Benchmarking

Benchmarking Methodology

The Performance Year Benchmark is the ACO’s expected per beneficiary, per month (PBPM) expenditure, against which actual spending is compared to determine shared savings or losses. LEAD calculates separate PBPM benchmarks for three beneficiary categories: Aged and Disabled (A&D), End-Stage Renal Disease (ESRD), and High Needs. Benchmarks are also calculated separately for claims-aligned and voluntarily aligned beneficiaries. (RFA §VIII)

Step 1: Historical Baseline Expenditures (RFA §VIII.A)

CMS calculates the ACO’s historical baseline using a three-year period comprising the three most recent calendar years before the ACO’s first performance year. For example, in PY 2027, the historical baseline years are CY 2024, CY 2025, and CY 2026.

For claims-aligned beneficiaries, CMS retroactively applies the claims-based alignment algorithm to the ACO’s current Participant TIN List for each base year. This means the baseline is recalculated each performance year to reflect changes in the TIN list, even though the base years themselves don’t change. (RFA §VIII.A)

Note: If the Participant TIN list used for baseline calculation does not change, the historical claims-aligned baseline for those calendar years does not change either. Instead, updates come from re-running alignment against the same fixed base-year claims with an updated TIN list.

For voluntarily aligned beneficiaries, the benchmark is based on historical spending of beneficiaries who are voluntarily aligned in the current PY, capped at no more than ±10% of the claims-aligned benchmark. This cap is new in LEAD and is designed to prevent selection arbitrage. (RFA §VIII.A)

Base Year Weighting:

ACO Type	BY1 Weight	BY2 Weight	BY3 Weight
Renewing ACOs	1/3	1/3	1/3
Newly Entering ACOs	10%	30%	60%

(RFA §VIII.A)

Step 2: ACO-Specific Benchmark Adjustments (RFA §VIII.B)

After calculating the baseline, CMS applies ACO-specific adjustments:

Regional Efficiency Adjustment: Available only to Global Risk ACOs with baseline spending below regional FFS averages (“lower-spending ACOs”). The adjustment equals 50% of the risk-adjusted difference between regional and ACO historical spending. Higher-spending ACOs do NOT receive a negative adjustment; this is a one-directional benefit. (RFA §VIII.B)

New, higher-spending TINs excluded from the REA (new policy): If a lower-spending ACO includes a TIN that is both higher spending and new to Medicare ACO models (no SSP or REACH participation in the two years before the ACO’s first LEAD PY), that TIN’s historical experience is excluded from the REA calculation. The TIN’s spending status is tested with the same three-year weighted blend methodology, with the region defined by the TIN’s aligned-beneficiary-weighted county average. Excluded TINs stay in the benchmark and aligned population, but their beneficiaries receive a $0 REA and are excluded from the benchmark adjustment cap calculation. This is LEAD’s counterpart to the SSP’s proposed growth adjustment for recruiting inexperienced providers, which LEAD will not adopt. (Methodology Paper §3.4.3; Policy Update 7/14/26)

Prior Savings Adjustment: Available to Renewing ACOs in either risk option, based on savings generated in the three calendar years immediately preceding LEAD. The methodology paper adds a predecessor continuity test: a predecessor REACH or SSP ACO may contribute only if at least 40% of the LEAD ACO’s Participant TINs participated in that predecessor (tested in BY3). If multiple predecessors qualify, CMS uses a beneficiary-weighted average of their per-beneficiary savings. For BY1 and BY2, a lower 30% threshold applies; if not met, that year’s savings are set to $0 and no substitute predecessor is used. Savings are measured gross (before the REACH Global discount or the CMS share under REACH Professional and SSP tracks) as the difference between the predecessor’s PY Benchmark and PY expenditures, divided by aligned beneficiary months. CMS takes a simple average of the three annual PBPM values; note the paper’s worked example includes a loss year in the average (superseding the May 5 office hours interpretation that loss years are set to $0). ACOs with a positive average are eligible. The proration factor equals average base-period beneficiary months divided by current PY beneficiary months, capped at 100%, and the adjustment equals 50% of the prorated average. The PSA is a single ACO-level PBPM amount; the benchmark adjustment cap is applied at the ACO level using beneficiary-category-weighted sub-population cap amounts. (Methodology Paper §3.4.4)eceding LEAD. (RFA §VIII.B)

The July 14 methodology paper describes both adjustments in sequence without restating the higher-of rule. This site retains the rule per the RFA and the May 5 office hours (the most recent explicit guidance), but renewing, lower-spending Global ACOs should confirm with the LEAD help desk before benchmark modeling.

