The MA Overpayment Ledger
What It Costs the Trust Fund and What Reform Would Save
This article sits at the intersection of two series. MCR-00.01 established the HI Trust Fund’s projected depletion date of 2033, possibly 2032 with the effects of the One Big Beautiful Bill Act. MCR-02.01 through MCR-02.04 explained the payment mechanics: the 0.09% rate shock, the $7.2 billion chart review exclusion, the V28 model reform, and the encounter-based risk adjustment trajectory. This article connects the two. It does the arithmetic that links MA overpayment to the trust fund depletion date and asks what the reform mechanisms CMS is pursuing would actually save.
MedPAC estimated in January 2026 that Medicare will overpay MA plans by approximately $76 billion in 2026. The Committee for a Responsible Federal Budget, using MedPAC’s methodology as a starting point, projected $1.2 trillion in cumulative overpayments through 2035, of which $520 billion would come from the HI Trust Fund. CRFB’s conclusion was direct: absent these overpayments, the HI Trust Fund would be solvent for the next decade and beyond. Instead, it is projected to run out of reserves in 2032.
The numbers are contested. AHIP argues that MedPAC’s coding intensity estimates are methodologically flawed and that CMS’s own decision to maintain the statutory minimum 5.9% coding pattern adjustment, rather than increasing it, effectively rebuts MedPAC’s higher estimates. The debate is real. But the fiscal trajectory is not a matter of opinion. The trust fund depletion date is an actuarial fact that advances or retreats based on the relationship between inflows and outflows, and MA overpayment is the single largest addressable outflow on the ledger.
How MA Overpayment Occurs#
Three mechanisms account for the bulk of MA overpayment relative to what the same beneficiaries would cost in Traditional Medicare.
Coding intensity differential. MA plans generate higher risk scores for the same beneficiaries than Traditional Medicare would generate. The same person, with the same conditions, receiving the same care, produces a higher risk score in MA because the plan’s coding practices capture more diagnosis codes. MedPAC has documented this differential consistently. For 2026, MedPAC estimates that coding intensity is 10.3% higher in MA than FFS after the V28 model’s full phase-in reduced it from 12.5% in 2025 and approximately 16% in prior years under V24. The statutory coding pattern adjustment of 5.9% partially addresses this differential but captures only a fraction of it. The net effect is a roughly 4 to 5 percentage point residual coding intensity advantage that translates directly into higher risk-adjusted payments. Among the ten largest MA organizations, MedPAC found a 26-percentage-point spread in average coding intensity, meaning the problem is concentrated but not uniform.
Coding intensity accounted for approximately $40 billion of MedPAC’s $84 billion overpayment estimate for 2025. The $76 billion estimate for 2026 reflects a reduction attributable to V28’s full phase-in, which MedPAC found reduced payments while maintaining stable supplemental benefits and plan availability.
Favorable selection. MA enrollees are, on average, healthier than their risk scores suggest. Risk adjustment is designed to account for health status differences between MA and FFS populations, but the current model does not fully capture the selection dynamics. Beneficiaries with lower actual spending relative to their risk score disproportionately enroll in MA, while those with higher actual spending relative to their risk score are more likely to remain in FFS. MedPAC estimated that favorable selection increased MA payments by roughly 11% above FFS spending in 2025, accounting for approximately $44 billion of the $84 billion total. The selection effect persists across MA market penetration levels and is larger for enrollees with higher risk scores, suggesting that it is a structural feature of the program rather than a transitional phenomenon.
The benchmark-risk score asymmetry. MA benchmarks are derived from Traditional Medicare per-capita costs, but the risk scores used to adjust payments are generated by MA coding practices that are systematically more intensive than TM. The result is a structural mismatch: the benchmark establishes what a beneficiary should cost (based on TM), but the risk score inflates what the plan is paid (based on MA coding). The product of TM-derived benchmarks and MA-inflated risk scores is overpayment by construction. This is not fraud. It is a payment system design that creates a structural asymmetry MA plans have exploited within the rules as written.
The Cumulative Overpayment Estimate#
MedPAC’s annual estimates provide the data points. The trajectory provides the picture.
