AMP is WAC -- 7/24/26

One Hump or Two?

Sticks Don’t Grow into Carrots. On Tuesday, President Trump announced a phased tariff schedule on imported generic drugs via Truth Social: zero tariffs through August 2028, then 100%, then 200% in 2029. The administration plans to invoke Section 232 of the Trade Expansion Act of 1962, the national security provision, which doesn’t require congressional action.

Generic drugs account for about 90% of American prescriptions. They are cheap specifically because they’re manufactured at scale overseas, mostly in India, with raw ingredients from China. Two-thirds of U.S. birth control pill prescriptions come from just two Indian generics companies. Ninety-nine percent of statin prescriptions are generic. The reason generic manufacturers haven’t rushed to build U.S. factories the way branded pharma has is straightforward: they operate on razor-thin margins. A 100-to-200 percent tariff doesn’t reshore manufacturing overnight; it raises prices and risks shortages. Supply chain experts and physicians have said so clearly.

(For what it’s worth, roughly half of all generics come from India, but a February U.S.-India trade pact apparently stipulated that India would “receive negotiated outcomes” on pharmaceuticals, which may limit the actual impact on Indian manufacturers. We’ll see.)

The contrast with branded pharma is worth noting. Pfizer, Merck, BMS, and AstraZeneca already cut private deals with the White House, and their tariff exposure is resolved. Their situation was a bilateral negotiation with a counterpart that had reasons to make a deal. The generics industry doesn’t have that leverage. Generic manufacturers compete on price, not on proprietary molecules.

Encouraging generic onshoring is awesome, but it isn’t like flipping a switch. And honestly, I think we’re okay with some of these drugs not being made here, but blanket policies can and will create weird dynamics.

In Plain Sight. Last week, Larry Levitt over at KFF published a piece in JAMA Health Forum asking a question I’ve been sitting with for a while: why do drug prices dominate the policy conversation when hospitals are a much bigger driver of health spending?

Look at the numbers -- hospitals account for 31% of all national health spending. Retail prescription drugs? Nine percent. Forty percent of the increase in health spending from 2022 to 2024 came from hospitals; 11% came from drugs. And yet here we are, with a Medicare negotiation program, a coupon website, a most-favored-nation (MFN) executive order, and now a tariff regime all aimed squarely at pharmaceutical manufacturers. On one hand, my career is grateful, but as someone who loves the industry, I could do with less focus on us.

Yes, there are real structural reasons drug prices register differently. People fill an average of about 12 prescriptions a year, roughly one a month. Two-thirds of adults are currently taking a prescription drug. Hospital stays happen to about 1 in 10 people in any given year. Visibility and out-of-pocket spending on drugs do make it more of a target. It doesn’t help that benefit designs load coinsurance onto pharmacy in a way they typically don’t for inpatient care.

But here’s what gets me. Private insurers pay hospitals more than two and a half times what Medicare pays. Hospital prices vary enormously across markets, driven largely by consolidation. The last president who tried to seriously rein in hospital spending was Jimmy Carter. Nearly 50 years ago. AND Congress didn’t let him do it.

A 2024 KFF poll found that 75% of respondents said pharmaceutical companies deserve “a lot of the blame” for rising health care costs. Forty percent said hospitals. Plus, hospitals are local employers, much harder to legislate against.

Drug prices are real and worth addressing, particularly when they drive patient out-of-pocket costs. But if we’re serious about health care spending, we’re currently putting enormous policy energy into nine cents of every health care dollar while the thirty-one-cent problem largely escapes scrutiny.

4.5 Stars (No Terms and Conditions Apply). Last week, Erin Socker and Katie Keith published a piece in Health Affairs Forefront on the Medicare Advantage (MA) quality bonus program, specifically what a string of recent court decisions reveals about how fundamentally broken the whole thing is.

The program has been quietly costing taxpayers billions for years, $16 billion projected for 2026 alone, more than triple what it cost in 2016. Critics have long argued it doesn’t improve quality or help beneficiaries make meaningful plan comparisons. MedPAC has said so repeatedly. The Congressional Budget Office (CBO) estimated back in 2018 that eliminating it would save nearly $100 billion over ten years. And yet it persisted, because MA plans love the money and reform is politically complicated.

