Why Insurers Are Exiting Medicare Advantage—and What It Reveals About Healthcare’s Risk Engine

The headlines say it all: major insurers are scaling back their Medicare Advantage (MA) footprints for 2026. UnitedHealthcare, Humana, CVS, and others are discontinuing specific plans, exiting certain counties, or pulling out of entire states.
To most people, this sounds like corporate repositioning. But underneath is something deeper: the economics of how insurers actually make money in Medicare Advantage are changing.
And that has ripple effects across the healthcare ecosystem—from hospitals and primary care networks to the analytics platforms trying to make sense of it all.
How Medicare Advantage Works
Medicare Advantage, also called Medicare Part C, lets private insurers deliver Medicare benefits. Instead of the government paying hospitals and doctors directly (as in traditional Medicare), it pays insurers a fixed monthly amount per enrollee—essentially a per-member-per-month capitation.
Each county in America has its own benchmark rate, built off local fee-for-service Medicare spending.
Each patient has a risk score that reflects expected costs based on their diagnoses. The sicker the member, the higher the payment. That score is applied to the benchmark and drives how much CMS sends the plan for that person.
Insurers then take that monthly payment and assume the full financial risk for that person’s healthcare. If they can keep medical costs lower than what they receive from CMS, they profit. If costs exceed the payment, they lose.
The Risk-Scoring Game
The financial engine behind Medicare Advantage is risk adjustment, a process run through the Hierarchical Condition Category (HCC) model. Every diagnosis—from diabetes to heart failure—adds weight to a patient’s risk score.
This creates a powerful incentive: document everything.
To illustrate scale: a relatively healthy retiree with few or no chronic conditions can generate on the order of ~$900 a month in CMS payments to a plan. A member with multiple serious chronic conditions like diabetes, COPD, and congestive heart failure can be worth $1,900+ per month, depending on the county benchmark, quality bonuses, and the member’s full risk profile.
Those numbers aren’t fixed national rates; they’re directional. But the gap (roughly 2x or more) is very real. Multiply that difference across hundreds of thousands of members and accurate—or aggressive—coding can swing billions in revenue.
That’s why regulatory scrutiny has intensified. The Department of Justice and CMS have accused several insurers of inflating risk scores, and CMS has said it is stepping up Medicare Advantage audits to fight overcoding.
At the same time, the model itself is changing. CMS is phasing in an updated risk adjustment model (often referred to as HCC “V28”) through 2024–2026. By 2026, that model is fully in effect and it generally pays less for certain diagnoses than the old model did, which lowers plan revenue per high-risk member compared to prior years.
Managing Care Like a Business
Once those CMS dollars arrive, insurers act more like care-management organizations than traditional payers. They control networks, negotiate rates, require prior authorizations, and build chronic-care programs to manage utilization.
In very simple terms, the margin equation looks like:
Profit = Revenue (from CMS) – Medical Costs – Admin Costs + Quality Bonuses
Plans that hit high Star Ratings (4 out of 5 or better) get bonus payments and higher rebate percentages from CMS. Those dollars are then required to be plowed back into richer member benefits—dental, vision, hearing, gym memberships, reduced copays, lower premiums.
Those extras attract more enrollees, creating a virtuous (and competitive) cycle. That’s a big reason Medicare Advantage enrollment has exploded; more than half of all people on Medicare are now in MA plans as of 2025.
But here’s the tension going into 2026:
- Seniors are using more care again after the pandemic lull—more outpatient visits, more procedures, more home health, more pharmacy spend. That’s driving medical costs per member sharply higher.
- CMS is tightening risk adjustment, which lowers how much “extra” revenue plans get for sicker members.
- And even though CMS is actually projecting an average increase of a little over 5% in Medicare Advantage payments from 2025 to 2026 (roughly $25+ billion more flowing to plans overall), plans argue that their cost trend is rising even faster than that.
So it’s not that CMS is slashing all payments. In fact, CMS finalized an average ~5% payment increase for 2026, which was higher than early 2025 projections.
The problem is: medical spend per senior is growing even faster, and the updated risk model makes it harder to “code your way out” of that gap.
That math is why plans are trimming geographies, narrowing networks, raising prior auth friction, and dialing back some supplemental perks for 2026.
Medicare Advantage vs. Medigap
It helps to contrast MA with Medicare Supplement (Medigap) plans.
Medigap policies don’t replace Medicare—they sit on top of traditional Medicare and cover the gaps, like deductibles and coinsurance. They’re regulated, premium-based, and more actuarial/predictable for the insurer.
Medicare Advantage, by contrast, is full-risk, full-management healthcare. It’s where private payers essentially become mini health systems: they’re responsible for total cost of care, they steer patients in-network, and they run care coordination programs.
When margins shrink in Medicare Advantage, they can’t just casually jack up premiums, because:
- CMS benchmarks and rebate rules constrain pricing,
- and seniors are conditioned to expect $0-premium plans that bundle dental/vision/hearing.
So instead of “charge more,” the lever becomes “redesign care and footprint.”
Why the Market Is Shifting Now
The 2026 environment is forcing hard choices:
- Medical costs are running hot. Seniors are using more services again, and that utilization is hitting plan budgets. Insurers have publicly said this unexpected spike in senior care use is eroding Medicare Advantage margins.
- Risk adjustment reforms are cutting into the upside. CMS’s updated HCC model (V28) is fully phased in by 2026. It generally pays less for certain common chronic conditions than the old model, so plans get less incremental revenue from coding those conditions.
- Revenue growth isn’t keeping up with cost growth. CMS actually finalized an average ~5.06% increase in Medicare Advantage payments from 2025 to 2026 (more than $25 billion systemwide), which on paper sounds generous.
- But plans argue that after you account for higher utilization, pharmacy trend, supplemental benefit expectations, and the new risk scoring, their per-member margins are still getting squeezed.
- Translation: “yes, rates went up — but our costs went up faster.”
- So payers are getting selective.
- • UnitedHealthcare is discontinuing certain Medicare Advantage plans and exiting 100+ counties for 2026, affecting roughly 180,000 members — mostly in lower-margin, often rural markets.
- • Humana is pulling out of entire states and roughly 194 counties, shrinking its footprint to about 46 states and ~85% of U.S. counties, and impacting an estimated ~500,000 members.
- • CVS/Aetna and others are also reducing certain Medicare Advantage and Part D offerings in less profitable regions.
Large insurers will concentrate on profitable counties where the math still works. Smaller regional plans — especially provider-sponsored plans and specialized plans that target complex or dual-eligible populations — may step in to fill some of the gaps the nationals leave behind.
It’s a shakeout—but also a reset.
The story isn’t “Medicare Advantage is dead.” It’s “the easy version of Medicare Advantage is dead.” The high-growth, bonus-funded, $0-premium, dental-plus-vision arms race is colliding with real cost, real utilization, and tighter oversight on risk scoring.
That tells you a lot about where U.S. healthcare is heading: less about selling “coverage,” and more about owning and managing the actual risk.
#MedicareAdvantage #Medicare #CMS #HealthcarePayments #GeospatialAnalytics


