) and CNC (≈75%) have corridor absorption; ELV (≈45%) and UNH (≈40%) capture directly as earnings. MOH cannot monetize the improvement as earnings for 12-18 months — until state rate updates reflect the better trend.

Medicaid year-end membership guide cut to 4.5M from ≈4.8M. 300K more attrition than previously expected. Sector-wide: ELV Medicaid -4.6% YoY, UNH -5.4% YoY. OBBBA eligibility changes plus PHE redetermination tail, not MOH-specific.

Florida CMS contract absent from filing. Zero mention of the $6B annual run-rate contract with "late 2026" go-live. No implementation cost disclosure in MD&A. Either normal deferral to earnings calls or schedule concern.

What the Market Thinks

Price $174.68 after +14.2% on the filing day, +23% month, RSI 85. Forward P/E 20.4x, implying consensus ≈$8.50 forward EPS (recovery beyond the $5+ 2026 guide). IV term compresses from 53% near-dated to 45% Jan 2027. Options-implied survival over 20 months: ≈100% above $130 (-25%), 93% above $85 (-50%). Near-zero tail.

Consensus price target is clustered around current. The market has priced covenant resolution and accepted the recovery narrative. What it has not obviously priced is the product-mix divergence.

Why the Gap Exists

Sell-side models tend to apply uniform recovery pacing across the MCO complex — rate cycle normalizes, MCRs revert, targets move together. The 10-Qs contain the corridor-absorption language plainly, but few consensus models decompose recovery speed by government-revenue share. If the corridor-lag mechanism is real, MOH (and CNC) should recover later but bigger than ELV and UNH. Consensus targets assume they recover together.

Separately, the +14% filing-day move extracted most of the obvious edge. The covenant resolution is priced. The structural frame is not.

Best estimate of the gap: MOH total return should exceed ELV total return by 10-25 percentage points over 18 months if the 2027 rate cycle reflects improved trend. If corridor lag is already priced, or if ELV's MA/commercial levers dominate, the gap collapses to zero. We can't narrow below 30 percentage points at this horizon.

Risks

1. 2026 is not the trough. Q1 at 92.0% is sub-guide but Q1 is seasonally best. If Q2-Q4 rate updates don't close the trend gap, FY Medicaid MCR prints above 92.9% and "2027 inflection" becomes "2027 trough."
2. Florida CMS slip. $6B run-rate contract silent in Q1. A push to 2027 start removes the near-term embedded-earnings catalyst.
3. OBBBA membership attrition. 15-20% Expansion reduction by 2029 on ≈1.2M members (≈25% of book). Sector-wide, but MOH has the highest Expansion loading.
4. Sub-dividend channel narrowing. Parent received $35M from regulated subs in Q1 vs $648M through 9mo 2025. If subs keep conserving capital, parent-level liquidity tightens despite the debt extension.
5. High idiosyncratic vol (≈57%). Trailing regression α_orth = -31.8% annualized; mean reversion to zero is not automatic. Sharpe on pure-Medicaid expression is structurally constrained by realized vol.

Catalysts

• ~May 5-10, 2026: CNC Q1 2026 10-Q. Tests whether corridor absorption replicates across the other pure-Medicaid operator.
• July 22, 2026: MOH Q2 2026 earnings. Florida CMS disclosure and Q2 MCR direction. Primary catalyst.
• Q4 2026: Florida CMS go-live target.
• January 2027: State 2027 rate cycle. Core thesis inflection point.
• February 2027: MOH FY2026 10-K.
• October 2027: 18-month corridor-release window.

What Would Change Our Mind

• CNC Q1 Medicaid HBR ≥93.5% → corridor thesis weakens; MOH Q1 was idiosyncratic, not a sector signal.
• Management confirms Florida CMS slip to 2027 on Q2 call → embedded earnings ramp pushes out ≈12 months.
• Management reinstates "trough" language on Q2 call → signals Q1 was in-line and H2 expected to worsen (counterintuitive, but the word matters).
• New covenant amendment needed, or subsidiary dividend extraction freezes → parent-level liquidity thesis re-opens.
• ELV earnings inflect 20%+ in H2 2026 on MA repositioning → commercial/MA levers dominate corridor lag; pair-trade frame breaks.

Evidence

Trailing regression (250d) against SPY/QQQ/XLV: β_SPY 0.21, β_QQQ -0.67, β_XLV 1.04; idio variance 92%; α_orth -31.8% annualized; σ_idio 57.6%.