From tracking ACA rate filings across multiple cycles, the 2027 season introduces a pricing challenge that has no clean precedent: actuaries must set morbidity assumptions using partial-year enrollment data from a market that is still mid-contraction. The enhanced premium tax credits expired at year-end 2025, and the consequences are now measurable. Wakely Consulting’s “Who Paid, and Who Stayed?” white paper, published April 15, 2026, provides the first carrier-level quantification of what happened to the ACA individual market risk pool when subsidies disappeared. Drawing on the Wakely National Risk Adjustment Reporting (WNRAR) dataset, which covers approximately 80% of the ACA-compliant individual market through data from more than 75 carriers across over 30 states, the analysis shows a market in rapid compositional flux.

This article walks through the actuarial methodology for converting Wakely’s emerging enrollment and effectuation signals into a defensible morbidity adjustment factor for 2027 rate filings, covering risk score recalibration, metal-tier migration effects, risk adjustment transfer volatility, and the professional judgment required under ASOP No. 8 when Q1 2026 experience data conflicts with a CY 2025 base period.

86%
January 2026 First-Month Premium Payment Rate
2.9–6.5%
Estimated Morbidity Increase Range
17–26%
Projected Effective Enrollment Decline

The Effectuation Gap as an Adverse Selection Signal

Plan selections during the 2026 open enrollment period fell by over one million to approximately 23.1 million, the sharpest single-year decline since ACA marketplaces launched (KFF). But plan selections measure only who chose a plan or was automatically renewed, not who paid. Wakely’s WNRAR data reveals the deeper story: only 86% of January 2026 enrollees paid their first-month premium, compared to historical effectuation norms above 92% and as high as 96% during the enhanced subsidy era from 2022 through 2025.

That six-to-ten-point effectuation gap is not random attrition. The members who failed to pay their first premium carried, on aggregate, approximately 10% lower morbidity than those who did pay (Wakely). This is the textbook adverse selection signature: healthier members exit when prices rise, leaving a sicker, costlier risk pool behind. The directional finding is consistent with every carrier-level morbidity assumption in the 2026 rate filing cycle, where actuarial memoranda uniformly projected that healthier members would lapse at disproportionately higher rates. Wakely’s data now confirms the direction and provides a quantitative anchor for the 2027 cycle.

The magnitude of the enrollment decline compounds the selection effect. When Wakely adjusts for effectuation, mid-year attrition, and ongoing non-payment beyond January, total effective enrollment in the ACA individual market is projected to decline 17–26% in 2026 relative to 2025 effectuated levels. At the midpoint, that represents roughly 3 to 5 million fewer covered lives. The market is not just losing members; it is losing the specific members whose departure worsens unit economics for everyone who remains.

Constructing the Morbidity Adjustment Factor

The pricing methodology for translating Wakely’s findings into a rate-level morbidity adjustment follows a two-step process anchored to concurrent HCC risk scores.

Step 1: Recalculate the population-weighted average risk score. Start with the CY 2025 base-period allowed PMPM and its associated population-level risk score (call it RS2025). Wakely’s data provides the key differential: non-payers carried approximately 10% lower morbidity (lower concurrent HCC risk scores) than payers. If the exiting cohort represents 14–26% of the 2025 enrolled population (the range implied by the effectuation gap plus mid-year attrition), removing their risk profile from the pool mechanically increases the average risk score of the remaining population.

For illustration, assume a 2025 population risk score of 1.000. If 20% of members exit with an average risk score of 0.90 (10% below the population mean), the remaining 80% carry an implied average risk score of:

(1.000 − 0.20 × 0.90) / 0.80 = 1.025

That 2.5% risk score increase translates directly into a 2.5% morbidity adjustment on the allowed PMPM, before any additional trend or utilization effects. Under a more aggressive exit scenario (26% disenrollment with exiting members at 0.88 relative risk), the math yields approximately a 4.2% adjustment. This range aligns closely with Wakely’s published 2.9–6.5% morbidity estimate, with the upper bound incorporating compounding effects from NBPP verification disruptions discussed below.

Step 2: Adjust for induced utilization changes from metal-tier migration. The risk score recalibration captures compositional selection, but it misses a second-order effect: the massive shift in metal-level enrollment changes how the remaining members use their benefits.

Bronze Migration and Metal-Level Pricing Relativities

The 2026 open enrollment produced one of the largest single-year metal-tier shifts in ACA history. Bronze plan enrollment surged from 30% to nearly 40% of all marketplace selections, while silver plan enrollment fell from 56% to 43% (ACA Signups/CMS data). Gold plan share actually increased from 13.2% to 17.2%, suggesting a bifurcation: price-sensitive healthy members moved to bronze, while sicker members who value lower cost-sharing moved to gold.

