HHS Finalizes OBBBA Marketplace Rules: Six Actuarial Repricing Triggers for 2027

HHS's final rule implementing the One Big Beautiful Bill Act eliminates automatic marketplace reenrollment, shortens open enrollment to five weeks, and narrows special enrollment periods across plan years 2027 and 2028. Each provision filters healthier, price-sensitive enrollees out of a pool insurers say already carries roughly 4 percentage points of subsidy-expiry morbidity from 2026.

Preliminary 2027 filings from 77 insurers across 16 states and the District of Columbia show a median proposed increase of 14%, a quarter above 21%, and none proposing a decrease.

Key Takeaways

Two clocks, not one. The shorter window and the low-income SEP elimination bite in 2027; auto-reenrollment ends for subsidy-eligible consumers in 2028 under Section 71303.
Open enrollment compresses from eleven weeks to six, November 1 to December 15, with state-based marketplaces barred from running past December 31.
Up to 2 million fewer enrollees in 2027 on HHS's own impact analysis, with a further 1 million leaving each year through 2030.
Passive reenrollees are the healthy segment. A book with 40% passive share and a 2-point load on that segment prices 0.8 points book-wide; at 65% passive share the same load prices 1.3.
2026 experience is already elevated by a Wakely-estimated 2.9% to 6.5% morbidity increase, so a fresh full load applied on top of it double counts.

What HHS Finalized, and on Two Different Clocks

CMS administrator Mehmet Oz framed the package around eligibility: "American taxpayers deserve to know their dollars are going only to people who truly qualify." Underneath that sit two implementation tracks with different morbidity timing.

For plan year 2027, effective with open enrollment starting November 1, 2026, HHS shortens the federal window from November 1 through January 15 to November 1 through December 15, and bars state-based marketplaces from extending past December 31. The same rule eliminates the ongoing special enrollment period for consumers below 150% of the federal poverty level, which had allowed monthly signup rather than only during open enrollment or after a qualifying event.

For plan year 2028, Section 71303 ends automatic reenrollment for subsidy-eligible consumers: coverage and premium tax credits lapse unless an enrollee actively confirms eligibility and plan selection each fall. A transitional mechanism requiring auto-reenrolled zero-premium enrollees to pay $5 a month until they confirmed was stayed before implementation.

HHS's own regulatory impact analysis puts the package at up to 2 million fewer marketplace enrollees in 2027, with another 1 million leaving each year through 2030.

Who Each Provision Filters Out

Automatic reenrollment is an invisible cross-subsidy, and removing it removes the enrollees who relied on it. Passive reenrollees skew younger and lower-utilization for a behavioral reason: an enrollee managing a chronic condition has a standing incentive to re-shop each fall around networks and formularies, while a healthy enrollee has none and lets the renewal happen. Adding an active step puts the friction on exactly the healthy, disengaged segment.

That makes the load calibratable from a number carriers can measure in their own book: passive reenrollment share. Applying a 1 to 3 point morbidity load to that segment rather than to the whole book is what keeps the estimate honest. A carrier with 40% passive share and a 2-point segment load prices about 0.8 points book-wide; at 65% passive share the same load prices 1.3.

The shorter window filters a different population, which is why the two cannot be blended. Signups cluster in the final weeks of any enrollment window, and late signups skew toward acute need: a new diagnosis, a new medication, a scheduled procedure. Compressing the window removes days rather than urgency, so the enrollee lost is disproportionately the one who would have signed up in early January, still comparison shopping and not yet facing an acute event. December 15 also lands before many households have received a final December paycheck, a factor the Urban Institute correlated with delayed shopping among lower-income enrollees.

The low-income SEP works the same way from the other end. A near-zero premium plan available in any month attracts opportunistic signups from people not managing an active condition, so removing the monthly on-ramp adds friction that falls hardest on those with least reason to push through it. HHS's modeling of the 2025 predecessor rule projected roughly 293,000 SEP verification issues and a premium effect up to 1% market-wide. That 1% is an average across all SEP types, and low-income SEP elimination concentrates far more heavily on a book with above-average low-income enrollment.

Trigger	Primary effective year	Directional load indicator	Source
1. Auto-reenrollment elimination	2028 (anticipatory in 2027)	1–3 pp on the passive-reenrollment segment	OBBBA Sec. 71303 / CMS, 2026
2. Shortened open enrollment window	2027	Component of KFF's ~4 pp 2027 morbidity estimate	KFF, July 2026
3. Narrowed SEPs (150% FPL elimination)	2027	Up to 1% market-wide (SEP administration bundle)	Health Affairs Forefront, 2025
4. Risk adjustment non-neutralization	2027 ongoing	No automatic offset for market-wide shifts	ACA risk adjustment methodology
5. 2026 baseline compounding	2027 (baseline year 2026)	2.9%–6.5% already embedded in claims baseline	Wakely, April 2026
6. State exchange variance	2027, state-dependent	Multiplier on triggers 1–3, varies by SBM vs. FFE	CMS / Wakely, 2026

Modeling all three against one population double counts the same lost enrollee. Existing passive reenrollees belong to the first, new and re-shopping enrollees to the second, and the low-income subsidy tier to the third.

Risk Adjustment Does Not Neutralize a Market-Wide Shift

The HHS-HCC formula transfers funds among carriers within a state market and is revenue-neutral by construction. That is exactly why it does not absorb a market-wide morbidity shift the way it absorbs a carrier-specific one.

If every carrier loses the same healthier segment to the same provisions in roughly the same proportion, relative risk scores barely move, transfers stay where they were, and the whole cost increase flows through as unhedged premium trend. A carrier that loses more than its share sees its relative score rise and receives a larger transfer, but that compensates for being worse affected than peers, not for the shift itself. The formula's compensating power depends on cross-carrier variance in who lost enrollees, and these provisions apply market-wide.

The baseline compounds the same error from the other direction. Wakely's analysis of January 2026 effectuated enrollment, across roughly 80% of the individual market, found enrollees who paid their January premium carried 10% higher morbidity than those who did not, and put the year's morbidity increase at 2.9% to 6.5% with effectuated enrollment down 17% to 26% by state. Insurer filing narratives attributed roughly 4 points of the 2026 increase to pool deterioration, on top of a 21.7% headline increase against the 2.0% average annual growth filed between 2020 and 2025.

So an actuary taking 2026 claims, already carrying that shift, and adding a fresh OBBBA load measured from a pre-2026 reference point counts the anticipatory portion twice. The defensible split prices the incremental 2027 provisions as a smaller increment on an already elevated base.

Two further asymmetries keep a single national factor from working. State-based marketplaces may set any window inside the December 31 ceiling, so a state exchange can run longer than the federal schedule, and Wakely found state-based exchanges lost less enrollment than federally-facilitated states through the same subsidy-expiry period.

The auto-reenrollment provision reaches full effect in 2028, so a 2027 filing pricing it at steady state overstates the year while one ignoring it leaves the 2028 filing carrying a full-magnitude load on top of whatever adverse selection the other provisions have already produced. The 2017 to 2019 cost-sharing-reduction episode resolved over two to three years partly through carrier exits concentrating the pool. These are permanent statutory changes, so there is no equivalent mean reversion to build in.

Sources