CMS FFS Normalization Factors by Year and Model

Last updated October 11, 2026 · Reviewed by Chin Ramamoorthi, CEO & Chief Architect, VBC Risk Analytics · Sources: CMS CY2026 and CY2027 Rate Announcements · Illustrative estimates

For CY2026 the CMS FFS normalization factor is 1.067, and for CY2027 it is 1.079, both for the standard Part C 2024 CMS-HCC (V28) model (non-PACE, non-ESRD). Raw risk scores are divided by the factor and then reduced by the separate 5.9% MA coding pattern adjustment. Together they lower the value of each raw RAF point by about 11.8% in CY2026 and 12.8% in CY2027.

What Is FFS Normalization?

CMS calibrates each CMS-HCC model so that the average fee-for-service beneficiary scores 1.0 in the model's base year. FFS risk scores drift upward over time, so for each payment year CMS projects that drift and divides payment-year scores by the projected factor. Normalization keeps a score of 1.0 meaning "average cost" in the payment year; it is not a penalty on MA coding.

Normalization vs the MA Coding Pattern Adjustment

The MA coding pattern adjustment is a different, statutory reduction that accounts for MA plans coding more intensively than fee-for-service. It is at least 5.9% and is applied after normalization. Payment RAF = raw RAF ÷ normalization factor × (1 − coding adjustment). ACO programs normalize differently and do not apply the MA coding adjustment; see normalization: MA vs ACO and the CMS LEAD Model.

Scores on MMR files are already payment-adjusted. Apply these factors only to raw scores, never twice.

Factor Table 2024–2027: Part C by Model and Part D

CY2025 blended two models. Normalize each model's raw score with its own factor, then blend 33% (2020 model) with 67% (2024 model); never average the factors. Example: raw scores of 1.10 (2020 model) and 1.05 (2024 model) give a CY2025 payment RAF of 0.9297; for 10,000 members at a $1,000 base rate and 100% of premium, that is $111,569,095 a year.

The method changed between years. For CY2026, "CMS calculated the normalization factors using a five-year multiple linear regression methodology and average historical FFS risk scores from 2020 through 2024" (CY2026 Rate Announcement, p. 5). For CY2027, CMS used "a four-year simple linear regression methodology and average historical FFS risk scores from 2022 through 2025" (CY2027 Rate Announcement, p. 3). Check each year's Rate Announcement for the final factors before you model a new year, or see what they do to one contract in the CMS Rate Announcement 2027 impact calculator.

Part C vs Part D: CMS-HCC and RxHCC Normalization

CMS-HCC models predict Part C medical costs; the RxHCC model predicts Part D drug costs, from a different set of condition categories. Each has its own normalization factors, and Part D is paid differently: the direct subsidy is the plan's own standardized bid times the RxHCC risk score, less the enrollee's basic Part D premium, with reinsurance and risk corridors on top. There is no county base rate, and the MA coding pattern adjustment applies only to Part C risk scores.

Since CY2025, CMS has set separate RxHCC factors for MA-PD plans and PDPs: after the Inflation Reduction Act redesign of Part D, the model over-predicted MA-PD costs (predictive ratio 1.106) and under-predicted PDP costs (0.879). For CY2027, the MA-PD factor falls from 1.194 to 1.109, which raises an MA-PD plan's normalized Part D risk score by about 7.7% for the same raw score, while Part C normalization moves the other way.

Revenue Impact: CY2026 vs CY2027

Take a provider group with 25,000 members, a $965 base rate, an average raw RAF of 1.08 and 100% of premium:

• CY2026: payment RAF 0.9525, annual revenue $275,738,575, value of 0.01 RAF $2,553,135
• CY2027: payment RAF 0.9419, annual revenue $272,671,974, value of 0.01 RAF $2,524,741

The higher CY2027 factor alone lowers revenue by about $3,066,601 a year before any base-rate change. Run your own numbers in the RAF revenue calculator.

How to Model Normalization in a Forecast

• Use published factors for published years (CY2026 and CY2027) and label later years as assumptions.
• Keep normalization and the coding adjustment as separate inputs, so a change in either is visible.
• Know your RAF source: raw scores from your own coding data need both adjustments; MMR payment scores need neither.
• Update after every Advance Notice and Rate Announcement, and show the revenue change in your Medicare Advantage revenue forecast.

How these factors flow into provider contracts (% of premium, corridors and caps) is covered in the MA risk contract profitability model.

FAQ: CMS Normalization

What is the CMS normalization factor for 2026 and 2027?

For the standard Part C CMS-HCC model, the FFS normalization factor is 1.067 for CY2026 and 1.079 for CY2027, both under the 2024 CMS-HCC (V28) model. Raw risk scores are divided by the factor. See the CMS Rate Announcement 2027 impact calculator.

What is the difference between HCC and RxHCC?

CMS-HCC models risk-adjust Part C (medical) payments; RxHCC risk-adjusts Part D (drug) payments. They use different condition categories and separate normalization factors, and RxHCC has separate factors for MA-PD plans and stand-alone PDPs. The 5.9% MA coding pattern adjustment applies only to Part C.

Is normalization the same as the MA coding pattern adjustment?

No. FFS normalization keeps the model calibrated to fee-for-service risk-score trend; the MA coding pattern adjustment is a separate statutory reduction (5.9% in CY2026 and CY2027) for MA versus FFS coding differences. Both apply to raw scores, one after the other. Compare with ACOs in normalization: MA vs ACO, and see why your ACO risk score differs from CMS’s.

How much does normalization reduce revenue?

With the 5.9% coding adjustment, CY2026 factors reduce the value of a raw RAF point by about 11.8% and CY2027 factors by about 12.8%. For 25,000 members at a $965 base rate and raw RAF 1.08, the move from CY2026 to CY2027 factors lowers annual revenue by about $3,066,601. Run your own contract in the RAF revenue calculator.