Will CMS RADV audit results for PY2018 result in material Medicare clawbacks (>$50M) for Molina by Q3 2026?

The vacated CMS extrapolation methodology is the dominant factor — it structurally compresses the right tail of clawback magnitude. Without extrapolation, CMS recoveries are limited to sampled beneficiaries (typically 200-500), making $50M+ findings mathematically implausible for a 14% Medicare segment. Q1 2026 materials show zero RADV mention and no accrual in $14M GAAP net income. Residual probability reflects timing risk (CMS could issue new methodology and release findings within window) and non-extrapolation pathway risk.

Three converging negative signals: (1) court vacating extrapolation caps maximum recovery at sampled-beneficiary level, almost certainly under $50M; (2) MAPD exit for 2027 signals management is reducing Medicare footprint, inconsistent with known material pending liability; (3) Q1 earnings call entirely silent on RADV. Base case is non-release by Q3 2026 OR release with immaterial findings. Tail probability reserved for CMS appeal + aggressive methodology reissuance.

Somewhat higher than consensus — CMS has been signaling aggressive Medicare Advantage oversight and may appeal the extrapolation ruling or use alternative recovery mechanisms. Committee's original 28% was based on binary unknown framing; Q1 evidence is modestly supportive of lower probability but doesn't eliminate tail. Also accounting for possibility that findings are released but below $50M (resolves NO anyway) vs above $50M (resolves YES).

The extrapolation vacate is a material positive for MOH, but it doesn't eliminate risk entirely. CMS could appeal, find concentrated errors in the sample, or issue revised methodology. However, the Q1 disclosure silence is meaningful — management typically accrues for known pending material liabilities under ASC 450. Net: lower than prior 28% but not collapsing to base rate.

Slightly higher than consensus — the Q1 silence could simply reflect timing (findings not yet received) rather than confirmed immateriality. The committee's original 28% priced in significant uncertainty about scope. Extrapolation vacate helps but (a) CMS has been openly aggressive on MA overpayments, (b) sample-only findings concentrated on outlier diagnosis codes could still reach $50M, (c) the court ruling may be narrowly applied. Timing risk also favors non-release (NO) but not overwhelmingly.

Q1 2026 evidence is directionally favorable across multiple vectors: extrapolation vacated (primary), no accrual in impaired Q1 results, no legal proceedings disclosure, no earnings call mention, MAPD strategic retreat. Offsets: CMS could appeal methodology ruling, timing of release is uncertain, committee's original framing acknowledged binary unknown nature. Central estimate ~18%.

Extrapolation methodology vacated by court limits CMS recovery to sampled beneficiaries. Sample-level findings on 14% Medicare segment unlikely to exceed $50M. Q1 earnings call and 10-Q silent on RADV. MAPD exit reduces future exposure.

Prior 28% reflected binary unknown. Q1 update supports lower probability — extrapolation vacated and no new disclosure. Residual uncertainty around CMS timing and appeal pathway. Base case NO by Q3 2026.

Court vacating extrapolation methodology is a material positive. Sampled-only findings unlikely to hit $50M threshold for Medicare segment at 14% of $38.6B revenue. Q1 financials consistent with no pending material clawback. Tail risk around methodology reissuance remains.