2027 Medicare Advantage Rate Announcement: County Benchmarks, Risk Scores, and Rebate Mechanics

CMS finalized a 2.48% average Medicare Advantage payment increase for CY 2027, adding $13 billion to program payments (CMS, April 2026). That national figure reaches individual plans through a three-stage translation: county benchmark updates driven by a 5.33% effective growth rate, risk score normalization that applies a 1.058 adjustment factor before payment, and rebate mechanics that convert the benchmark-to-bid margin into member benefits or premium reductions. Each stage varies enough by plan geography and coding documentation that the national average tells bid actuaries almost nothing on its own.

County Benchmark: From National Growth Rate to County-Specific Dollar

The 5.33% effective growth rate CMS finalized in April 2026 is not a flat multiplier that every county receives. It represents the expected change in national fee-for-service per-capita costs as estimated by CMS's Office of the Actuary, incorporating updated fourth-quarter 2025 FFS expenditure data that came in stronger than the January Advance Notice had projected (Georgetown Medicare Policy Initiative, April 2026). From that national rate, CMS derives county-specific benchmarks by multiplying the estimated national FFS per-capita level by a county-level geographic adjustment index. That index measures each county's historical FFS spending relative to the national average, updated using the CY 2020 through CY 2024 claims window that underpins the CY 2027 rebasing.

The practical result: a county whose FFS spending runs at 90% of the national average receives a smaller absolute dollar benchmark than a county at 115%, even when both see the same percentage growth rate applied. For a plan concentrated in a Midwest market where per-member FFS spending is flat, the 5.33% growth rate translates into fewer new benchmark dollars than for an urban Florida or Texas county with high FFS utilization. That asymmetry means the industry-average 2.48% payment increase carries a wide plan-level distribution long before any risk score or rebate considerations enter the calculation.

Quality bonus payments (QBP) add another layer of county-level variation. Contracts earning 4 or more Stars receive a 5 percentage point bonus applied to their applicable county benchmark (CMS, April 2026); those at 3.5 Stars or above receive 3.5 percentage points; contracts below 3.5 receive no bonus. A plan that crossed from 3.5 to 4 Stars between the 2026 and 2027 contract years picks up 1.5 percentage points of additional county benchmark, a gain invisible in the national rate announcement but material in the plan's revenue PMPM projection. At a $1,200 county benchmark, that 1.5-point gain adds $18 per member per month of available revenue before bid construction begins.

CMS finalized a rebasing and re-pricing impact of -0.17% for CY 2027 (CMS Rate Announcement Fact Sheet, April 2026), reflecting the net effect of resetting the geographic adjustment index to the updated FFS claims window. Counties where the refreshed data shows spending below prior index estimates see benchmark reductions on top of whatever the growth rate produces; this offset is small on average but can be concentrated in markets where historical FFS data was overestimated.

Risk Score Translation: Normalization, Coding Intensity, and Encounter Data

A plan does not receive its county benchmark per enrolled member. The county benchmark is the ceiling against which the plan's submitted bid is compared; the plan's actual payment per member is calculated from the benchmark scaled by that member's risk score, divided by the normalization factor that holds the national average risk score at 1.00. For CY 2027, the CMS-HCC normalization factor is 1.058 (CMS Advance Notice, January 2026). A member whose assigned HCC risk score is 1.058 is paid at 100% of the benchmark; a member whose score is 0.90 is paid at 0.90 divided by 1.058, or 85 cents on the benchmark dollar.

Layered on top of normalization is the statutory MA coding pattern difference adjustment. The coding intensity adjustment for CY 2027 remains at 5.9% (CMS, 2027), reflecting the empirically observed tendency of MA contracts to assign more diagnosis codes per member than fee-for-service does for the same population. This adjustment reduces all MA risk scores proportionally before payment. A plan whose population carries an average raw HCC score of 1.25 does not receive 1.25 times the benchmark rate; the effective multiplier is 1.25 divided by 1.058, then further reduced by the 5.9% coding adjustment, producing a payment rate closer to 1.12 times benchmark for that member cohort.