Adjustment Hierarchy: If an ACO qualifies for both a Regional Efficiency Adjustment and a Prior Savings Adjustment (PSA), it receives the higher of the two, not the sum. (RFA §VIII.B)

Overall Cap: The cap on combined benchmark adjustments depends on prior MSSP participation and spending level: (RFA §VIII.B)

• 3% cap: Applies to lower-spending former MSSP ACOs — specifically those with more than 40% of Participant TINs from an MSSP ACO in which they participated within the previous two years.
• 5% cap: Applies to higher-spending former MSSP ACOs and to all non-MSSP ACOs.

Adjustment	Eligibility	Calculation
Regional Efficiency Adjustment	Global Risk ACOs with baseline spending below regional FFS averages	50% of risk-adjusted gap between regional and ACO historical spending
Prior Savings Adjustment	Renewing ACOs (either risk option)	50% of prorated average per-capita gross savings across 3 years preceding LEAD; requires 40% Participant TIN continuity with a predecessor ACO (30% for BY1/BY2); proration = base-period months / PY months, capped at 100%
1.5% Administrative Add-On	Higher-spending ACOs (any risk option)	1.5% of benchmark, paid monthly. Not included in PY expenditures; not repayable.

If an ACO qualifies for both a Regional Efficiency Adjustment and a Prior Savings Adjustment, it receives the higher of the two (not the sum). The Administrative Add-On is separate and additive. See the Capitation page for full details on the add-on. (RFA §VIII.B, §X.A.5)

Step 3: Trending Benchmarks to the Performance Year (RFA §VIII.C)

LEAD introduces a blended update factor:

• Two-thirds: Two-way blend of national and regional FFS growth rates (per MSSP methodology at 42 CFR § 425.652).
• One-third: Accountable Care Prospective Trend (ACPT), a prospectively set growth rate based on CMS Office of the Actuary modified USPCC projections. ACPT values are updated and announced annually before each Performance Year, rather than fixed for the duration of the agreement period as the RFA described. Separate ESRD and non-ESRD values apply; the non-ESRD ACPT covers both the A&D and High Needs categories. (Methodology Paper §3.5.1)

Two-way blend weight: The weight on the national growth rate is calculated using the ACO’s share of aligned beneficiaries in its regional service area for the applicable Performance Year, rather than for BY3. (Methodology Paper §3.5.1)

The ACPT is subject to guardrails that widen over time:

Performance Year	Upper Guardrail	Lower Guardrail
PY 2027 (Year 1)	+0.3%	−0.2%
PY 2028 (Year 2)	+0.6%	−0.4%
PY 2029 (Year 3)	+0.9%	−0.6%
PY 2030 (Year 4)	+1.2%	−0.8%
PY 2031+ (Year 5+)	+1.5%	−1.0%

(RFA §VIII.C)

Guardrail sequencing, a 500-beneficiary minimum, and a High Needs BY3-only rule: The plus or minus 10% guardrail is applied to the Historical Benchmark after the three base years are blended but before the Regional Efficiency or Prior Savings Adjustments. If an ACO has fewer than 500 voluntarily aligned beneficiaries (across all categories) that can contribute BY3 historical expenditures, CMS will not build separate VA benchmarks and will apply the corresponding claims-aligned benchmarks to VA beneficiaries. For PY 2027, VA Historical Baseline Expenditures for the High Needs category use only BY3, and a beneficiary contributes (and counts toward the 500) only if they met High Needs criteria in BY3. Those beneficiaries are correspondingly excluded from the BY3 High Needs risk score reference average. (Methodology Paper §3.2.5)

Step 4: Risk Adjustment (RFA §VIII.D)

After trending, CMS applies risk adjustment by beneficiary category. See the Risk Adjustment page for full detail.

