UnitedHealth, Presbyterian and Humana's 2027 Medicare Advantage Exits Trace to County Benchmark Math

UnitedHealthcare's preliminary 2027 Medicare Advantage list covers 34 counties across 12 states and more than 20,000 members, a fraction of the 109 counties and 180,000 members it already cut for the current plan year (Modern Healthcare, August 2026). Presbyterian Health Plan is dropping roughly 30,000 members and Humana 600,000; each retreat traces to the same county-level benchmark and star-rebate math.

Every one of those exits is a bid decision made county by county, not a single corporate switch. A Medicare Advantage plan does not fail nationally; it fails in Otero County, New Mexico, or in a specific rural Nebraska service area, when the local benchmark, the plan's risk-adjusted bid, and its star-rating rebate share stop covering the medical cost the plan expects to pay. Forbes reported August 2, 2026 that the wave of 2027 exit announcements was still escalating, with more carriers expected to disclose county-level cuts before the October 15 start of open enrollment. What the coverage of member counts has not done is walk through why a specific county tips into unprofitability while its neighbor does not. That is a benchmark-quartile, risk-score, and rebate calculation, and it is the same calculation a bid actuary runs every June.

The Three Confirmed 2027 Retreats

UnitedHealthcare's 2027 list, shared with third-party marketing organizations and reported by Modern Healthcare on August 5, spans Minnesota, Nebraska, California, Wyoming, Missouri, Montana, New Hampshire, Wisconsin, Alabama, Arkansas, Kansas, and Kentucky, and the company has been explicit that the list is preliminary and subject to change until CMS's official October 1 plan-finder release. That caveat matters: the carrier's 2026 list also started smaller and grew before finalization. For the plan year already underway, UnitedHealthcare pulled out of 109 counties across 16 states, displacing 180,000 members effective January 2026, in a notice mailed to affected enrollees on October 28, 2025. Bobby Hunter, UnitedHealth Group's CEO of government programs, framed that cut in blunt margin terms at the time: "We need a model that is sustainable and allows us to bring care to folks in those areas in a cost-effective way" (Bobby Hunter, Investing.com/Reuters, October 2025). UnitedHealthcare's total Medicare Advantage enrollment fell by roughly 1.1 million members over 2026 as a result of that and related repricing moves (Modern Healthcare, August 2026), the base its 2027 county-level list is now layered on top of.

Presbyterian Health Plan, the Albuquerque-based system that dominates several New Mexico counties, is discontinuing most of its Medicare Advantage business for 2027 and cutting 150 administrative jobs, a move affecting roughly 30,000 members while it retains its Dual Plus Special Needs Plan covering 13,000 dual-eligible beneficiaries. The Medicare Advantage book lost more than $59 million in 2025, and CEO Rishi Sikka told staff the discontinued plans reflect "the loss profile last year" tied to that business (Rishi Sikka, Healthcare Finance News, June 2026). Presbyterian is a provider-sponsored plan without a national risk pool to smooth its Medicare Advantage losses against other business lines, which is precisely why a single county's benchmark shortfall can force an exit decision that a diversified national carrier might instead absorb for another cycle.

Humana's exit is the largest by member count: plans covering roughly 600,000 members, about 8% of its 7.2 million Medicare Advantage base, will not be offered for 2027, and the company expects to recapture only about 240,000 of them into retained plans at its historical 40% rate (Healthcare Dive, July 2026). Humana's CFO, Celeste Mellet, described the cuts on the July 29 earnings call as trimming "the lower tail of profitability and return," a framing that fits the county-benchmark mechanics below more precisely than a simple member-count reduction does. actuary.info covered the recapture-selection dynamics of that specific move in depth, including why the 360,000 members Humana expects to lose outright are unlikely to select randomly with respect to morbidity (Humana Sheds 600,000 Medicare Advantage Members for 2027, Recaptures Fewer Than Half). This piece takes a different cut through the same 2027 retreat: not which members Humana keeps, but which counties, across all three carriers, stop clearing a bid in the first place.

How a County Benchmark Decides What's Unwritable

Every Medicare Advantage bid is judged against a county-specific benchmark, the maximum CMS will pay for an average-risk enrollee in that county, set as a percentage of local fee-for-service spending. CMS ranks counties nationally by per-capita FFS cost and assigns each to one of four quartiles: the highest-spending quartile gets a benchmark of 95% of FFS, the next two get 100% and 107.5%, and the lowest-spending quartile gets 115% (Congressional Research Service, 2026). A plan that bids below the benchmark keeps a rebate; a plan that bids above it gets nothing and cannot charge the difference back to CMS. Rural counties skew toward the lower end of that FFS-spending distribution because traditional Medicare in sparsely populated areas already runs cheaper per capita, which sounds like it should make them more generously benchmarked, and it does on paper, but a thin enrollee base in those counties means a single high-cost member or a modest utilization swing moves the plan's actual cost experience far more than the benchmark tolerance allows.

