ACA MLR Rebates in 2026: $759M and the Cycle Behind the Number

Every year around September, a slice of the health insurance industry writes checks back to its customers. On July 13, 2026, KFF - drawing on data compiled by Mark Farrah Associates from insurers’ state filings - projected that those checks and premium credits will total $759.2 million this year across the individual, small-group, and large-group markets. It is the smallest rebate pool since 2018, and it pushes the running total since the program began in 2012 to roughly $15.1 billion.

The number is small because insurers, on the whole, are no longer wildly profitable - not because the rule stopped biting. The interactive below traces the full 2012–2026 cycle; toggle to the cumulative view to watch the $15.1 billion accumulate, and use the calculator to see exactly when a plan tips into owing a rebate.

What the 80/20 rule requires

The Affordable Care Act’s medical loss ratio provision - the “80/20 rule” - obliges insurers to spend a minimum share of premium dollars on medical care and quality-improvement activities rather than on administration, marketing, and profit. The floor is 80% in the individual and small-group markets and 85% in the large-group market. Fall below it and the insurer must return the difference to enrollees or, for group coverage, to the employer.

Two details make the mechanic less intuitive than it sounds. First, the ratio’s numerator counts not just claims but also documented quality-improvement spending, and the denominator is premium revenue after subtracting taxes and regulatory fees. Second, rebates are calculated on a three-year average MLR, so the 2026 payout reflects 2023, 2024, and 2025 combined. A plan can have a rough single year and still owe nothing - or owe a rebate off strong prior years despite a thin current one. (Governing regulation: 45 CFR Part 158.)

Why 2020 was the peak - and why it is ending

The two-year rebate boom of 2020 and 2021 has a specific, policy-driven origin. In October 2017, the federal government stopped reimbursing insurers for cost-sharing reductions - the discounts they are still legally required to give lower-income marketplace enrollees. Insurers responded by “silver loading”: concentrating the lost revenue into sharply higher silver-plan premiums for 2018.

The result was that individual-market insurers had their most profitable year of the ACA era in 2018, with unusually low loss ratios. Because rebates run on a three-year average, those fat 2018 margins flowed into record rebate checks in 2020 ($2.5 billion) and 2021 ($2.1 billion). In the years since, insurers largely held premiums flat or trimmed them as claims costs caught up, compressing margins and pulling rebates back toward the ~$759M projected for 2026. As KFF puts it, the current environment reflects “margin normalization rather than continued insurer profitability.”

Who gets the money

Rebates paid in 2025 (the most recent completed cycle) reached 5.1 million people in individual-market plans and 3.5 million with employer coverage. The per-person amounts were modest and uneven across markets:

• Individual market: about $233 per person.
• Small group: about $190 per person.
• Large group: about $91 per person.

For people in individual plans, the rebate arrives as a check or a premium credit. For employer coverage, the insurer pays the employer, who is then responsible for passing the enrollees’ share back to them in proportion to what they contributed - a step that is easy to overlook and subject to ERISA plan rules. Rebates are due out by September 30 following the reporting year.

The signal underneath the number

The 2025 simple loss ratio - 93% in the individual market - is the more forward-looking figure. It says insurers spent 93 cents of every premium dollar on claims, well past the 80% floor and a sign of genuinely thin margins amid policy and enrollment uncertainty. A shrinking rebate pool, in other words, is not evidence the 80/20 rule has lost its teeth; it is evidence that pricing and medical spend have converged.

That convergence may not hold. Preliminary 2026 filings show premiums rising at the steepest rate since 2018 - over 20% in the ACA marketplaces. If those increases prove to have out-run actual claims, the three-year averaging that produced the 2020 peak could push rebates back up in 2028 and 2029.

What it means for plans and the teams that run them

MLR is the visible edge of a much larger machine: the relationship between what a plan charges and what it actually spends on care. The rebate is a downstream artifact of that balance - and everything a quality organization does to close care gaps, document quality-improvement activity accurately, and keep members healthy feeds the same numerator the rule measures. Plans that manage HEDIS and Star Ratings performance are managing the medical-spend-and-quality side of the very ratio that determines whether they rebate.

Quality Health’s quality portal gives quality and actuarial teams a live view of measure performance, care gaps, and the documentation that supports both - the operational layer beneath the numbers KFF reports each summer.

Sources

• KFF, “2024 Medical Loss Ratio Rebates” (Matt McGough, Jared Ortaliza, Cynthia Cox), updated July 13, 2026. kff.org
• Paige Minemyer, “Insurers set to pay out $759M in 2026 MLR rebates: KFF,” Fierce Healthcare, July 13, 2026. fiercehealthcare.com
• Mark Farrah Associates, “A Brief Summary of the 2024 Health Insurance MLR and Rebates Results.” markfarrah.com
• KFF, “Explaining Health Care Reform: Medical Loss Ratio (MLR).” kff.org
• CMS, Medical Loss Ratio (45 CFR Part 158) data and resources. cms.gov
• Louise Norris, “What is the medical loss ratio?” healthinsurance.org (payment-year rebate history). healthinsurance.org

Figures are rebates by payment year (the fall payouts go out), the convention used by KFF and Mark Farrah Associates. The annual series sums to KFF’s published cumulative anchors ($14.4B through 2025, $15.1B through 2026); 2026 is a projection. Per-market dollar splits for the 2026 projection were not published at the time of writing.