MedPAC's September 4 presentation on plan-level MA risk adjustment accuracy shifts the analytical frame from a single aggregate overpayment figure to plan-by-plan examination. The aggregate number produced a program-wide V28 recalibration whose financial effect varied by plan. A plan-level finding, if it leads to policy, would not be symmetric — and it arrives as major MA sponsors tell investors 2027 is a margin-repair year.
For several years, debate over Medicare Advantage risk adjustment has focused on the aggregate difference between MA payments and projected traditional Medicare spending for comparable beneficiaries. MedPAC estimates that Medicare will pay MA plans approximately $76 billion, or 14%, more in 2026 than it would spend if MA enrollees were in traditional Medicare. MedPAC attributes roughly 11 percentage points of that difference to favorable selection and approximately 4 percentage points — about $22 billion — to coding intensity after accounting for CMS's statutory coding-intensity adjustment.
The aggregate number matters, but it has one feature that limits its market impact: it is applied at the program level. CMS's statutory coding-intensity adjustment is program-wide. V28 was a risk-adjustment model recalibration that changed diagnosis coefficients and removed or revised condition categories. Its financial effect was not uniform across plans, beneficiaries, or carriers; it varied with coding mix, morbidity profile, documentation practices, and population composition. It was not a carrier-neutral adjustment in practical effect.
What MedPAC staff presented on September 4 is different.
The session — "Accuracy of Medicare Advantage risk adjustment at the plan level" — examines whether risk adjustment accuracy varies across individual plans rather than across the program as a whole. This is not a minor methodological shift. If MedPAC finds that risk adjustment accuracy varies meaningfully at the plan level, any policy response — a differentiated coding intensity adjustment, a plan-specific reconciliation mechanism, or enhanced audit authority over above-average coders — would affect carriers asymmetrically.
MedPAC staff presented preliminary analysis of risk-standardized 2024 MA plan bids. After adjusting bids for beneficiary risk and geographic differences, staff found substantial remaining variation: bids at the 90th percentile remained roughly 40% above bids at the 10th percentile. That finding does not establish that coding intensity explains the variation, nor does it identify carriers or plans as overpaid. It does establish why the Commission is examining whether the current system more accurately compensates plans for the populations they enroll.
MedPAC discussed possible future directions, including an MA encounter-based model, alternative inputs that reduce reliance on discretionary diagnoses, tiered coding adjustments, use of utilization or survey data, and other model refinements. No formal recommendation was adopted at the September meeting.
A future plan-level risk adjustment policy could affect several dimensions of the MA market.
Payment and bidding. A plan-specific or tiered policy could create asymmetric exposure across plans, unlike a purely program-wide adjustment. If policymakers ultimately adopt plan-specific or tiered adjustments tied to discretionary diagnosis coding, plans with higher measured coding intensity could face greater exposure. MedPAC has not yet made such a recommendation or identified plan-level exposure by carrier.
County and product decisions. A future plan-specific adjustment could become one factor in county-level product and bid decisions, particularly for plans with limited margin. Whether it would materially change exits would depend on the policy design, transition period, bid timing, and each carrier's local economics.
2027 bid assumptions. Plans are currently finalizing their 2027 bids. A plan-level finding suggesting higher exposure for above-average coders is, at minimum, an additional uncertainty in the bid math.
A plan-level risk-adjustment policy could also affect audit exposure and compliance workload. Any connection to Star Ratings would depend on separate rating measures and should not be assumed.
MedPAC's September discussion did not produce a formal recommendation. Staff plans to continue the work in spring 2027, with a plan-level risk-adjustment chapter anticipated for the Commission's June 2027 Report to Congress. Any subsequent CMS or congressional policy action would require a separate process, and no implementation timeline has been established. MedPAC recommendations are advisory, not self-executing.
What moves faster is investor pricing. The market does not wait for final rules. The September 4 presentation is public, MedPAC's track record on coding intensity analysis is well-established, and the direction of the work — toward plan-level examination rather than program-wide averaging — is a visible shift in framing.
Several large MA sponsors have discussed benefit redesign, geographic rationalization, and margin-restoration measures for 2027. Any carrier-specific comparison should be sourced to the company's own earnings materials or SEC filings.
The next concrete milestone is MedPAC's continuing work in spring 2027 and its anticipated June 2027 report chapter. The meeting transcript is available at medpac.gov and will tell you more about the commissioner discussion than any summary can.
This is analysis, not investment advice. Continuum does not make recommendations to buy or sell securities.
Discover how Addie helps health systems, post-acute providers, and payers improve throughput, reduce avoidable days, and deliver better transitions of care.
Get Started
