Newsletter · · Ashutosh Agarwal

The 100% Drug Tariff Is Now Days Away as 340B Clawbacks Begin - Drug Pricing & IRA Round 2 - Week of September 21, 2026

Drug Pricing & IRA Round 2 for the week of September 14 to September 21, 2026. Podcast synthesis on the 100% Section 232 tariff on branded drugs going live at the end of September with exemptions for deal-signers, drugmakers using new claims data to claw back 340B duplicate discounts from safety-net hospitals, an activist investor at Novartis, Lilly's legal win against a compounder, and FDA wins for Merck and Lilly.

Drug Pricing & IRA Round 2

Week of September 21, 2026: The 100% Drug Tariff Is Now Days Away as 340B Clawbacks Begin

TL;DR

  • The single biggest live event on this beat is a calendar entry: the 100% tariff on branded, patented drugs starts "at the end of September," roughly a week out. For pharma, the old baseline tariff was zero, so this is a genuine step-change, not a tweak. Companies that cut deals with the White House (agreeing to "most favored nation" pricing plus building factories in the US) get an exemption or a lower rate; everyone else eats the full 100%. Source: PwC's accounting podcast (specialist advisers).
  • A quieter but real earnings story: drugmakers are quietly clawing back money from safety-net hospitals. New data pipes now show manufacturers exactly which prescriptions were filled under the 340B discount program, and they're firing off "good faith inquiries" demanding repayment for discounts they say they shouldn't have paid. This is the 340B gross-to-net fight turning from theory into cash. Source: 340Banter Podcast (industry consultants).
  • Novartis's bad month got worse: an activist investor showed up. After roughly $40B of market cap vanished on back-to-back trial failures, Artisan Partners wrote a letter attacking the board over its dealmaking and pay, and declared the company's "party" over. Source: Biotech Hangout (analyst panel).
  • Eli Lilly's war on knock-off weight-loss drugs cleared a legal hurdle. A California judge let Lilly's "corporate practice of medicine" claims against a compounding telehealth outfit proceed. That enforcement campaign is one reason realized GLP-1 prices are firmer than the cash-pay headlines suggest. Source: On The Pen (GLP-1 pharmacist commentator).
  • Merck and Lilly both had good news at the FDA. Merck's Keytruda-plus-cancer-vaccine combo hit in melanoma, and Lilly won an oral breast-cancer approval. For Merck especially, that's fresh ammunition against the "what happens when Keytruda's patent runs out" worry that sits at the center of its pricing exposure. Source: Telltales (market recap).
  • The action this week was the tariff clock ticking down, plus name-specific stories at Novartis, Lilly and Merck.

What's New

1. The 100% drug tariff is a week away, and here's exactly how it works.

The one dated catalyst everyone should have circled: the 100% tariff on patented, branded drugs. On PwC's accounting podcast ("Navigating tariff uncertainty," Sept 17), the firm's tariff specialists (advisers who help companies actually calculate and file this stuff, so insiders on the plumbing) spelled out the mechanics.

The headline number, in their words:

"One of the more notable in Section 232 is the pharmaceuticals... that was 100%, which will be starting on certain products at the end of September."

Why this is a bigger deal than it sounds: pharma has historically paid no import tariff at all.

"The MFN level for pharmaceutical was zero. So this is a very notable progression."

("MFN" here is trade jargon for the standard baseline tariff rate. Confusingly, it is the same three letters the administration uses for "most favored nation" drug pricing. Different concept, same acronym.)

The escape hatch is the deal. Companies that went to the White House and agreed to price their drugs at international levels and build US factories negotiated their way out:

"Certain companies went in and said, we will agree to certain MFN pricing... as well as doing onshoring, producing in the U.S.... and thereby were able to negotiate lower rates or no rate, you know, kind of an exemption from the rates for a period of time."

Two more things the specialists flagged that matter for the book:

  • The terms are public. The deal specifics "are in the annexes of the agreements," so as more companies sign, we'll be able to read exactly what each one gave up.
  • Generics may be next. Right now it's "patent-in products," but there's active "conversation whether generic products will also be brought into the fold." That would drag the whole generics complex (think Teva, Viatris) into a tariff fight it has so far mostly dodged.

Separately, they're watching a second, slower threat: a Section 301 "excess capacity" investigation that names pharma manufacturing in Ireland, Switzerland, Singapore and Germany specifically. It "began last year. We still don't have the results, although there were rumors that they may be coming out soon." For companies with heavy manufacturing in those countries, that's a second tariff shoe that hasn't dropped.

