Summary

Trump sent letters to 17 pharma CEOs demanding four MFN actions within 60 days, with no statutory or regulatory authority cited. Sam Fazeli’s triage: MFN for Medicaid is relatively easy, and direct-to-consumer sales could allow 30%, 40%, or 50% price cuts without materially hurting P&L because of insurer rebates. He finds repatriating ex-US revenue incoherent, while Washington analyst Dwayne Wright expects the policy ultimately to be implemented through harder IRA price negotiations around the September 29 target.
Investors are treating MFN much like tariffs, but the outcome could still be meaningful. Paul Matteis says investors are “still massively discounting this becoming a reality,” despite potentially major effects on companies with large ex-US price discrepancies, including Vertex and Neurocrine. Matt Gline warns that skepticism about a legal pathway underestimates the administration’s ability to use regulatory agencies to make life unpleasant.
Vinay Prasad is out at CBER after three months — “around eight Scaramuccis” — and investors viewed his removal as at least a near-term positive for stocks. Brian Skorney says Prasad reportedly overruled reviewers six times in about as many weeks and took aggressive positions, including in the Sarepta controversy, which attracted MAHA attention and, according to reporting, perhaps Trump’s. At Makary’s listening session the next day, Gline noted that Prasad’s name never came up in three hours while CDER–CBER alignment under Tidmarsh was discussed explicitly.
China now accounts for 30% of global clinical-trial starts versus 35% for the US, and Brad Loncar predicts “something politically big” within two to three years. His Apple analogy is that biopharma is outsourcing invention rather than low-value manufacturing: GSK’s China-sourced drugs would be labeled “designed by Hengrui in China.” Gline’s pushback is that generating many novel CD19 T-cell engagers and Phase 1 data has become a commodity, so blocking China would not reverse that commoditization.
GSK–Hengrui was the third-largest deal ever by buyer dollars and functions as an options book on fast, inexpensive Chinese Phase 1 development. Fazeli says GSK gets good human data, including in non-Chinese patients, for a fraction of the cost. Another panelist argued that a century-old science leader outsourcing work to a company that was recently a traditional generics maker shows the structural problem. Elsewhere, BMS put a mostly SLE Phase 3 TLR7/8 program into a Bain SpinCo, and AbbVie’s rumored roughly $1 billion Gilgamesh deal further validates psychedelics after J&J’s Spravato.
Shorting launches has become more dangerous. Paul says small and midsize companies are better resourced, more sophisticated, and better at managing expectations: Alnylam beat high expectations for vutrisiran in ATTR cardiomyopathy, Neurocrine’s Crenessity approached 1,000 patients starting therapy, and Verona and Madrigal also had strong launches. Josh says the change is multifactorial and includes better operational execution, not just expectation-setting.
Alzheimer’s diagnosis may be helped by blood-based biomarkers — up three-fold in a year per Biogen — while the upcoming Lilly AHEAD readout is the wildcard. Leqembi finally outpaced expectations, but Sam says growth remains fairly linear because of diagnosis and neurologist capacity. Daphne notes AHEAD’s 85%–90% screen-failure rate and questions whether it could double the class rather than expand it fivefold.

Deep dive

1. Trump’s MFN letters: two easy asks, two that don’t parse

The setup: letters to 17 CEOs with a 60-day deadline and four demands — extend MFN pricing to Medicaid, guarantee MFN for newly launched drugs across Medicare, Medicaid and commercial markets, repatriate increased ex-US revenues, and sell direct-to-consumer at MFN — with no statutory authority cited. Sam’s triage: Medicaid is “a relatively easy lift,” and a direct-to-consumer mechanism could work because pharma “should be able to cut prices without really hurting their P&L by 30, 40, 50%” given rebates surrendered to insurers.
The middle two are Sam’s problem. New-launch MFN at least isn’t retroactive (“it says newly launched drugs”), though a cheaper new GLP-1 raises questions for incumbents; repatriation he “can’t get my head around” — what’s the incentive to push ex-US prices up if you must send the money back? “That sounds like a not well thought through process,” especially while demanding US manufacturing and R&D investment.
Brian’s take: conceptually MFN isn’t terrible given US–UK disparities, but the administration’s superficially good ideas are “so aggressive and not necessarily well thought out, they create much more chaos.” Regeneron’s example from its call: it doesn’t set Eylea’s ex-US price — Bayer does — and Bayer has no incentive to raise it. Brian also noted Trump crossed out formal titles and hand-wrote first names, missing “the ultimate power move” of writing the wrong ones.
One panelist’s read of the subtext was that bipartisan support for drug reform is the threat: “play ball with us because if you don’t play ball with us, it’s going to be potentially much uglier.”