This migration has three pricing implications that interact with the morbidity adjustment:

Utilization suppression in bronze plans. Bronze deductibles average $7,186 nationally for 2026, compared to $5,304 for silver plans (Peterson-KFF). The $1,882 deductible gap suppresses utilization at the point of service: members with high deductibles defer discretionary care, avoid specialist visits, and reduce prescription fills. This suppression lowers the observed allowed PMPM for bronze members but does not eliminate their underlying health risk. When these members do incur claims (emergencies, hospitalizations, chronic disease flares), the per-event severity is often higher because deferred care escalates acuity.

Concentration of high-cost claims in the shrinking silver and gold pools. Members who stayed in silver or upgraded to gold tend to be higher utilizers who value actuarial values of 70–80%. As the silver pool contracts by 13.6 percentage points, the remaining silver members are disproportionately those with chronic conditions, ongoing specialist relationships, and regular prescription needs. Their per-member costs rise not because their health changed, but because the healthier silver members left for bronze or exited entirely.

Actuarial value recalibration under the 2027 AV Calculator. The 2027 federal AV Calculator incorporates updated utilization data and, for the first time, includes out-of-network emergency care spending (SHVS). The updated continuance tables particularly affect platinum plans and 94%-AV silver CSR variants. Pricing actuaries must reconcile the metal-level experience emerging from 2026 (which reflects a fundamentally different enrollment composition than the AV Calculator’s calibration data) with the AV targets required for plan certification. Where the calculator’s built-in utilization assumptions diverge from the actual utilization patterns of the post-subsidy population, plan designs may need to be restructured to hit target AV bands within the de minimis tolerance.

Risk Adjustment Transfer Recalibration

Shrinking enrollment amplifies risk adjustment transfer payment volatility. The ACA risk adjustment formula redistributes costs between plans within a state based on relative risk scores, but it does not inject new money into the market. When enrollment drops 17–26% and the remaining pool is sicker on average, several dynamics compound:

  • The statewide average risk score rises, compressing the spread between high-risk and low-risk plan averages. Plans that previously received large transfer payments may see receipts shrink as the market average converges toward their own acuity level.
  • Smaller absolute enrollment means each individual member’s risk score contributes more to the plan-level average. For smaller issuers with 5,000 to 15,000 members, a handful of high-cost claimants can swing the plan’s relative risk position by several points, increasing per-member risk adjustment uncertainty.
  • The 2027 NBPP final rule recalibrates HHS risk adjustment using 2021–2023 EDGE data and reduces the FFE user fee from 2.5% to 1.9%. The user fee reduction is a direct PMPM benefit that partially offsets the morbidity increase, but the model recalibration changes HCC coefficient values in ways that may amplify or dampen the morbidity shift depending on the specific diagnostic profile of the remaining population.

For multi-state carriers, the risk adjustment interaction is particularly complex. The transfer formula operates at the state level, so a carrier whose national book shows stable aggregate risk scores may still face material transfer payment swings in states with concentrated disenrollment. The actuarial memorandum should document state-level risk adjustment assumptions separately rather than relying on a blended national estimate.

State-Based Exchange vs. FFE Divergence

Wakely’s data reveals meaningful divergence in effectuation rates between state-based exchanges (SBEs) and the federally facilitated exchange (FFE). SBE states generally retained more enrollees and experienced less disruption than FFE states (Wakely). This divergence likely reflects several factors: SBE states tend to have more robust outreach and enrollment assistance infrastructure, some SBE states layered state-funded subsidies or reinsurance programs on top of expiring federal credits, and SBE platforms had more flexibility to customize the renewal and payment experience.

For pricing actuaries at carriers operating across both exchange types, this divergence requires bifurcated morbidity assumptions rather than a single national adjustment. An FFE-heavy carrier may need a morbidity load at or above 5%, while an SBE-concentrated carrier in a state with supplemental subsidies might justify an adjustment closer to 2.5–3%. Filing a blended rate across exchange types where the underlying risk pool compositions diverge by several points of morbidity creates cross-subsidization that understates rates in FFE markets and overstates them in SBE markets.

The NBPP Compounding Effect

On top of the subsidy-driven enrollment cliff, CMS’s 2027 NBPP layers additional verification requirements that threaten further risk pool deterioration. The proposed rule requires exchanges to verify income when external data sources indicate household income falls below 100% of the federal poverty level or when IRS tax data is unavailable. CMS projects this change would reduce marketplace enrollment by up to 2 million people and cut federal premium tax credit spending by $10.4 billion in 2027 (CMS).

