Humana Stock and the 2027 Medicare Star Ratings: A Payment, Not a Moat

Late on October 8, Medicare published its 2027 star ratings for Medicare Advantage, the private plans people on Medicare can pick instead of the government's own coverage. Each plan gets one to five stars, built from up to 43 measures of care, drug safety and customer service (CMS). A plan with four stars or more gets a bonus from the government. Humana went from about 20% of its members in those plans to over 90%. Its stock rose almost 12% the next day. UnitedHealth, CVS's Aetna and Elevance each lost 8 to 12 percentage points, and their stocks barely moved.

Analysts read it as a turnaround that had been years in the making. J.P. Morgan's Lisa Gill called it a "clear win for a company that has been working toward this outcome for several years" (Healthcare Dive). I don't think the rating works like that. Medicare resets it every October, and Humana's own number has gone from 94% to 25% to 20% and back to the 90s in four rating years. I wouldn't call anything that moves like that a moat. This note asks two things: how much the market paid for Humana's rebound, and what still earns money for each of these insurers in a bad star year.

The reactions don't line up with the ratings. UnitedHealth and Elevance lost about as many points as Aetna did and both rose with the sector. CVS fell 1.9%. Humana's 73-point gain added about $5.4 billion of market value in a day, on about 120 million shares (StockAnalysis). That number is the one I'd test.

Humana's last bad star year gives a rough price for one year of bonus. Its adjusted earnings per share, which strip out one-off items, fell from $17.14 in 2025 to a 2026 guide of at least $9.00, and Humana put the drop down to the star ratings, net of what it did to soften it (Humana FY2025). About $8 a share on about 120 million shares is roughly $1 billion of earnings. That's my estimate, and other things moved too. But on that rough math, the market paid about five years of bonus for one year's rating. That only makes sense if the rating sticks.

The chart is the case against assuming it sticks. UnitedHealth went down, up and down again by about ten points each time. Aetna held for two years and then lost 18 points in two. Humana fell 69 points in one year and gained 73 in another. Across the whole industry, about 71% of members are in 4-star plans for 2027 (CMS). So the total barely changes from year to year, while the share each insurer holds moves a lot.

Three things make the rating unstable, and none of them is about how good a company's health plan is over time.

• The cut-offs move. Medicare sets the score needed for each star level from how all plans did, so a plan can improve and still lose a star. For 2027 it added two drug-safety measures, tripled the weight of two health-outcome survey measures and retired two others (CMS). It also raised many cut-offs, and five-star contracts fell to 15 from 22 (Healthcare Dive).
• Small events carry big weight. Humana says its collapse in the 2025 ratings turned on three test calls to its call centre that Medicare graded as poor. It sued twice and lost both times, and it has appealed. Analysts at Capstone estimated the drop cost Humana a windfall of about $3 billion (Fierce Healthcare).
• The courts can redo the scores. In May 2026 a judge ruled Medicare had used 20 measures in Clover Health's 2026 rating without the authority or the public notice it needed. Medicare recalculated 2026 ratings, upward only (MedCity News), and in July it appealed. Elevance, Alignment and SCAN then sued to get the same treatment, and Elevance put its own gap at $115 million of bonus (Fierce Healthcare). Alignment Healthcare went from 98% of members in 4-star plans to 25% this year and says it may sue again (Healthcare Dive).

The money behind the score is real. Medicare pays each plan up to a per-member ceiling called the benchmark, and the plan bids what it expects care to cost. A plan with four stars or more gets a higher benchmark, and it keeps more of the gap between its bid and the benchmark to spend on extra benefits. The table shows what that adds up to. Double-bonus counties are ones with low traditional Medicare costs and high private-plan enrollment, where the bonus is doubled.

The money doesn't seem to steer members much, though. Humana's 2026 bonus rests on its 2025 ratings, when about one in four members was in a 4-star plan. The table shows what happened.

Members came anyway, in a year when Humana credits new sales and better retention and some rivals were pulling back. I can't separate those causes, so I won't claim stars never steer members. What I can see is that the bad rating years cut Humana's earnings roughly in half while its membership grew. In Humana's case, the stars moved profit much more than they moved customers.

If stars are a payment, the moat has to be somewhere else. On the new Humana and CVS stock pages, the framework rates their Medicare contracts and licences under regulatory lock-in and says in the notes that the star rating isn't a lock. The pillars that matter here are the ones that keep paying when the rating doesn't: owning the pharmacy and care businesses next to the insurer (bundling), the claims data, the provider network and scale.

