Your Growth Plan Books Revenue the Payment System Doesn't Pay
A new Medicare Advantage member arrives without twelve months of claims history, so CMS pays the plan on demographics alone — no coding-driven revenue for a year. The acquisition cost, meanwhile, is paid up front: commissions past $694, administrative payments a court just ruled CMS cannot cap, all-in broker compensation reported above $1,300 before the plan's own marketing. Revenue lags cost by construction. Six things from building a growth instrument around that lag.
1. The clawback protects the broker relationship, not the plan. Rapid-disenrollment recovery returns the commission when a member leaves within three months. The demographic-only revenue year runs twelve. A member who leaves in month eight cost full acquisition, paid discounted revenue, and returned nothing. Months four through twelve are the plan's alone.
2. Nobody has published the payback. The churn evidence is solid. The acquisition-cost evidence is partial. Nothing public joins them: there is no independent lifetime-value or payback-period study for MA membership — none. Every LTV in every growth deck is proprietary modeling that has never met a counterfactual. Ask for the survival curve and the year-one haircut before believing the integral.
3. The growth engine is the churn engine. Special-needs plans carried 85% of net MA enrollment growth in 2026 — and high-need dual-eligibles are precisely where documented exits run highest: 14.8% leave MA for traditional Medicare annually, reaching 42.8% in low-star plans, with star rating the strongest exit predictor in the peer-reviewed record. A blended churn assumption from the legacy book, applied to an SNP-heavy growth plan, forecasts a book nobody is buying.
4. Sign-ups are not members. Marketplace: 96% of plan selections effectuated into paying members — with enhanced subsidies. Those expired in December 2025; 2026 sign-ups fell, net premiums rose 58%, and projections dropped as low as 17.5 million effectuated. Medicaid: 10.3% of enrollees churn out and back in normal times, before any unwinding — which disenrolled 25 million, 69% of them for paperwork rather than eligibility.
5. The rebid is the risk term everyone deletes. States rebid Medicaid managed-care contracts perhaps once or twice a decade. The growth risk isn't annual attrition — it's a single procurement event that can remove the entire book. A Medicaid growth projection without the rebid calendar has deleted its largest risk term.
6. The age-in share is an assumption in a fact's costume. No current public decomposition splits MA growth into age-ins versus switchers — the last one uses 2016 data. With penetration at 55%, the switcher pool thins every year. State the source or state the assumption.
We built these into an instrument: the survival-weighted payback sketch (every month of margin weighted by the probability the member is still there), the mix-versus-churn check, and the program-specific machinery — election periods, effectuation, redeterminations, rebids — every figure cited.
Growth that ignores its own survival curve is acquisition spend buying churn.
What did last year's new-member cohort actually earn in months one through twelve — and who in your building has looked?