KFF's July 2026 tally puts this year's Affordable Care Act rebate at $759.2 million across the individual, small-group and large-group markets, down from $1.6 billion in 2025 and $958 million in 2024, the smallest payout since 2018's $706.7 million (KFF, July 2026).
That is not a consumer refund story. A rebate is owed when premium was priced rich against realized claims, so a smaller total means fewer carriers cleared the medical loss ratio floor by a wide margin.
Key Takeaways
- $759.2 million is a three-year average of 2023, 2024 and 2025 experience, not a read on 2026 pricing. The checks go to people covered in 2025, the most recent year in the window.
- The individual market carries close to 72% of rebate dollars while covering about 11% of the population, at $233 per person against $190 in small group and $91 in large group.
- Blocks under 1,000 life-years are exempt entirely, and blocks between 1,000 and 75,000 get an upward MLR adjustment of 1.2 to 8.3 points, so identical experience produces different obligations.
- Rebates peaked at roughly $2.46 billion in 2020 and $2.01 billion in 2021 when insurers priced against pandemic-depressed claims; this is the first cycle where the window has fully rotated past that distortion.
- Preliminary 2027 filings from 77 insurers show a median increase of 14%, with 20 above 20%, which is the opposite direction from the rebate.
The Number Describes a Pricing Cycle That Already Closed
CMS requires individual and small-group insurers to spend at least 80% of premium on medical claims and quality improvement, and large-group insurers 85%, with any shortfall in a given state and market segment owed back. The calculation is never a single year. CMS averages three consecutive years before determining whether a rebate is due, so 2026's $759 million reflects 2023, 2024 and 2025 financial data.
The lag is deliberate smoothing. A single bad claims year or a utilization swing would otherwise whip an insurer's obligation from zero to material and back annually, which is the volatility a credibility-weighted calculation exists to dampen. The tradeoff is that the published figure always describes an already-completed pricing cycle.
The historical series shows the smoothing at work. Rebates peaked near $2.46 billion in 2020 and $2.01 billion in 2021, when insurers overestimated pandemic-depressed claims and priced 2021 too rich, then fell to $1.03 billion in 2022, $947 million in 2023 and $958 million in 2024 as the windows absorbed that overpricing (healthinsurance.org). The 2025 spike to $1.6 billion carried residual excess margin.
This year's figure is the first cycle where the rolling average has fully rotated past the CSR-era distortions into what KFF calls "margin normalization rather than continued insurer profitability stemming from the earlier CSR-related premium increases." Reading the total as a statement about how insurers are pricing today, rather than how they priced from 2023 through 2025, inverts what the formula measures.
Three Variables Stack Before Anyone Owes Anything
Rebates concentrate unevenly. The individual market accounts for close to 72% of rebate dollars while covering about 11% of the population, at average per-person rebates of $233 against $190 in small group and $91 in large group, across 5.1 million individual-market recipients and 3.5 million with employer coverage. Self-funded and ASO arrangements sit outside MLR entirely and dilute the group figures; fully-insured individual policies carry no such exemption.
Market-level ratios do not resolve the pattern. KFF's 2025 data puts the individual market's aggregate simple loss ratio at 93%, against 87% in small group and 91% in large group, all above their floors on a market-wide basis. But MLR is computed separately for each legal entity, in each state, in each market segment, so a market clearing the floor on average can still contain carrier-state blocks well short of it. Those specific blocks generate the obligation.
The numerator is the second variable. What CMS compares against the floor is incurred claims plus documented quality improvement expenditures, divided by earned premium net of certain taxes and licensing fees, reported on each insurer's Supplemental Health Care Exhibit. A carrier with a 76% simple loss ratio in a state block is not automatically 4 points short of an 80% floor; with 3 points of substantiated QI spend it is 1 point short, and the rebate shrinks proportionally. QI credit is audited at the filer level and varies enormously: a national insurer with established care management typically substantiates more than a regional plan building it.
| Block size (life-years) | Credibility status | MLR treatment |
|---|---|---|
| Under 1,000 | Not credible | Exempt from rebate requirement |
| 1,000 to 75,000 | Partially credible | Reported MLR adjusted upward by 1.2 to 8.3 points, sliding by size |
| Above 75,000 | Fully credible | No adjustment; reported MLR compared directly to the 80/85 floor |
Credibility is the third. Insurers covering fewer than 1,000 life-years in a state and market are exempt because the experience is not statistically meaningful. Blocks between 1,000 and 75,000 are partially credible and receive an upward adjustment of 1.2 to 8.3 percentage points, largest for the smallest blocks (KFF's MLR explainer).