When CMS projects MA payment growth at 4.98% including estimated risk score trend (CMS, April 2026), that figure incorporates the 2.45% annual increase in MA risk scores from coding practices and population aging (CMS Advance Notice, January 2026). The difference between the 2.48% base rate change and the 4.98% total-growth figure is precisely that expected risk score trend: plans with growing, aging, or increasingly coded member populations can expect their effective revenue PMPM to rise faster than the national policy rate, assuming coding documentation holds.

Coding documentation for CY 2027 is where plan-level outcomes diverge most sharply. CMS finalized two diagnosis exclusions that reduce risk scores: first, diagnoses from audio-only telehealth encounters (modifier 93 or FQ) are excluded when no other risk-adjustment-eligible service appears on the encounter or FFS record; second, diagnoses from unlinked chart review records are excluded, meaning chart review diagnoses require a linkable encounter data record to count toward risk calculation. CMS estimated the unlinked chart review exclusion at -1.53% on average across all MA contracts (CMS, April 2026). The agency noted that while the affected beneficiary population is small at roughly 0.04 to 0.05% of all MA enrollees nationally, the practice is "not rare at the contract level," with approximately half of all MA contracts having at least some beneficiaries with unlinked chart review records (CMS CY 2027 Advance Notice, January 2026).

The 1.53% national average obscures a highly skewed distribution. A contract whose retrospective chart review program drove 8 to 10% of total diagnoses through unlinked records faces revenue exposure several times the national average. A contract that embedded diagnosis capture in prospective care delivery visits, where an encounter record always anchors the chart review, may see exposure at or near zero. Bid actuaries must pull contract-level linkage rates from encounter data submissions before finalizing the revenue assumption underlying the bid.

CY 2027 MA Payment Components: Advance Notice vs. Final Rate Announcement
Component	Advance Notice (Jan 2026)	Final Announcement (Apr 2026)	Change
Effective growth rate	+4.97%	+5.33%	+36 bps
Rebasing / re-pricing	TBD	-0.17%	n/a
Star Ratings / QBP	-0.03%	-0.03%	0 bps
Normalization (with model update proposed)	-3.32%	N/A (model retained)	+220 bps
Normalization (without model update)	N/A	-1.12%	n/a
Unlinked CRR / audio-only exclusion	-1.78% (proposed)	-1.53% (finalized)	+25 bps
Net average payment change	+0.09%	+2.48%	+239 bps

Sources: CMS Rate Announcement Fact Sheet (April 2026); CMS Advance Notice Fact Sheet (January 2026); Georgetown Medicare Policy Initiative analysis (April 2026); American Action Forum (February 2026).

Rebate Mechanics and Supplemental Benefit Compression

When a plan's submitted bid falls below its applicable county benchmark, CMS calculates a rebate equal to a fixed percentage of the difference. For contracts at 4.5 Stars or above, 70% of the benchmark-to-bid spread flows back to members; for 4.0 Stars, 65%; for contracts below 4.0 Stars, 50%. Plans must apply the rebate to member-facing benefits: supplemental benefits such as dental, vision, hearing, and over-the-counter allowances; reduced Part B or Part D premiums; reduced cost sharing; or some combination (Social Security Act, Part C bid framework).

The connection between encounter data exclusions and benefit compression runs through this rebate calculation. Expected risk score-adjusted payment PMPM determines the maximum revenue the plan can project for a given membership mix at a given bid level. When encounter data exclusions compress that expected risk score, the plan either absorbs the reduction as margin or passes it through as a lower bid, which reduces the benchmark-to-bid spread and thus the rebate available for supplemental benefits. A plan projecting $1,200 in risk-adjusted revenue PMPM against a $1,350 county benchmark can allocate up to $150 per member per month in rebate funds (at the 65% tier, returning $97.50 to members). If encounter data exclusions reduce projected risk score revenue by 2%, the PMPM projection drops to $1,176 and the available rebate narrows proportionally, directly compressing benefit budgets.