The matrix shows what each company owns that the rating can't reach, which isn't the same as how the stocks moved. UnitedHealth owns Optum, whose pharmacy benefits, clinics and data business earn money whether a Medicare plan has four stars or three. CVS owns Caremark, one of the three pharmacy benefit managers (the middlemen that run drug coverage for insurers and employers) that together handle about 80% of US prescription claims (Drug Channels), plus about 9,000 pharmacies. Elevance has its Blue Cross licences, which no rival in its states can take. Humana's CenterWell clinics and pharmacy are growing fast, but they're built to serve its own Medicare members, so in my reading they rise and fall with the plan.

Run the callout's test on the cohort and the reactions make more sense. For Humana, one bad star year cut adjusted earnings roughly in half. For CVS, the insurance business is where its recovery lives: Aetna earned $2.43 billion of the $5.63 billion of adjusted operating income across its three segments in Q2 2026, up from $1.31 billion a year earlier (CVS Q2 2026). That's why CVS is the one stock that fell. The rating touched the part investors are paying for, even though Caremark and the stores sit outside it. UnitedHealth and Elevance lost a year of bonus on a slice of much bigger businesses. So the market was pricing how much of each company the rating touches, and I agree with that part. Where I part with it is Humana. It earned one bonus year, worth something like $1 billion of profit in 2028, and the stock was priced as if it had earned several.

The best argument against me starts with Humana's own history. It had 94% of members in 4-star plans in the 2024 ratings, before the three phone calls. On that reading, 2025 and 2026 were the accident, and 2027 is the return to normal. Getting from 20% to over 90% in one year took real work. Analysts at Evercore pointed to gains on drug-plan quality, health-plan quality and hospital readmissions, all things management targeted after the fall (Yahoo Finance). J.P. Morgan had expected 60% to 70%. If a company can push a score that far on purpose, the score is measuring a skill, and a skill can be a moat.

The bonus also compounds. A plan with four stars gets a higher benchmark and keeps more of its rebate. It spends that on lower premiums and richer benefits, which draws members, which spreads its fixed costs. Kaiser has almost all of its members in 4-star plans and gets the biggest bonus per member of any large insurer (KFF). If that loop holds, a good rating feeds itself.

And Humana may be right to stay focused. UnitedHealth's bigger bundle didn't protect it in 2025, when medical costs ran ahead of its pricing. CVS's bundle didn't stop Aetna's costs running ahead of its pricing in 2024. Owning more businesses means more ways to be wrong. A pure Medicare insurer that runs its plans well might earn more on the program than a giant that runs everything adequately.

I take the skill point seriously. It's why the test at the end asks whether Humana keeps its stars, not whether it got them. Humana also changed how its members are spread across contracts for the 2026 enrollment season, which it tied to getting back to the top quarter of ratings (Humana 8-K). That's skill too, but it's skill at working the rulebook, and the rulebook changed again this year. Kaiser is a different kind of company, a non-profit that runs its own hospitals and works with its own medical groups, so I don't think its record transfers to a pure insurer. The same Humana operation lost 69 points on three phone calls, by its own account.

The bundle point is right about 2024 and 2025, and I'm not claiming a bundle prevents pricing mistakes. Those were cost-trend misses, and a bundle doesn't stop an insurer from bidding too low. My claim is narrower: in a bad star year, the businesses outside the plan keep earning. Optum Rx and Caremark fill prescriptions for members of every insurer, whatever stars their own plans got. The compounding argument is real too, but Humana's 2026 runs the other way. It grew members while its stars were at the bottom and paid for it in profit, because the loop runs through margin, and margin is the thing the rating moves.

For Humana, I'd treat the 2027 ratings as one restored bonus year, worth billions of revenue in 2028, and not as proof the moat is back. The things to watch are how many members it keeps through this fall's enrollment period after the 2027 benefit changes its CEO flagged in April (Becker's), and whether its appeal on the 2025 ratings goes anywhere. For UnitedHealth, which reports on October 13, and for CVS, the lower ratings are a 2028 earnings headwind on part of the business. They don't touch what makes either company hard to compete with. Elevance's drop matters least of the four, because Medicare Advantage is a small part of its book.

The same question works outside health insurance. When a business is paid by a score a regulator recomputes each year, such as a quality rating, a reimbursement tier or a bonus formula, I'd ask what it earns when the score goes the wrong way, and look for the moat there.