Humana Sheds 600,000 Medicare Advantage Members for 2027, Recaptures Fewer Than Half

Humana will exit Medicare Advantage plans covering roughly 600,000 members for the 2027 plan year, about 8% of its 7.2 million MA lives, and expects to recapture only about 240,000 of them into retained plans (Healthcare Dive, July 2026). That leaves roughly 360,000 members Humana expects to lose outright, and the actuarial question those exits raise is not how many people leave but which ones stay.

Humana's CFO, Celeste Mellet, framed the exits on the July 29 second-quarter call as a deliberate cut of "the lower tail of profitability and return," prioritizing plans with greater value-based care penetration for retention (Celeste Mellet, Healthcare Dive, July 2026). The stated goal is a sustainable pretax Medicare Advantage margin of at least 3% by 2028, and the company is running that goal against a Q2 2026 pretax margin of just 1.8%, a level that has fallen from 6.7% in 2020 and shows how far the recovery still has to travel. The quarter's other numbers were genuinely strong: revenue of $40.9 billion, up 26% year over year, and net income of $694 million, up 27% (CNBC, July 2026). None of that changes what a 600,000-member exit does to the morbidity mix of the members who remain.

The Recapture Number That Actually Drives the Model

Humana's own precedent sets the benchmark it is now repeating. The insurer recaptured just over 40% of the roughly 500,000 members displaced by its 2025 plan exits into other Humana plans, and it is guiding to a similar share for 2027 (Healthcare Dive, July 2026). Applied to 600,000 exiting members, a 40% recapture rate produces the roughly 240,000 figure both Healthcare Dive and Becker's have reported. The remaining 360,000 members do not simply vanish from the Medicare rolls; they scatter into competitors' plans, traditional Medicare, or a Humana plan they were not defaulted into, and each of those destinations pulls a slightly different demographic than the pool Humana designed its 2027 bids around.

Recapture in Medicare Advantage is not a random draw from the exiting population. CMS's own default-enrollment and crosswalk mechanics tend to auto-assign a share of displaced members into a company's nearest comparable plan unless the member actively shops, which means the members most likely to be recaptured are disproportionately those least engaged with plan selection, often the same group carrying higher chronic-condition burden and lower health literacy. Conversely, the members most likely to actively shop, compare star ratings, and switch carriers during the October 15 to December 7, 2026 annual enrollment period tend to skew healthier and more price-sensitive. If that selection pattern holds, Humana's 40% recapture is not simply capturing 40% of the exiting risk pool's average morbidity; it may be capturing a segment tilted toward higher, not lower, expected cost.

Where the Exits Are Concentrated

The exits are weighted toward lower-rated plans, but Humana executives were explicit that star ratings were not the primary driver. Most of the affected plans are rated 3.5 stars or below for the 2027 bonus year, though the company said the decisions rested on profitability and value-based care penetration rather than the rating itself (Healthcare Dive, July 2026). That distinction matters because it decouples the exit decision from the metric CMS uses to allocate quality bonus payments, meaning a plan can be jettisoned for margin reasons even where its star trajectory was improving.

The star-rating backdrop is itself a separate deterioration Humana has been managing since 2024. Only 20% of Humana's Medicare Advantage members are in plans rated 4 stars or higher for the 2026 bonus year, a collapse from 94% of members in 4-star-or-better plans as recently as 2024 (Fierce Healthcare, 2026). CEO Jim Rechtin has said the company is targeting top-quartile star performance again by the 2028 bonus year but has stopped short of guaranteeing it: "We don't know industry thresholds. So while we feel good about our substantial progress, we cannot guarantee an outcome" (Jim Rechtin, Healthcare Dive, July 2026). A prolonged run of sub-4-star ratings compounds the margin problem the exits are meant to solve, since quality bonus payments flow only to 4-star-and-above contracts and low-rated plans generate less bonus revenue to offset benefit costs in the first place, part of what pushed Humana's Q2 2026 insurance segment benefit ratio to 91.2%, up from 89.9% a year earlier even as premium revenue grew (Investing.com, July 2026).

Why the Stayers, Not the Leavers, Set 2027 Pricing

Every actuary pricing a 2027 MA bid off 2025 or early-2026 experience data is implicitly pricing a risk pool that will not exist by the time the bid takes effect. actuary.info has previously documented how the broader 2026 exit wave forced 2.9 million MA enrollees to switch plans while carriers filed 2027 bids before that reshuffled population had actually arrived in their books, against a finalized 2.48% effective CMS rate increase for 2027 that trails an 8% to 10% medical trend assumption most actuaries are carrying (actuary.info's analysis of the 2026 disenrollment wave, 2026). Humana's 600,000-member exit is the single largest carrier-specific instance of that dynamic reported so far this cycle, and it isolates the mechanism cleanly because it is one company's book, not an industry-wide reshuffle blurred across dozens of carriers.

The mechanism runs through selection, not through the exit itself. A member who is auto-crosswalked or defaults into a retained plan without actively shopping contributes their existing utilization pattern and diagnosis history to the surviving risk pool with minimal friction. A member who actively compares options during open enrollment, researches competitor benefits, and switches carriers is, by construction, engaged enough to act on new information, a trait that correlates with better self-management of chronic conditions and, empirically, lower per-member cost. If Humana's 240,000 recaptured members lean toward the passive, higher-morbidity end of that spectrum while the 360,000 who leave lean toward the active, lower-morbidity end, the retained book's average risk score rises even though total membership fell by only 8%. That is a materially different pricing input than a flat 8% membership reduction with unchanged average morbidity, and it is not one CMS's rate notice or Humana's own guidance quantifies directly; it has to be inferred from the recapture mechanics and validated against early 2027 enrollment data once it becomes available.