CMS found in 2026 data that supplemental benefit values declined meaningfully year-over-year across the MA market, tracking the encounter data and normalization pressures that accumulated as V28 implementation phased in. The unlinked CRR exclusion effective CY 2027 adds another identifiable mechanism of compression for documentation-heavy plans. Plans that previously used retrospective chart review to offset other payment pressures now face a two-stage problem: the revenue from that coding strategy is gone, and the benefit budgets funded by it must be reconstructed or cut.

Four Conditions That Separate Plan Outcomes From the 2.48% Average

The national 2.48% rate is an enrollment-weighted average across more than 4,500 contracts operating in more than 3,000 counties. Four conditions push individual plan outcomes above or below that average in predictable directions.

County FFS spending relative to national: Plans concentrated in counties where FFS utilization grew faster than the national average over the 2020 to 2024 rebasing window receive larger absolute benchmark increases per the geographic index. High-cost urban and Sun Belt markets with strong FFS growth capture more of the 5.33% growth rate in dollar terms than flat-spending rural counties at the same percentage rate.

Stars tier crossing: The shift from 3.5 to 4.0 Stars moves a contract from the 50% rebate tier to the 65% tier and from a 3.5-point QBP to a 5.0-point QBP. Both changes compound: the higher QBP raises the county benchmark ceiling and the higher rebate percentage returns more of the benchmark-to-bid spread to members, expanding the benefit budget simultaneously. Plans that crossed 4.0 Stars heading into the CY 2027 bid season face a materially different revenue equation than those that stayed at 3.5.

Unlinked CRR and audio-only exposure: Plans with retrospective chart review programs as a primary coding strategy face plan-level revenue reductions that can be two to four times the 1.53% national average, depending on their CRR linkage rate. Those plans will see effective rate changes below the national 2.48% headline, possibly negative in contracts with high CRR reliance, even as the sector-average registers a positive swing.

New membership and plan-switching dynamics: CMS softened the unlinked CRR exclusion for one specific case: beneficiaries who switch from one MA organization to another mid-year may still have diagnoses from unlinked CRRs counted, because those members lack a mechanism for their new plan to submit a linkable encounter data record for services delivered under a prior contract (CMS, April 2026). Plans with high voluntary or plan-exit-driven enrollment growth benefit from this exception. Plans with stable membership and retrospective coding programs receive no such offset.

Bid Revenue Documentation for the June 2026 Submission

The June 2026 bid submission requires actuarial certification of the revenue assumptions embedded in each MA-PD bid. Under the finalized CY 2027 payment rules, that certification must reflect encounter data exclusion assumptions at the contract level, not the national average. A bid that applies the 1.53% industry average CRR reduction to a contract whose linkage rate analysis shows 6% exposure will understate revenue compression and overstate the margin available to cover medical costs or fund supplemental benefits.

The encounter data documentation process also feeds directly into RADV audit exposure. CMS has signaled that RADV will treat encounter data adequacy as a permanent audit dimension (Veradigm, April 2026). A plan that cannot demonstrate linkage between chart review diagnoses and encounter data records during a RADV audit faces retrospective payment adjustments extrapolated to the full membership from a sample, creating a materially larger risk than the prospective bid revenue adjustment would have been. The bid documentation process and the RADV readiness process are now the same work.

What bid actuaries should quantify before the June deadline: the contract-level rate of CRR diagnoses lacking a linkable encounter data record, segmented by condition category and coding vendor; the expected revenue PMPM impact under base and adverse scenarios for that linkage gap; and the sensitivity of the benchmark-to-bid spread to a range of risk score outcomes, recognizing that the 2.45% projected risk score trend (CMS, January 2026) is a national average that individual plan demographics will either amplify or dampen. Plans that frame this analysis in the bid documentation reduce both the actuarial certification risk and the RADV exposure in a single body of work.