CMS rebases those FFS estimates periodically, and the CY2027 rebasing rolled the claims window forward, dropping 2019 data and adding 2024 data to the five-year rolling average used to calculate the benchmark. A county that had been sitting in a favorable quartile can drop a tier when its relative cost position shifts against the rest of the country, and CMS gives plans only a one-year transition before the new, lower benchmark applies in full. Layered on top of that rebasing, the finalized CY2027 rate announcement projects a 2.48% net average payment increase, worth more than $13 billion industry-wide, or 4.98% once estimated risk-score trend is included, down from the 5.06% net increase finalized for CY2026 (CMS, April 2026). A national average increase near 2.5% still means individual counties see negative benchmark movement once their quartile reclassification is applied, and those are precisely the counties where a marginal plan's bid stops clearing.

The benchmark backdrop is generous in aggregate even as it tightens at the margin. MedPAC's most recent data book put average Medicare Advantage benchmarks at 124% of FFS spending nationally, meaning CMS paid plans an estimated $76 billion more in 2026 than the same beneficiaries would have cost in traditional Medicare, a gap actuary.info has tracked in detail as MA's enrollee base has grown (MedPAC's July 2026 data book). That aggregate favorability is exactly why the county-level exceptions stand out: a plan does not walk away from a market where the average national benchmark runs 24 points above FFS unless that specific county's benchmark, risk mix, and bid economics diverge sharply from the national picture.

The Star-Rating Rebate: A Second Lever on the Same Bid

Benchmark movement alone rarely tips a county from viable to unwritable. The second lever is the star-rating rebate share, the percentage of the gap between a county's benchmark and a plan's bid that the plan actually gets to keep and spend on supplemental benefits. Statute sets that share at 50% for contracts rated below 3.5 stars, 65% for 3.5 to 4.5 stars, and 70% for 4.5 stars and above (KFF, 2026). A plan that drops from the 65% tier to the 50% tier does not just lose its quality bonus; it loses 15 percentage points of rebate on every dollar of benchmark-bid spread in that county, a much larger dollar impact in most markets than the bonus itself. That rebate compression is happening at scale. CMS's 2026 star ratings pushed the 4-star cut point higher across roughly 60% of measures, and the share of Medicare Advantage membership in 4-star-or-better contracts slipped to 63.5% from 64.1% the year before, even as the overall average rating held roughly flat (Healthcare Dive, 2026). actuary.info's own coverage of the quality bonus program found the industry's bonus-qualified enrollee share falling to 68% from 75%, the lowest since 2018, even as total bonus spending rose to $13.4 billion from $12.7 billion, a sign that fewer contracts are clearing the bar even as the dollars at stake for the ones that do keep climbing (Medicare Advantage's Quality Bonus Hits $13.4 Billion While the Qualified Base Shrinks to 68%). Elevance's ongoing litigation over $115 million in disputed 2026 bonus dollars, tied to CMS dropping quality measures unevenly across competing contracts, illustrates how directly a single measure removal can move a plan's rebate tier and, from there, its bid economics in every county that contract serves (Elevance Sues CMS Over $115M in Medicare Advantage Star Ratings).

The mechanism that ties star tier to a specific county exit runs through supplemental benefits. A shrinking rebate share means fewer dollars available to fund dental, vision, hearing, over-the-counter allowances, and reduced cost-sharing, the features that make a Medicare Advantage plan competitive against a neighboring carrier's offering in the same county. In a thick urban market, a plan can absorb a rebate cut and still field a competitive benefit package because its enrollee base is large enough to spread fixed administrative costs and smooth utilization variance. In a county with a few thousand members, the same rebate cut forces a choice between a benefit package too thin to retain enrollment or a bid too rich to clear the benchmark, and that is the point at which a plan exits rather than compete on economics it cannot win.