Enter Clover Insurance. Clover is an MA plan that received a 3.5-star rating instead of 4 stars for 2026, a distinction worth $120 million in lost bonus payments for 2027. They sued the Centers for Medicare & Medicaid Services (CMS). And in May, a federal district court in Georgia sided with them, finding that CMS improperly relied on 20 quality measures: ten exceeded the agency’s statutory authority under the Medicare statute, and ten should have gone through notice-and-comment rulemaking. (Loper Bright is doing a lot of work here. Courts are much more willing to second-guess CMS right now, and the agency knows it.)

And get this, CMS responded by voluntarily recalculating star ratings for all Medicare Advantage plans, not just Clover. And they set it up as a one-way ratchet. Plans whose recalculated rating is higher get the new number. Plans that would score lower keep their original rating. No plan’s payment goes down. I mean, I get it and I’m also annoyed by it.

Clover went from 3.5 to 4.5 stars. Then Elevance filed suit seeking $115 million more. SCAN Health Plan filed for $125 million. Alignment Healthcare wants $50 million. CareFirst’s case is on hold pending recalculation.

A quality program that can only go up in its payments regardless of what plans do is not a quality program. Congress has the tools to fix this; the question is whether anyone has the appetite.

Fun Never Stops. Last week, CMS released draft guidance on how manufacturers must effectuate the Maximum Fair Price (MFP) in 2028, and the biggest development in it isn’t getting nearly enough attention. Personally? I missed it in the IPAY 2029, MPFS, holiday haze.

The Inflation Reduction Act (IRA) drug negotiation program has been mostly a Part D story. Round one, round two, all pharmacy-dispensed drugs, all Part D. IPAY 2028 changes that. For the first time, the MFP can also apply to drugs payable under Medicare Part B. Comments are due September 18.

Here’s why Part B is operationally harder. Part B billing uses HCPCS codes, which are procedure codes. A single HCPCS code can cover multiple drugs. It might cover both a negotiated biologic with an MFP and a biosimilar that isn’t subject to negotiation. CMS can’t always determine from the claim alone which product was administered.

CMS is requesting comments on three approaches to fix this identification problem: require providers to use billing modifiers flagging whether an MFP-eligible product was used, require NDC-11 reporting on Part B claims, or create separate HCPCS codes for selected drugs. Each option carries different administrative burden for physicians and hospitals, providers who are already not exactly thrilled about being in the middle of this. And will lose money on each MFP drug because reimbursement will be based on a % of that and not Average Sales Price (ASP).

Then there’s the SDRA, the Standardized Default Refund Amount, which kicks in when a manufacturer doesn’t effectuate MFP prospectively and a refund has to be calculated. For Part B, CMS is soliciting comments on four options that involve either WAC (Wholesale Acquisition Cost) or ASP, the existing reimbursement benchmark for Part B drugs. ASP is already how Medicare pays for Part B drugs, which argues for consistency. WAC, as readers of this newsletter will appreciate, is a list price that doesn’t reflect actual transaction prices.

And yes, 340B is back. Manufacturers don’t owe MFP access if the 340B ceiling price is lower, but CMS is still not taking responsibility for nonduplication enforcement. Manufacturers have to figure that out themselves. It has been fun.

Most people talking about IRA negotiations are focused on which drugs get selected and what prices come out. That’s fair. But the question of how the negotiated price reaches the provider, especially for physician-administered drugs, is genuinely unresolved, and CMS is asking for public comment on some pretty fundamental questions with eighteen months until go-live.

8,000 and Counting. On Monday, IQVIA published a report on retail pharmacy closures, and the numbers are striking. The United States lost nearly 8,000 retail pharmacies between 2018 and 2025. More than 2,000 of those closed in 2025 alone. And yet the prescription volume going to the pharmacies that remain is rising, up 11% since 2023, from about 70,000 scripts per store annually to 78,000. What the report describes is a system under serious structural strain.