An estimated 4.7 million enrollees could receive data matching issues (DMIs) under the expanded income verification requirements. As ACAP noted in its comment letter, healthy individuals are more likely to be deterred by additional documentation requirements, meaning the verification-driven attrition carries the same adverse selection signature as the premium-driven effectuation gap (Georgetown CHIR). The joint probability of subsidy-cliff disenrollment and NBPP verification disruption is not simply additive; it is correlated, because the populations most affected by both forces overlap substantially (low-income, intermittent employment, hourly or seasonal workers).

Stress-testing the high-end morbidity scenario requires modeling the combined impact: the 6.5% upper bound of Wakely’s estimate incorporates aggressive disenrollment compounded by verification-driven attrition. Carriers should document this scenario explicitly in the actuarial memorandum as a plausible adverse case, even if the filed rate reflects a point estimate closer to the midrange.

Documenting the Assumption Under ASOP No. 8

ASOP No. 8 (Regulatory Filings for Health Benefits, Health Insurance, and Entities Providing Health Benefits) requires the filing actuary to disclose the assumptions used in developing the rate filing, review those assumptions for reasonableness, and document the basis for professional judgment where actuarial experience is limited or inapplicable. The 2027 ACA filing cycle presents a specific challenge under Section 3.4 (Assumptions): the CY 2025 base period reflects a fundamentally different enrollment environment than the projected 2027 coverage year.

CY 2025 experience data captures a market with enhanced premium tax credits still in effect, effectuation rates above 96%, and a risk pool composition that no longer exists. Q1 2026 emerging data, by contrast, reflects the post-subsidy reality but covers only one quarter, carries limited statistical credibility for smaller carriers, and may not yet reflect the full extent of mid-year attrition.

The professional judgment required here involves credibility-weighting two fundamentally different data sources. Under ASOP No. 25 (Credibility Procedures), the actuary should consider both the volume and the relevance of each data source. CY 2025 has volume (a full year of claims data) but limited relevance (the subsidy environment has changed). Q1 2026 has relevance (same policy environment as the projection period) but limited volume (one quarter of experience). A reasonable approach documents a credibility blend that gives increasing weight to 2026 emerging data as subsequent quarters become available, with the filed rate reflecting the best estimate at the time of submission and explicit sensitivity testing around the morbidity assumption range.

The actuarial memorandum should also address the interaction between the morbidity adjustment and the GLP-1 pharmacy trend loading. Both factors increase expected PMPM costs, but through distinct mechanisms. If the morbidity adjustment implicitly captures higher GLP-1 utilization among sicker stayers (who are more likely to carry chronic conditions warranting GLP-1 therapy), and the pharmacy trend assumption separately loads for GLP-1 utilization growth, the filing double-counts. Document the separation explicitly and confirm that the GLP-1 trend loading reflects utilization and unit cost growth for the projected population, not a population mix effect already captured in the morbidity factor.

Why This Matters for Pricing Actuaries

The Wakely WNRAR data gives 2027 ACA rate filings something the 2026 cycle lacked: observed, carrier-level evidence of adverse selection in the post-subsidy market. The 2026 filings relied on analogues, professional judgment, and directional estimates. The 2027 filings can anchor morbidity assumptions to actual effectuation differentials and measured risk score gaps between payers and non-payers.

Three dimensions of this pricing problem require ongoing attention:

The calibration window is narrow. Rate filings are due to state departments of insurance in the summer of 2026. By filing deadline, actuaries will have at most two quarters of post-subsidy claims experience, and the second quarter data will be immature. The credibility-weighted blend between CY 2025 base-period data and 2026 emerging experience is the central professional judgment call of this filing cycle.

The spiral risk is asymmetric. Conservative morbidity assumptions (closer to 2.9%) risk rate inadequacy if actual adverse selection reaches the upper bound. Aggressive assumptions (closer to 6.5%) push premiums higher, which accelerates further disenrollment, which worsens morbidity, which validates the aggressive assumption. The feedback loop between pricing and enrollment makes this a self-fulfilling prophecy in either direction. Carriers with limited market share have the least ability to influence the market-wide equilibrium and the most exposure to getting the assumption wrong.

State regulators face an information asymmetry. DOI rate reviewers evaluating 2027 filings may benchmark morbidity assumptions against 2026 filed values (3–5 percentage points). But the 2026 assumptions were pre-Wakely estimates, made before effectuation data existed. A 2027 morbidity load above the 2026 level is not an aggressive pricing action; it is a response to observed data that confirmed the directional assumptions carriers made a year ago. The actuarial memorandum should explicitly frame the Wakely data as newly available evidence supporting the filed assumption, not as a departure from prior methodology.

Further Reading

Sources

ACA Marketplace Adverse Selection Actuarial Pricing Health Insurance Pricing Risk Adjustment