CMS Proposes 75% AMP Drug Price Ceiling for Part D Starting IPAY 2029: Actuarial Pricing Framework

On June 16, 2026, CMS published a 132-page proposed rule in the Federal Register (91 FR 36236, CMS-4215-P) that would, for the first time, embed the Medicare Drug Price Negotiation Program into binding 42 CFR rulemaking. The prior two negotiation cycles, covering 10 drugs with Maximum Fair Prices effective January 1, 2026, and 15 drugs with MFPs effective January 1, 2027, operated under sub-regulatory guidance. Starting with Initial Price Applicability Year 2029, the proposed rule codifies the MFP ceiling at 75% of non-federal average manufacturer price for small molecules and 65% of AMP for biologics, expands the annual drug selection to up to 20 drugs per cycle, and establishes a renegotiation trigger for material post-agreement changes in a drug's commercial trajectory (Federal Register, June 16, 2026). Comments are due August 17, 2026.

Part D bid actuaries building 2028 projections in 2027 cannot treat the proposed rule as a distant compliance problem. The drugs likely to be selected in the fourth and fifth negotiation cycles are already on formulary. The ceiling formula is the quantitative input that bounds drug cost projections from above for any drug not yet selected. The effectuation mechanics are already live on the first 10 drugs, generating the first real-world IBNR data. And the comment window closes before most 2028 bid teams begin final pricing. This piece develops the three-stage projection framework the rule demands, the IBNR reconfiguration it forces, and the dual-track problem introduced when MFN executive order pricing overlaps with IRA negotiated prices.

The 75%/65% AMP Ceiling: What the Number Is and Is Not

The AMP ceiling is an upper bound, not a target. For any drug entering the negotiation program, CMS's opening offer will be at or below 75% of non-federal AMP for small molecules (or 65% for biologics), but the actual negotiated MFP is typically lower. Across the first 15 IPAY 2027 drugs, CMS achieved negotiated discounts ranging from 38% to 85% below non-negotiated net prices (Milliman, November 2025). The ceiling formula, though, is the actuarially useful number: it sets the outer bound on a modeled MFP distribution for drugs not yet selected, allowing a plan to construct a probability-weighted drug cost projection without waiting for a negotiated price announcement.

AMP itself is not a mystery figure. Manufacturers report quarterly AMP to CMS under 42 U.S.C. 1396r-8, the same statutory authority that underlies Medicaid best price. The quarterly AMP data releases give actuaries the denominator for the 75% calculation. Applying that ceiling to a candidate drug's current AMP and then applying a negotiation haircut distribution produces a range of projected MFPs. The haircut distribution should be calibrated to CMS's historical negotiation outcomes: the 38%-to-85% range from IPAY 2027 suggests a wide spread, and the actuary should build scenario weights rather than a point estimate.

The proposed rule also includes a Small Biotech Drug provision: for drugs manufactured by companies with annual gross revenues below $1 billion, CMS is prohibited during IPAY 2029 and 2030 from offering or agreeing to a counteroffer below a specified statutory floor (CMS, June 2026). This creates a bifurcated ceiling calculation: large-manufacturer drugs are bounded at 75%/65% of AMP from above, small-biotech drugs are also bounded from below during the transition years. For plans with meaningful small-biotech exposure, the projected MFP distribution is more compressed, with less downside to the price.

A Three-Stage Projection Framework for 2028 Bids

Building a 2028 bid under this proposed rule means projecting drug costs for formulary items that may or may not be selected in the fourth cycle (IPAY 2029) or the renegotiation pool. The framework has three stages that run sequentially, each producing a probability distribution that feeds the next.

Stage One: Drug Selection Probability. CMS's selection screens are public and specific. Negotiation-eligible drugs must have seven or more years of exclusivity remaining for small molecules, eleven or more for biologics. CMS ranks candidates by total Part D gross covered drug cost and excludes already-selected drugs, orphan drugs with a single rare disease designation, and plasma-derived biologics. Beginning with IPAY 2029, the annual selection expands to up to 20 drugs, up from 15 in IPAY 2027 and 10 in IPAY 2026. The actuary applies these screens to the current CMS Drug Spending Dashboard and arrives at a candidate pool of roughly 50 to 70 drugs per cycle. A Poisson-based selection model weighted by spend rank and exclusivity proximity assigns selection probabilities. Drugs in the top decile of Part D gross spend with exclusivity windows expiring in the 2030 to 2033 range carry the highest selection probability and warrant individual projections rather than pooled treatment.