MedPAC estimated overpayments of approximately $88 billion in 2024, $84 billion in 2025, and $76 billion in 2026. The decline from 2025 to 2026 reflects V28’s full phase-in, which reduced the coding intensity component. CRFB’s projection of $1.2 trillion through 2035 applies MedPAC’s 2025 and 2026 starting points to a decade of projected MA enrollment growth, medical cost inflation, and the continuation of current payment policies with incremental reform.
The magnitude of these estimates depends on contested methodological assumptions. MedPAC measures coding intensity by comparing risk scores for beneficiaries who switch between MA and FFS, controlling for demographics and clinical characteristics. CMS maintains the statutory 5.9% CPA and has declined to increase it, arguing in the CY 2026 Rate Announcement that its own analysis does not support the higher estimates. AHIP’s position is that MedPAC’s methodology conflates coding accuracy (MA plans documenting conditions that FFS misses) with coding intensity (MA plans inflating risk scores beyond clinical reality). The distinction matters: if MA plans are genuinely capturing conditions that FFS undercodes, some portion of what MedPAC calls “overpayment” reflects more accurate clinical documentation in MA rather than inflated payments.
MedPAC’s counter is empirical. The 26-percentage-point spread in coding intensity among the ten largest MA organizations means that some plans are coding at rates far above what even the most generous interpretation of clinical accuracy would justify. If coding intensity reflected genuine clinical documentation, the variation across plans serving similar populations would be much narrower. The concentration of high coding intensity in plans with the most aggressive chart review operations reinforces MedPAC’s position that the differential is driven substantially by administrative coding practices rather than clinical documentation accuracy.
What the overpayment represents in trust fund terms: every dollar of MA overpayment above what the beneficiary would cost in TM is a dollar drawn from the HI Trust Fund that does not reflect the actual cost of caring for that person. MA overpayment also increases Part B premiums for all Medicare beneficiaries, including those in Traditional Medicare, because Part B financing absorbs a share of the excess MA payments. CRFB estimates $230 billion in higher premiums over the next decade attributable to MA overpayment.
Connecting Overpayment to the Depletion Date#
The HI Trust Fund is projected to deplete in 2033 under the 2025 Medicare Trustees Report, possibly 2032 with the effects of the One Big Beautiful Bill Act’s Medicaid provisions and other fiscal interactions (MCR-00.01). CRFB’s analysis establishes the counterfactual: if the $520 billion in HI Trust Fund costs attributable to MA overpayment over the next decade were eliminated, the trust fund would remain solvent through 2035 and beyond.
That counterfactual is not achievable in practice. No reform eliminates all MA overpayment. Plans will adapt to coding changes, encounter-based RA will phase in gradually, and favorable selection is a structural feature that coding reform alone does not address. But the counterfactual establishes the scale of the opportunity. Even partial overpayment recovery extends the depletion date.
A rough scenario framework: if CMS recovers $10 billion annually in overpayment through the combined effect of the chart review exclusion, CPA maintenance, V28’s continued impact, and incremental encounter-based RA transition, that is $100 billion over a decade in reduced HI Trust Fund draws. The actuarial effect would extend the depletion date by approximately one to two years, depending on assumptions about enrollment growth, FFS cost trends, and payroll tax revenue. If recovery reaches $20 billion annually through more aggressive CPA levels and full encounter-based RA implementation, the extension could reach two to four years. The range is wide because it depends on variables CMS does not fully control, but even the conservative end represents a material fiscal outcome achieved entirely through administrative action.
The political economy of this arithmetic is CMS’s strongest reform argument. Trust fund depletion triggers automatic 11% Part A payment cuts to every hospital, skilled nursing facility, and home health agency serving all 67 million Medicare beneficiaries. Overpayment reform reduces revenue to MA plans. The political asymmetry is stark: trust fund depletion harms everyone; overpayment reform affects plan margins.