Why it moves numbers: the tariff is the stick that forced foreign pharma to accept US-style price controls. It goes live in days. Signers are shielded; hold-outs are exposed. And the "generics too?" question is a live overhang for the low-margin generic makers.

2. 340B: manufacturers now have the receipts, and they want their money back.

The 340B program forces drugmakers to sell to safety-net hospitals and clinics at steep discounts. For years, manufacturers suspected they were being double-billed (paying a 340B discount and a separate Medicaid rebate on the same pill) but couldn't prove it. That's changed, and on 340Banter Podcast ("Good Faith Inquiries," Sept 17), two 340B consultants (Chelsea and her co-host, industry advisers who manage these programs for hospitals) explained what flipped.

The key change is data:

"The manufacturers have more data to go off of than they used to... now they're getting a lot more of that information with the medication prescription transaction information going through 340B ESP and now TRUSO. They know exactly which claims are 340B. And when they get that chargeback information from the MCO plans, they can actually see where those duplicate discounts are occurring."

The result is a wave of "good faith inquiries," formal letters demanding hospitals justify a claim or pay it back:

"The manufacturers are coming back and requesting repayment for those prescriptions because they were Medicaid claims, managed care Medicaid... and the covered entities did not submit the 20 codes on them and they are 340B eligible."

The consultants' advice to hospitals was blunt: don't ignore these, because a non-response can escalate into a manufacturer audit and a "pretty significant repayment." Their read on the balance of power is telling: even when a hospital did nothing statutorily wrong, "it can amount to a pretty significant repayment to a manufacturer."

Why it moves numbers: this is the gross-to-net story getting concrete. Every dollar of 340B leakage manufacturers recover is a dollar back to branded gross margin, and a dollar out of the hospital systems and contract pharmacies that have leaned on 340B spread. It's slow, it's unglamorous, and it's real money.

3. Novartis: after the wipeout, the activists arrive.

We flagged Novartis's brutal early September last week: roughly $40B of market cap erased on three back-to-back failures. This week Biotech Hangout (Ep. 195, Sept 15, an analyst/investor roundtable of sell-side and buy-side voices, so pundits, but well-sourced ones) added the sequel: an activist investor.

The panel walked through the trigger. The big one was the LP(a) heart-drug trial (LP(a) is a genetic cholesterol-type risk factor; the drug, pelacarsen, was meant to be the first to prove lowering it saves lives). The ~8,000-patient study "was completely a negative study," dropped "on Friday, literally at 4:30 p.m. after the market closed into the Labor Day weekend," timing that told everyone it was bad before they read a word. Novartis fell "about 14%."

Then came Artisan Partners:

"Artisan Partner... essentially wrote a letter raising concerns about board oversight of acquisitions, whether they need... stronger board talent in an acquisition committee. Also advocating for compensation overhaul... and ultimately called the party sort of quote unquote being over and really criticizing the chair."

The specific sore point is Novartis's ~$12B purchase of Avidity, far above its usual "$5 billion and below" bolt-on deals, whose lead program also just failed. The panel's own firm "removed about $5 billion in sales from the model" and "shaved off the EPS [CAGR] by about 200 basis points to 4%" long-term. The gut-punch: those cuts land "at the time that they needed these revenues to be there when they're facing loss of exclusivity."

Why it matters: Novartis is one of the large-cap pharma names most exposed to upcoming patent cliffs, which is exactly why its dealmaking matters, and now its M&A strategy is under public fire right as it needs deals to work. There's also a clean read-through: pelacarsen cut LP(a) by ~80%, while Amgen's rival (olpasiran) and Lilly's version cut it "95 to 100." Whether that extra potency rescues the category is the next big test for both.

4. Lilly's crackdown on knock-off weight-loss drugs clears a legal hurdle.

A recurring theme on this beat: the gap between scary GLP-1 cash-pay headlines and what companies actually realize per prescription. A big reason that gap exists is Lilly and Novo aggressively suing the compounders and telehealth firms selling cheaper, non-branded copies. On On The Pen (Sept 15, hosted by a GLP-1 pharmacist who tracks this litigation closely), the host detailed the latest ruling in Lilly v. Mochi Health:

"On September 8th, Judge Jacqueline Scott Corley partially granted and partially denied the defendant's latest motion to dismiss... the judge allowed Lilly's California unfair competition law theory against Equida involving allegations around the corporate practice of medicine to continue."