Step 5: Benchmark Discount and Quality Withhold (RFA §VIII.E)

For Global Risk ACOs, a benchmark discount rate reduces the benchmark before settlement:

Year	Higher-Spending ACOs	Lower-Spending ACOs
PY 2027	1.75%	3.0%
PY 2028	2.0%	3.0%
PY 2029	2.25%	3.0%
PY 2030	2.5%	3.0%
PY 2031	2.75%	3.0%
PY 2032–2036	3.0%	3.0%

No benchmark discount applies to Professional Risk ACOs. ACO REACH applied 3.5% (PY2025) and 4.0% (PY2026) flat for all Global ACOs. (RFA §VIII.E)

A 3% quality withhold applies to all ACOs (reduced from 5% in ACO REACH). See the Quality page for how quality scoring determines earn-back. (RFA §VIII.E)

Regional Rate Book Transition (RFA §VIII.G)

After Year 5, CMS will phase in a regional rate book-based benchmark. The transition pace varies by region based on: 1) proportion of beneficiaries aligned to ACOs, 2) cumulative savings generated, and 3) proportion of higher-spending ACOs. The rate book will fully replace historical-expenditure-based benchmarks by the end of the model.

Treatment of CY 2026 as Base Year 3

Complete CY 2026 experience will not exist when PY 2027 benchmarks are first calculated, so CMS operationalizes a phased approach. The December 2026 Preliminary Benchmark Report uses January to September 2026 claims paid through October 2026, with three estimation layers derived from the LEAD National Reference Population in CY 2024 and CY 2025: a claims completion factor, a Q4 seasonality adjustment (the ratio of Q1 to Q4 PBPM over Q1 to Q3 PBPM, averaged across the two reference years), and an alignment completion factor. Subsequent releases replace estimates with actual experience. From 2028 forward, complete run-out exists before each Performance Year and no completion factors are needed. (Methodology Paper §3.2.3)

Report	Date	CY 2026 Claims (Lookback)	Run-Out	Adjustments
Preliminary Benchmark Report	December 2026	Jan to Sep 2026	Paid through Oct 2026	Q1 to Q3 completion factor, Q4 seasonality
Preliminary Update	February 2027	Jan to Dec 2026	No run-out	Q1 to Q4 completion factor
Q1 Benchmark Report	May 2027	Jan to Dec 2026	March 2027	None
Q2 Benchmark Report	August 2027	Jan to Dec 2026	March 2027	None
Q3 Benchmark Report	November 2027	Jan to Dec 2026	March 2027	None
Q4 Benchmark Report	February 2028	Jan to Dec 2026	March 2027	None

Spending designation timing and a one-time election: ACOs receive a preliminary higher-/lower-spending designation with the December 2026 report and a final designation in May 2027. If the designation flips between the two, the ACO elects which designation to use for PY 2027. From PY 2028 forward, final designations arrive in Q4 of the prior year. The designation is recalculated each PY using that year’s Participant TIN List against the fixed CY 2024 to CY 2026 base period, and never changes mid-year. Because the designation drives the discount rate, REA eligibility, and the 1.5% Administrative Add-On, ACOs near the boundary should model both designations before May 2027. (Methodology Paper §3.2.3, §3.4.2)

SAHS and Skin Substitute Exclusions From the Baseline

Two categories of spending are removed from BY1 and BY2 (CY 2024 and CY 2025) Historical Baseline Expenditures and from the corresponding national and regional calculations. First, 100% of expenditures for designated SAHS codes: HCPCS A4353 and A5057 for CY 2024; HCPCS A4352, A4353, A6197, L0486, L1852, and L3916 for CY 2025. Second, and separate from SAHS policy, 65% of skin substitute expenditures are excluded (35% retained), because the CY 2026 PFS final rule payment changes are expected to reduce Medicare skin substitute spending by roughly 90%, making base-period levels unrepresentative. The 65% figure approximates what SSP truncation removed in CY 2025, promoting cross-program consistency. (Methodology Paper §3.2.1)

Hybrid Alignment Benchmark Adjustment

Beneficiaries who align mid-year contribute expenditures for only part of the year, and Medicare spending is seasonal (lowest in Q1, highest in Q4). To correct for this, the benchmark for mid-year additions is multiplied by an adjustment factor equal to average PBPM during the aligned partial-year period divided by full-year average PBPM, using model-wide factors from the LEAD reference population in CY 2024 and CY 2025. Claims-based hybrid additions (effective April 1) use a Q2 to Q4 factor. Voluntary additions use the factor for their effective quarter: Q1 effective dates receive no adjustment; Q2, Q3, and Q4 effective dates use progressively larger factors. For PY 2027 the factors are model-wide, not beneficiary-category-specific; CMS may refine this in future years. (Methodology Paper §3.5.3)