Stage Two: MFP Ceiling Calculation. For each high-probability candidate, the actuary pulls the drug's most recent quarterly AMP from CMS data and applies the 75% ceiling (or 65% for biologics). This produces an MFP ceiling estimate. The actuary then applies a negotiation discount distribution calibrated to CMS's published outcomes: a central estimate somewhere between 50% and 65% of the AMP ceiling, with scenario weighting toward the tighter negotiated price for drugs facing therapeutic competition and a wider range for drugs with dominant market position. The result is a probability-weighted projected MFP per candidate drug, expressed as a percentage of the current adjudicated price to produce a per-unit cost reduction assumption.

Stage Three: Utilization Response Modeling. When an MFP drug moves from specialty tier to preferred-brand tier, member cost-sharing drops substantially. A drug priced at 33% coinsurance specialty tier may shift to a $47 fixed copay on preferred-brand. KFF has documented that for chronic-condition drugs under similar cost-sharing reductions, utilization increases of 15% to 30% are consistent with observed responses (KFF). Milliman's analysis of the IPAY 2026 and 2027 selections found meaningful adherence improvements, particularly in diabetes and cardiovascular drugs, where the utilization response skews toward the upper end of that range. The actuary must run a demand-elasticity model calibrated to the drug class, net the unit cost reduction against the volume increase, and produce a pharmacy trend adjustment for the bid. For several chronic-condition drugs near selection probability thresholds, the net effect of lower price plus higher utilization is near-neutral or slightly favorable for plan costs; the dangerous assumption is that cost falls proportionally with price while ignoring adherence recovery.

MFP Effectuation and the IBNR Triangle Problem

The effectuation mechanics of MFP pricing break the historical pharmacy claim development pattern in a specific and measurable way. Under legacy rebate contracting, a plan adjudicates a drug claim at gross price at point of sale and carries a rebate receivable until the manufacturer remits, typically six to twelve weeks later. The pharmacy IBNR triangle reflects this lag: three-month reserves carry a gross claim liability offset by a rebate receivable that is still settling.

Under MFP effectuation, the plan pays the net MFP price at the point of sale. The 14-day payment window in the CMS effectuation framework requires the primary manufacturer to transmit payment within 14 calendar days of the Medicare Transaction Facilitator verifying the claim (CMS, Final Guidance IPAY 2027). There is no gross claim and no rebate receivable: the claim develops fully at adjudication. This collapses the IBNR lag structure for MFP drugs to near-zero, because pharmacy claims adjudicate in real time and MFP settlement occurs almost simultaneously.

A worked example quantifies the reserve difference. Suppose a large Part D plan has monthly pharmacy spend of $200 per member per month, with 20% of that spend concentrated in the first 10 MFP-selected drugs. Under historical rebate contracting for those drugs, an eight-week rebate lag on the 20% MFP portion creates an IBNR carry of approximately 0.20 x (8/4.33 months) x $200, or roughly $7.40 PMPM in reserve for the MFP drug segment. Once effectuation eliminates the lag, the appropriate three-month IBNR carry for those drugs drops to near zero, reducing total pharmacy reserve by that $7.40 PMPM. For a plan with 100,000 members, that is approximately $740,000 in reserve adjustment per three-month period. Plans that do not segment their IBNR triangles by MFP versus non-MFP drugs will carry inflated reserves for the MFP segment and understate any adverse development in the non-MFP segment where the rebate lag persists.

The correct method is triangle segmentation. The actuary maintains two pharmacy development triangles: one for MFP-selected drugs, where development factors should compress to near 1.00 by month two, and one for non-MFP drugs, where the historical rebate lag patterns apply. As MFP penetration grows across IPAY 2029 and subsequent cycles, the non-MFP triangle shrinks as a share of total pharmacy spend, and the plan's aggregate pharmacy IBNR reserve will decline structurally. This is not an error; it is the correct actuarial treatment of the changed cash flow timing. Plans that anchor tail factors to pre-MFP development patterns will overstate their IBNR as the MFP drug share grows.

The proposed rule does not yet specify how MFP-driven IBNR methodology should appear in statutory reserve opinions. That is the opening for the August 17 comment period. Actuaries who want specific reserve guidance for MFP effectuation mechanics should submit comments requesting it. The comment cycle is pre-final; the methodology question is live.

Renegotiation Trigger: A New Uncertainty in Multi-Year Projections

The proposed rule introduces a renegotiation mechanism: CMS may reopen MFP negotiations for a selected drug if that drug's sales volume, clinical use indication, or market share changes materially after the original agreement. The prior guidance-based cycles had no equivalent provision. For bid actuaries, this introduces a probability-weighted scenario that was not in any prior Part D model: a drug may enter a bid year at one MFP, then receive a revised MFP mid-period if CMS exercises the renegotiation trigger.