In plain terms: Lilly's claim that a compounding pharmacy and telehealth operation were mass-producing "individualized" prescriptions (which isn't allowed) survives, and Lilly now gets discovery into their internal communications. The host's sharper point is the uncomfortable mirror it holds up to Lilly's own LillyDirect platform, where listed telehealth partners "still receive a massive commercial benefit of being presented to patients." Where exactly the line sits between a "commercially valuable relationship" and "improper influence over medical care" is the question the whole direct-to-consumer model now hinges on.

Why it matters: every compounder Lilly shuts down protects branded Zepbound volume and price. But the LillyDirect scrutiny cuts the other way: the same legal theory could eventually be pointed back at Lilly.

5. Merck and Lilly get FDA wins that speak to their pricing exposure.

On Telltales (Weekend Update W2638, Sept 20, an AI-produced market recap, so treat it as a pundit summary), the FDA highlights tied straight to two of our most pricing-exposed names:

  • Merck had "the best of it": its Keytruda-plus-Moderna-cancer-vaccine combo "hit both its primary and key secondary endpoints" in resected melanoma, and the FDA expanded the label on Merck's Winrevair in pulmonary hypertension. As the show put it, "for a company that has spent three years being asked what replaces Keytruda when the patents run out, that's two answers inside seven days."
  • Eli Lilly won approval for Inlurio, an oral breast-cancer drug (an "estrogen receptor degrader") paired with Verzenio, "which pushes Lilly further ahead in the oral CERD race."

Why it matters: Merck's entire pricing risk is concentrated in Keytruda, its mega-blockbuster facing both patent loss and IRA/formulation exposure. Anything that credibly answers "what comes after Keytruda" directly de-risks that story.

The Debate

The core argument on this beat hasn't changed (a bounded, well-modeled headwind versus a structural squeeze on US branded-pharma margins), but this week's evidence tilted toward "it's real, it's just slow."

"It's manageable, and reversible."

  • The tariff's whole design is voluntary: sign an MFN-pricing-plus-onshoring deal and you're exempted or get a lower rate (PwC). Most of big pharma has signed.
  • The 340B clawbacks and tariff exemptions are administrative and negotiated, not a permanent law; they can be unwound.
  • The drama has moved from a "signing spree" to "grinding implementation," with no new MFN mega-deals or IRA negotiation headlines this week.

"It's structural: the machinery is now live."

  • The tariff is no longer a threat on a whiteboard; it starts in days, with a base rate that jumped from 0% to 100% (PwC). Once that's the world, it's hard to go back.
  • The 340B repayment wave shows the enforcement infrastructure (ESP, TRUSO data feeds) is built and working; manufacturers now have the receipts and are collecting (340Banter).
  • The Section 301 excess-capacity probe (Ireland, Switzerland, Singapore, Germany) is a second, unpriced tariff threat still working through the system.

Net: this week argued that the near-term P&L is still contained, but the plumbing of price control keeps getting more built-out and harder to reverse. The bear case doesn't need a bad quarter; it needs the tariff to stick, generics to get pulled in, and the 340B clawbacks to compound.

Stocks in Play

Eli Lilly (LLY)

  • Bull: Two wins this week: the Inlurio oral breast-cancer approval, and a legal ruling advancing its campaign against knock-off compounders that protects Zepbound economics (On The Pen, Telltales). An MFN signer, so tariff overhang is largely lifted. Its next-gen combo (tirzepatide + the amylin drug eloralentide) delivered 17% weight loss in just 16 weeks in an early cohort, versus 10% for tirzepatide alone.
  • Bear: The same LillyDirect telehealth model it's using to sue rivals could face the identical "corporate practice of medicine" scrutiny. Its LP(a) drug shares the category cloud hanging over Novartis's failure.
  • Watch: Whether the compounding-enforcement momentum holds; LP(a) trial design and readout; the eloralentide combo's later data.

Novartis (NVS)

  • Bear: The one to avoid being long into: ~$40B wiped out, a fully negative LP(a) trial, a failed ~$12B Avidity bet, and now an Artisan Partners activist letter attacking the board, its dealmaking and its pay, declaring "the party over" (Biotech Hangout). Sell-side models are cutting ~$5B of sales and ~200bps off the long-term EPS growth rate, right as loss-of-exclusivity looms.
  • Watch: Whether management concedes anything to the activist; any change in M&A discipline; the next LP(a) readouts from Amgen and Lilly (which could either validate or bury the whole target).

Amgen (AMGN) (read-through)

  • Bear/Watch: Its LP(a) drug (olpasiran) is the next big test of a target Novartis just failed. The one edge: it lowers LP(a) by "95 to 100%" versus Novartis's ~80%; the bull case is that potency is the difference. Data is the catalyst.