Rural communities are taking the hardest hit. When one pharmacy closes in a metro area, about 15% of the zip code’s residents had been receiving care there. When one closes in a rural area, it’s 54%. That’s patients displaced.

The vaccination data is especially striking. Among rural flu vaccine recipients whose pharmacy closed between 2021 and 2025, somewhere between 63% and 70% didn’t get a flu vaccine the following season.

What’s driving the closures? The report names pharmacy benefit manager (PBM) preferred network design, reimbursement that doesn’t cover the actual cost of dispensing branded drugs, vertical integration, and opaque payment models that disadvantage independent community pharmacies. These are the same structural issues that show up every time rural pharmacy access comes up in policy conversations, and they’ve been largely unaddressed.

The question is whether any of this lands in legislation before more communities lose their only option.

Reimbursement Fundamentals – Innovation is Not a Guarantee

Earlier this month, the Incubate Coalition released a survey of C-suite executives at early-stage, mostly pre-revenue biotech companies. These are small developers working in oncology, rare disease, and neuroscience, often with a single asset and investors who are watching the policy environment closely. The results gave me pause. We say we want innovation, but we seem to be shooting ourselves in the foot (feet?) from a policy perspective.

Sixty-five percent of small molecule developers said they would pursue additional indications if the IRA’s pill penalty were eliminated. Eighty percent named FDA instability as their top policy concern. Sixty-eight percent had heard directly from investors who cited FDA instability as a reason to pull back or reduce investment. Seventy-two percent said they would make meaningful operational changes if most-favored-nation (MFN) pricing becomes law. One in three had already lost funding because investors were spooked by MFN pricing threats.

The pill penalty is worth unpacking. The IRA set different negotiation timelines based on drug type: small molecules, basically pills, become eligible for negotiation seven years after FDA approval. Biologics, which are injectable or infused drugs, get eleven years. That four-year gap creates a structural incentive to develop biologics over small molecules, even when a pill would be clinically equivalent or easier for patients to take. Pills are cheaper to manufacture, better for adherence, and generally preferable for patients managing chronic conditions.

Early-stage biotech is not a mature industry with stable cash flows. It is a high-risk, capital-intensive ecosystem that is genuinely sensitive to investor confidence. Their investors are making decisions right now, in 2026, about which programs to fund based on what they think the pricing and regulatory environment will look like in 2034 or 2038 when a drug might commercialize. The pipeline consequence of capital flight is years away from being visible, which is exactly what makes it easy to discount.

When the survey says FDA instability, they mean something specific. Reviewer departures, budget pressures, and uncertainty around user fee reauthorization have made the FDA review environment less predictable than it was two years ago. For a company with a single asset and a development timeline built around a specific approval window, that unpredictability translates directly into cash burn, extended runways, and conversations with investors about whether the return still pencils out.

Biotech competes for dollars against every other asset class: real estate, technology, infrastructure, public equities. Investors, particularly the institutional ones writing the checks that fund early-stage drug development, are not sentimentally attached to pharmaceutical innovation. They’re looking at risk-adjusted returns. When policy uncertainty compresses the expected value of a successful drug launch, they move the money somewhere else.

The timeline makes this especially acute. A drug entering Phase I clinical trials today might reach patients in 2036 or 2038, if it works, if it clears FDA review, if the regulatory environment cooperates. Investors are making decisions right now about whether the return they might see in twelve years justifies the capital they’d deploy today. Pricing policy that hasn’t even been finalized is already factoring into that calculation.

There’s also an international competition piece that doesn’t get enough attention. Other countries are actively building biotech ecosystems, offering incentives, and creating favorable regulatory environments specifically to attract the investment, the companies, and the jobs that come with drug development. If U.S. policy makes U.S. biotech investment structurally less attractive, that problem doesn’t stay inside U.S. borders.

We say we want cures for cancer, for Alzheimer’s, for rare diseases that currently have no treatment options. Those cures require someone to fund the companies trying to develop them, for a decade or more, with uncertain odds of success. If the policy environment makes that investment less competitive relative to other options, we get less of it. That consequence won’t show up in any data set for another ten years. By then it will be too late to course correct.