The effect is mechanical. A block of 5,000 life-years at a 76% simple loss ratio might take a 6-point adjustment to 82% and owe nothing. A fully credible national carrier reporting the same 76% in the same state receives no adjustment and owes the full 4-point shortfall. The rule is compensating for small-sample noise rather than for size, and the practical incidence still skews toward larger books.
The obligation is also carried in real time rather than discovered at the end. Insurers accrue an estimated rebate liability during the plan year and update it quarterly as claims develop, so a carrier whose interim development assumptions run optimistic faces the same adverse true-up dynamic a casualty reserving actuary would recognize: a liability that looked adequately provided at an early valuation date turns out understated once the experience period matures.
The Rebate and the Filing Point Opposite Directions
Marketplace insurers pushed through average increases of roughly 20% to 25.5% entering 2026, the steepest since 2018, and preliminary 2027 filings from 77 insurers across 16 states and the District of Columbia show a median proposed increase of 14%, with 20 requesting above 20% (KFF).
Insurers attribute roughly four points of the 2026 increase, and four more expected in 2027, to a sicker pool left after the enhanced premium tax credits expired and healthier enrollees exited, a shift that already pushed average post-subsidy premium payments up 58% this year (Peterson-KFF).
The two are not contradictory. They describe the same dynamic from two points in the cycle. The $759 million reflects a period when insurers were still working through post-pandemic cost catch-up and had not absorbed the subsidy-cliff shift that began with 2026 enrollment. The 2027 filings reflect a current-year view of a materially worse pool, priced defensively because several thin-rebate years left little cushion.
That is what makes the low rebate total the harder read. A market running close to its floor has less room to absorb a fresh adverse-selection shock without either raising rates sharply or eroding margin, and the 14% median ask is that arithmetic resolving. The buffer between priced premium and realized cost has narrowed enough that the next adverse development, whether from pool deterioration or a slower-than-modeled contracting cycle, surfaces as a missed rate filing rather than a rebate check.
Credibility compounds it for smaller books specifically. A partially credible block that has avoided rebates on a favorable adjustment, rather than on claims experience that genuinely cleared the floor, is carrying statistical forgiveness that growth removes. A clean multi-year rebate history on a small block splits into earned economics and credibility-table arithmetic, and only one of those halves survives the block getting larger.
Further Reading
- New York Cut Filed Health Rates Nearly in Half. The Gap Is an Adequacy Problem: How a regulator-cut rate can squeeze a carrier's MLR from both directions, compressing the margin cushion the 82% New York floor already limits.
- UnitedHealth and Elevance Split the Q2 2026 Managed-Care Recovery: How current-year medical loss ratio movement at two national carriers diverged in the same reporting window, a useful contrast to the three-year rolling MLR that drives ACA rebates.
- Molina's 92.7% Medicaid MLR and a Trough Call the Acuity Math Hasn't Confirmed: A Medicaid-specific MLR read that shows the same margin-normalization dynamic playing out on the government side of the book.
- UNH's 1.3M Member Exit and the Margin Recovery Calculus: How deliberate membership shedding, rather than rebate exposure, has been the dominant margin lever at the largest national carrier.
- The P&C Soft-Market Reserve Adequacy Playbook: A parallel margin-floor problem in property-casualty pricing, for actuaries comparing how different lines respond to compressed underwriting cushion.
- Social Inflation and the Casualty Loss Development Factor Adjustment: Another case where a lagging, multi-year actuarial calculation understates a faster-moving current-year cost trend.
Sources
- KFF: 2026 Medical Loss Ratio Rebates (July 2026)
- Fierce Healthcare: Insurers Set to Pay Out $759M in 2026 MLR Rebates: KFF (July 2026)
- InsuranceNewsNet: 2026 Medical Loss Ratio Rebates (July 2026)
- healthinsurance.org: Billions in ACA Rebates Show the 80/20 Rule's Impact (2026)
- KFF: Explaining Health Care Reform: Medical Loss Ratio (MLR)
- KFF: In Preliminary Rate Filings, ACA Marketplace Insurers Largely Propose Double-Digit Premium Increase for 2027 (July 2026)
- Peterson-KFF Health System Tracker: How Much and Why ACA Marketplace Premiums Are Going Up in 2027 (July 2026)