Merck (MRK)

  • Bull: Its biggest single risk is Keytruda's patent cliff and pricing exposure, and it just posted "two answers inside seven days" on what comes next: a positive Keytruda-plus-cancer-vaccine melanoma trial and a Winrevair label expansion (Telltales).
  • Watch: Whether the Moderna vaccine combo advances toward approval; Keytruda subcutaneous conversion; IRA/formulation exposure timing.

Novo Nordisk (NVO)

  • Bull: Still the volume leader in oral GLP-1s; the broader market remains wildly underpenetrated. One telehealth operator pegged it at "only about 2%" of eligible patients actually on the drugs (Hims House).
  • Bear: As a foreign manufacturer, it's squarely in the tariff frame unless fully shielded by its deal; the same compounding/cash-pay pressure that hits Lilly hits Wegovy.
  • Watch: Tariff exemption specifics; cash-pay price trajectory (see Read-throughs).

Tariff-exposed branded complex (JNJ, PFE, BMY, ABBV, AZN)

  • Setup: No name-specific news this week, but all sit under the same two clocks: the 100% tariff going live in days (most acute for heavy ex-US manufacturers like AstraZeneca) and the 340B clawback wave (a modest gross-to-net tailwind for every branded name). The unpriced wildcard is whether generics get pulled into the tariff and whether the Section 301 probe hits Ireland/Switzerland manufacturing footprints.

Read-throughs

  • Hospitals & safety-net systems (the other side of 340B): The manufacturer "good faith inquiry" wave is a direct hit to hospital drug economics. Health systems that carved Medicaid claims into 340B without the right codes are now getting repayment demands they mostly can't win (340Banter). Read it as a slow bleed for 340B-dependent providers and a slow tailwind for branded manufacturers.
  • PBMs / managed care (CVS, CI, UNH): The same data transparency (ESP, TRUSO) that lets manufacturers spot duplicate discounts also shines a light on the rebate plumbing PBMs sit on. More manufacturer visibility into claims is a structural negative for opaque middleman margin.
  • GLP-1 cash-pay trajectory (NVO, LLY, and HIMS as a read-through): On Hims House (Sept 16), OpenLoop's CEO (an operator running clinical infrastructure for telehealth) agreed with Hims CEO Andrew Dudum's call that GLP-1 cash-pay prices fall to "$40 to $50 per month" by around 2030, and thinks it "could come down even lower." That's the long-term pricing gravity under the branded franchises, offset near-term by the enforcement campaigns keeping copies off the market.
  • Peptides, a new adjacent market forming: The same episode sized the gray market for non-GLP-1 peptides at "$6 to $8 billion" today, with an FDA advisory committee having voted 6-of-7 to move several toward legal status, possibly before year-end. Wall Street estimates range from Needham's "$3 billion right now but $33 billion eventually" to Leerink's ~$2B. Samsung's ~$2B purchase of contract manufacturer Polypeptide Group is the tell that big players see it coming. One to watch for the DTC health platforms (HIMS), not yet a pricing-policy story.
  • Generics / biosimilars (incl. TEVA): The live risk got sharper. PwC flagged active "conversation whether generic products will also be brought into" the 100% tariff (PwC). If that happens, the low-margin generics complex has the least room to absorb it.

What Changed vs. Last Week

More company-specific. Last week was thick with policy: MFN carve-outs (Incyte, BridgeBio), the GLOBE/GUARD Medicare demos slipping to January, Dr. Oz conceding the deals are temporary. This week the policy machine mostly just kept grinding toward its deadlines, and the fresh news was at the ticker level.

  • The tariff moved from "scheduled" to "imminent": last week it was a September 29 date on the calendar; this week it's days away, with the exact mechanics (0% → 100%, exemptions for deal-signers, a live question over generics) laid out.
  • 340B advanced from policy fight to cash collection: last week's story was the $100B program and the rebate-model pilot; this week it's manufacturers actually issuing repayment demands using new claims data.
  • Novartis gained an activist: last week was the ~$40B, three-strike wipeout; this week Artisan Partners turned it into a governance fight.
  • New name texture: FDA wins for Merck (Keytruda combo) and Lilly (Inlurio); a legal win for Lilly against compounders; a fresh look at the peptide market.

The dated markers to keep watching: end of September (the 100% tariff goes live), the pending GLOBE/GUARD Medicare demo announcements, the still-unreleased Section 301 excess-capacity results (Ireland/Switzerland/Singapore/Germany), January 1, 2027 (340B changes and, likely, the Medicare MFN demos), and the midterms as the switch on whether any of this becomes permanent law.