Metric	Value	Source
Total MA membership exiting for 2027	~600,000 (8% of 7.2M base)	Healthcare Dive, July 2026
Expected recapture into retained plans	~240,000 (40% rate)	Healthcare Dive, July 2026
Net member loss	~360,000	Derived
2025 exit recapture rate (precedent)	~40% of ~500,000	Healthcare Dive, July 2026
Members in 4+ star plans, 2026 vs. 2024	20% vs. 94%	Fierce Healthcare, 2026
Q2 2026 insurance segment benefit ratio	91.2%, up from 89.9%	Investing.com transcript, July 2026
CY2027 CMS effective growth rate / net rate change	5.33% / 2.48%	CMS final rate notice, April 2026

Coding Intensity and the Risk-Adjustment Backdrop

The exits are landing at the same moment the industry's coding-intensity cushion is shrinking under a full V28 risk-model implementation. The Medicare Payment Advisory Commission estimates that Medicare Advantage plans in 2026 were paid $76 billion more than traditional Medicare would have spent on the same beneficiaries, with $28 billion of that gap attributable to coding intensity, the tendency of MA plans to document diagnoses more thoroughly than fee-for-service providers do for the same clinical picture (KFF, 2026, citing MedPAC). That coding-intensity effect has declined sharply as the new risk model phased in fully, from adding an estimated 10% to MA payments in 2022 to about 4% in 2026, the first year the V28 model has run without a blended transition (KFF, 2026, citing MedPAC).

A shrinking coding-intensity cushion raises the stakes on Humana's member-mix problem rather than lowering them. When coding intensity was contributing a full 10 percentage points of payment uplift, a carrier had more room to absorb an unfavorable shift in its retained population's true morbidity without the risk-adjustment mechanism fully offsetting it in the following payment year. At roughly 4 percentage points of uplift, that buffer is thinner, and CMS finalized additional restrictions for 2027 excluding diagnoses sourced from unlinked chart reviews and audio-only telehealth visits from risk score calculations, tightening the documentation pathways plans have historically used to capture HCC codes (Crowell & Moring, citing CMS CY2027 final rate notice, April 2026). If Humana's post-exit risk pool skews toward higher true morbidity, as the recapture-selection dynamic suggests it might, the plan will need those diagnoses captured cleanly under a stricter coding regime to have its risk-adjusted revenue reflect the population it is actually covering, a harder task than it would have been under the pre-V28 model.

Star Ratings, Bonus Payments, and a Second Feedback Loop

The exits interact with star-rating dynamics in a way that can either stabilize or destabilize Humana's 2028 quality trajectory, depending on which members actually leave. Exiting predominantly sub-4-star plans should mechanically raise the enrollment-weighted average star rating of Humana's remaining book, a genuine improvement independent of any change in clinical performance. But star ratings themselves are partly a function of member mix: measures like medication adherence and controlling blood pressure are easier to hit in a healthier, more engaged population, and harder in a population with more comorbidities and lower engagement. If the recapture-driven morbidity shift described above is real, it works against the mechanical star-rating improvement the exits are supposed to deliver, a tension embedded in Rechtin's own hedge about not being able to "guarantee an outcome" on star performance (Jim Rechtin, Healthcare Dive, July 2026).

This is also not a Humana-specific phenomenon in isolation. Roughly 1 in 10 Medicare Advantage beneficiaries enrolled in HMO or PPO plans nationally are expected to be forced to disenroll in 2026 because their plan exited the market, a rate MedPAC's public comment letter attributes to a combination of tightening MA payment growth and unanticipated utilization increases across the sector (Health Affairs Scholar, 2026, citing MedPAC). UnitedHealthcare has run a parallel, larger-scale version of the same trade, shedding 1.3 million MA members in 2026 to push its medical cost ratio from near 90% down to 83.9%, a margin-over-volume choice actuary.info has covered in detail (UnitedHealthcare's 1.3 million-member MA repricing). Humana's 600,000-member move is smaller in absolute count but proportionally similar, roughly 8% of its base against UnitedHealthcare's larger cut, and both carriers are converging on the same conclusion: shrinking the book deliberately is now a faster path to margin recovery than growing through it.

What the Bid Actuary Cannot Yet See

The practical constraint facing anyone pricing Humana's 2028 MA bids this cycle is timing. The 600,000-member exit takes effect for the 2027 plan year, with affected members notified during the October 15 to December 7, 2026 annual enrollment period, but the actual recapture outcome, who follows into a retained plan and who does not, will not be observable in claims experience until well into 2027. Humana's 2028 bid, due to CMS in the first half of 2027 under the normal Medicare Advantage bid cycle, will therefore need to be built substantially on assumption rather than on fully mature post-exit experience, the same structural lag actuary.info's earlier disenrollment analysis identified in the 2026 exit wave. The company's public guidance, a 40% recapture rate carried over from 2025 precedent, is a reasonable starting assumption, but it is a population-average rate applied to a exit tranche that, by the mechanics described above, may not select randomly with respect to morbidity.

A bid actuary working this book has a narrower set of levers than the headline recapture percentage suggests. Segmenting the recapture assumption by whether a member's transition into a retained plan was passive (crosswalk or default) versus active (independent shopping and enrollment) would better isolate the morbidity difference between the two populations than a single blended 40% figure. Comparably, watching early first-quarter 2027 utilization data against the prior book's baseline, rather than waiting for a full calendar year of experience, would surface a morbidity shift faster than the annual bid cycle alone allows. Neither approach eliminates the uncertainty Humana's own exit decision has introduced into its 2027 and 2028 numbers, but both narrow the range around a recapture assumption that, as filed, treats a selection-driven population shift as if it were a simple headcount reduction.