Two scenarios matter most. A drug that gains a major new clinical indication after MFP effectuation will see sales volume rise substantially, potentially triggering renegotiation toward a lower MFP. A drug that loses significant market share to a generic or biosimilar competitor may no longer meet the trigger threshold, but the actuary should model the scenario where CMS renegotiates upward under a changed-competition argument. The direction of renegotiation is ambiguous from a regulatory text standpoint; the actuary should carry a symmetric uncertainty band around the modeled MFP for high-probability renegotiation candidates, rather than assuming the published MFP is stable for the bid period.

The Dual-Track Problem: MFN Pricing Runs Alongside IRA MFPs

Two simultaneously operative federal price regimes now apply to semaglutide products in Part D. The Trump administration's most-favored-nation executive order produced manufacturer agreements to price Ozempic, Mounjaro, Wegovy, and Zepbound at $245 per month for Medicare. The IRA negotiated MFP for semaglutide in the IPAY 2027 cycle is $274 per month. CMS has stated that the lower MFN price supersedes the IRA MFP when both apply to the same drug claim (AMCP, 2026).

The proposed rule does not fully resolve how Part D sponsors should adjudicate claims when the two price regimes produce different prices on the same transaction. CMS's statement that the MFN price takes precedence is currently administrative, not codified. For bid actuaries, the practical problem is which price to use in the 2028 projection: the IRA MFP ceiling calculated from AMP, or the MFN agreement price, which was negotiated through a separate legal framework entirely. The MFN deal with Novo Nordisk and Eli Lilly does not run through 42 CFR; it runs through voluntary manufacturer commitments secured by the executive order. The actuarial conservatism position is to use the lower of the two applicable prices as the projected drug cost but carry a scenario weight for the possibility that one regime is invalidated or that the effectuation infrastructure does not support simultaneous application of both.

GLP-1 drugs already represent one of the largest single line items in the pharmacy trend projection for most large Part D sponsors. A $29-per-month difference between MFN and MFP pricing across millions of member-months is not a rounding error. Plans that assume the IRA MFP governs are projecting costs $29 higher per semaglutide user per month than plans that assume MFN supersedes. That divergence needs a documented methodology decision in the bid workpapers, not a tacit assumption about which number to pull from the pricing system.

EGWP Creditable Coverage: A Quiet Compliance Risk

Employer Group Waiver Plans carry an additional layer of complexity. When MFP pricing reduces a drug's plan-share cost, it may alter whether the underlying employer benefit remains creditable under the Part D minimum creditable coverage standard. Creditable coverage requires that the actuarial value of the employer's drug benefit meet or exceed the actuarial value of standard Part D, currently set at 72% for 2026 with 73% mandatory in 2027 under the IRA Part D redesign, as analyzed in the creditable coverage analysis.

When a drug moves from specialty tier to preferred-brand tier under MFP pricing, the employer's plan-share cost on that drug decreases. For benefit designs where the employer was carrying a thin actuarial value margin above the creditable coverage threshold, a reduction in plan-share cost on one or more high-volume drugs can push the plan below creditable coverage, requiring recertification and potentially benefit redesign before the next plan year. The EGWP actuary who is modeling the 2028 bid must run the creditable coverage actuarial value calculation both with and without the projected MFP drug cost reductions and document whether the benefit redesign scenarios hold.

Bid Workpaper Documentation Under a Proposed Rule

A proposed rule is not yet final. CMS expects to finalize the regulation in fall 2026, before the 2028 bid cycle opens. But the methodology decisions a Part D actuary makes now, encoding AMP-based ceiling calculations, segmented IBNR triangles, and dual-track price scenarios, will run through multiple bid cycles whether or not the specific regulatory text changes between proposed and final. The Part D bid is a one-year instrument but the underlying modeling assumptions, once embedded, tend to persist.

The documentation requirement is not merely CMS compliance. When a plan's pharmacy trend comes in materially different from bid for 2027, and the post-effectuation IBNR is lower than reserved, the bid-to-actual analysis should show exactly which methodology choices drove each variance. Plans that cannot decompose their 2027 IBNR favorability into MFP-effectuation lag compression versus genuine trend favorability will be carrying noise into their 2028 projections.

The comment period through August 17, 2026 is the most direct mechanism available. CMS is making a permanent transition from sub-regulatory guidance to 42 CFR rulemaking for a program that will govern Part D drug cost pricing for every cycle after 2029. Reserve methodology guidance, dual-track price adjudication rules, and renegotiation trigger specificity are all areas where actuarial input, submitted formally, can shape the rules that will govern the next decade of Part D bid construction.