Mechanism	Value / Rule	Source
County benchmark quartiles (% of local FFS)	95% / 100% / 107.5% / 115%	Congressional Research Service, 2026
CY2027 benchmark rebasing window	2020–2024 claims (drops 2019)	CMS, April 2026
CY2027 net average payment change	2.48% (4.98% incl. risk-score trend)	CMS, April 2026
CY2026 net average payment change (comparison)	5.06%	CMS, 2026
Star-rating rebate share, <3.5 / 3.5–4.5 / ≥4.5 stars	50% / 65% / 70% of benchmark-bid gap	KFF, 2026
MA membership in 4+ star contracts, 2026 vs. 2025	63.5% vs. 64.1%	Healthcare Dive, 2026
National MA benchmark average, 2026	124% of FFS ($76B above FFS spend)	MedPAC, cited by actuary.info

Where the V28 Risk Model Adds Pressure

A third mechanism compounds the first two: the risk-adjustment model that converts a county's raw enrollee mix into the risk score a plan's bid is actually paid against. The CMS-HCC V28 model reached full 100% phase-in for the current plan year, having run one-third weighted in 2024 and two-thirds in 2025, and CMS chose not to update its underlying calibration data for CY2027, retaining coefficients built on 2018 diagnoses and 2019 expenditures rather than refreshing them, a decision CMS said was meant to "allow the MA market more time to adjust" after the recent transition. actuary.info's earlier analysis of that rate reversal walked through the plan-level revenue exposure the decision to hold the old calibration created heading into the June 2026 bid deadline (CMS 2027 MA Rate Reversal: What 2.48% Means for Plan Actuaries).

CMS did finalize two narrower changes for CY2027: diagnoses sourced from unlinked chart reviews and from audio-only telehealth encounters no longer count toward a member's risk score. Both exclusions bite hardest in exactly the counties most likely to appear on an exit list. Rural and low-provider-density counties lean more heavily on telehealth, including audio-only visits, to close annual wellness and diagnosis-capture gaps than urban markets with dense in-person provider networks, and those same counties are disproportionately the ones sitting in the lower-benchmark quartiles where every risk-score point matters more to the bid. actuary.info has separately documented this as a second, distinct 2027 risk-score compression, layered on top of the benchmark and rebate pressure described above, that plan actuaries need to model independently rather than blend into a single trend assumption (MA Crosses 51.6% Penetration: What the Majority Threshold Means for Plan Actuaries).

Portfolio Management, Not a Blanket Retreat

None of the three carriers described their 2027 moves as an industry-wide pullback, and the data supports a more selective reading. Oliver Wyman's analysis of the 2027 bid cycle describes carriers moving into a more aggressive "portfolio management posture," trimming counties, degrading select benefits, and curating networks rather than exiting Medicare Advantage broadly, while scaled plans concentrating in high-benchmark urban markets can realize 3% to 5% efficiency gains relative to a more geographically dispersed footprint (Oliver Wyman, March 2026). That framing matches what the three confirmed 2027 exits actually look like on a map: UnitedHealthcare's preliminary list is concentrated in Minnesota, Nebraska, and other lower-density Midwest and Mountain West counties; Presbyterian's exit is essentially a single-state retreat from a market it could not subsidize internally; and even Humana, cutting the largest absolute number of members, is retaining plans with stronger value-based care penetration while shedding the "lower tail" Mellet described. The carriers are not leaving Medicare Advantage. They are systematically exiting the specific counties where the benchmark, rebate, and risk-score math no longer works, while defending or expanding in counties where it does.

The Bid-Pricing and Risk-Pool Implications

For a bid actuary pricing 2027 or 2028 Medicare Advantage business, the county-level mechanics above translate into three concrete modeling questions rather than one national trend line. First, pricing discipline now has to be run at the quartile level rather than the market level: a county's benchmark trajectory following its five-year rebasing window matters more to bid viability than the industry-wide 2.48% headline, and a plan holding share in a county whose quartile just dropped is pricing against a shrinking ceiling regardless of what the national number implies. Second, benefit richness in marginal counties is now directly constrained by star tier in a way that compounds rather than offsets a weak benchmark; a plan cannot fund its way out of a rebate-share downgrade with benchmark strength alone if the county was already near the exit threshold. Third, the risk-pool churn created by all three 2027 exits, plus the 180,000 members UnitedHealthcare already displaced for 2026, adds up to a population that will re-shop or get auto-crosswalked into competitors' plans well before those competitors' 2027 bids were finalized, the same selection-driven morbidity mismatch actuary.info identified in the broader 2026 disenrollment wave that forced 2.9 million MA enrollees to switch plans ahead of carriers filing bids on a population that had not yet arrived in their books.

That last point is the one most likely to be underweighted in a standard bid file. A plan that gains enrollment because a competitor exited a shared county is not simply picking up market share; it is inheriting a risk pool shaped by which members were engaged enough to shop during open enrollment versus which were passively crosswalked, the identical selection dynamic that determines whether Humana's own 240,000 recaptured members skew healthier or sicker than the 360,000 it expects to lose outright. Multiplied across three carriers' county exits landing in the same October 15 to December 7 enrollment window, the 2027 Medicare Advantage risk pool in the affected counties will not resemble the 2026 experience data most bids are currently built on, a gap that will not close until well into 2027 claims runout.