What the Current Reforms Would Actually Save#
Chart review exclusion ($7.2 billion annually, gross). The $7.2 billion CMS estimate reflects the full exclusion of unlinked chart review diagnoses from risk scores. Plans will adapt by shifting chart review activity into encounter-linked processes, converting some unlinked revenue into linked revenue. The behavioral adaptation discount probably reduces the actual annual savings to $4 to $5 billion. Over a decade, the chart review exclusion generates $40 to $50 billion in gross savings before behavioral response, and $30 to $40 billion after adaptation. This is still the single largest administrative savings mechanism available to CMS.
Coding pattern adjustment (5.9%). The CPA saves approximately $20 to $25 billion annually at its current level. MedPAC has argued consistently that the 5.9% statutory minimum is insufficient and that the actual coding intensity differential justifies a higher adjustment. MedPAC’s 2026 estimate of 10.3% net coding intensity (after the 5.9% adjustment) implies that increasing the CPA to levels MedPAC considers adequate would capture an additional $20 to $30 billion annually. CMS has not increased the CPA above the statutory minimum, and doing so would require either legislative action to raise the floor or a CMS determination that the statutory minimum is insufficient, a determination the agency has so far declined to make.
Encounter-based RA (future). The savings from moving to encounter-only risk adjustment depend on implementation design and phase-in timeline. A five-year phase-in would generate savings gradually, with the trust fund impact back-loaded. The actuarial range is broad: encounter-based RA could reduce MA overpayment by $5 to $15 billion annually at full implementation, depending on how effectively it constrains the coding intensity differential that chart review exclusion and the CPA do not fully capture. Encounter-based RA is the single highest-leverage administrative action available for trust fund extension because it requires no legislation, no benefit cuts, and no program restructuring. It changes only how existing diagnoses are validated for payment.
The Political Economy of Reform#
The total revenue at stake is substantial. CRFB’s $1.2 trillion ten-year estimate represents the upper bound of what full overpayment elimination would save. No realistic reform scenario captures all of it. But the scale explains the intensity of the lobbying response.
AHIP’s advocacy positions the 0.09% rate, the chart review exclusion, and potential future CPA increases as threats to beneficiary benefits and plan viability. The argument is that reducing MA payments forces plans to cut supplemental benefits, increase premiums, exit counties, and reduce provider networks, all of which harm the 35 million beneficiaries enrolled in MA. The Better Medicare Alliance amplifies this framing, citing research showing that MA beneficiaries save an average of $3,500 annually compared to Traditional Medicare and that forced disenrollments from plan exits already affected 2.9 million beneficiaries in 2026.
CMS and MedPAC’s counter-argument is fiscal and structural. Overpayment reform extends the trust fund. The alternative to reform is trust fund depletion and automatic payment cuts that affect all 67 million Medicare beneficiaries, not just the 35 million in MA. MedPAC’s finding that V28 reduced coding intensity while maintaining stable benefits and plan availability undermines the industry’s claim that payment accuracy reforms inevitably harm beneficiaries.
Congress has acted on the margins. The Senate Judiciary Committee’s investigation of UnitedHealth Group documented coding intensity practices that support the administrative reform trajectory. The No UPCODE Act, reintroduced in March 2025, would prohibit chart review and HRA diagnoses from risk adjustment entirely. The Medicare Advantage Reform Act, introduced by Representative Schweikert, would go further by reducing benchmarks and eliminating quality bonuses. Neither bill has advanced to a vote, but their introduction signals congressional attention to the overpayment problem. The administrative path remains faster: CMS can implement chart review exclusion, encounter-based RA, and CPA adjustments through the annual rate-setting process without legislation and without a congressional vote (MCR-03.04).
The MA overpayment question is not an abstract policy debate. It is a trust fund arithmetic problem. Every dollar of overpayment accelerates the depletion date. Every dollar recovered decelerates it. The reforms CMS is pursuing are the administrative tools available to address the overpayment without legislation. Whether they are applied aggressively enough to materially extend the trust fund depends on the same political dynamics that have prevented MedPAC’s recommendations from being enacted for more than a decade. The depletion date does not wait for political consensus.
Related Reading#
MCR-00_01 The Trust Fund Clock MCR-04_01 Is MA Still Worth It? The Strategic Recalculation for Insurers MCR-12_01 The MA Plan Landscape Under Pressure: UnitedHealth, Humana, CVS/Aetna, Elevance, and the Regional Plans