2. Enforcement risk vs. market complacency — and what Daphne heard in DC

Paul’s positioning check: investors are “still massively discounting this becoming a reality,” trading the tariff template — a first offer that “decimated the stock market” followed by something “considerably more benign.” The letter day saw pressure but “wasn’t at all like day one Liberation Day trading,” even though the policy could be highly meaningful for companies with large ex-US price discrepancies.
Gline’s rebuke of the legal-framework skepticism: “if the administration got really serious about this, they would absolutely be able to find tools to make life unpleasant, including via the regulatory agencies.” Another panelist noted that the discussion often ignores China, where prices are a fraction of Europe’s.
Sam’s postulated workaround: launch a new drug at $600,000 instead of $300,000 in the US, wait the roughly two years to European launch, and bank enough early to cover the later step-down — “kind of like the practice in Germany,” per another panelist. Gline: in that band companies already do this — a drug such as Vyvgart can be priced roughly at US parity in Europe while the company effectively gives up on European access, as argenx’s sales mix illustrates.
Daphne’s DC readout after meeting administration advisers and Dr. Oz: the administration is “actually looking to collaborate,” wants industry proposals, and CMS is targeting middlemen — PBMs and 340B misuse — noting one path to MFN is eliminating the US middleman layer that doesn’t exist ex-US. Her verdict: directionally set but “not written in stone,” so wait-and-see “is probably correct.”

3. China: commodity engineering or outsourced innovation?

Daphne’s numbers frame the stakes: China is 30% of global clinical-trial starts versus the US’s 35%, and Chinese assets were 24% of the global pipeline in 2023, up from 2% a decade earlier — with IP copying plus fast-track development the core threat. Her DC proposals: regulatory, reimbursement, and funding penalties tied to IP theft, and amending the tax code so US biotechs retain NOLs through dilutive financings and M&A.
Brad’s framing: every other industry outsources what doesn’t matter — “designed by Apple in California” — while biopharma outsources invention itself; GSK’s drugs “would say designed by Hengrui in China.” Politicians took years to grasp semiconductors and “we’re even more technical than that,” but once they understand, “they’re going to freak out.” Something big within two to three years unless relations improve — odds of that “practically zero.”
Gline’s pushback — worth keeping: “generating 50 CD19 T-cell engagers, each one a novel molecule… it just turns out that’s a commodity good now.” The US built an industry designed to treat that as the innovative thing, and even blocking China won’t stop others from doing it.
Brad’s rebuttal: China is “on thinner ice than we are” — 100% reliant on US success, with Chinese pharmaceutical innovation and drug pricing having “never materialized” as expected — but US capital is funding Chinese companies’ growth. With five people there for the cost of one here, a next phase comes where China out-innovates on truly new things and “it’s impossible to change the regulatory structure here to level that playing field.”

4. Prasad gone in eight Scaramuccis; Makary’s charm offensive continues

Skorney, who called the exit: Prasad reportedly overruled reviewers six times in about as many weeks, “tried to do too much too quickly,” and drew MAGA/MAHA attention and, according to reporting, perhaps Trump himself — with the Sarepta response (two DMD deaths, an LGMD death on a related product, and another Elevidys death via potentially related immunosuppression and infection) the flashpoint. His removal “has been viewed by investors as a net at least near-term positive for stocks,” with small, ambiguous datasets in tough indications possibly getting more leniency.
The attribution disagreement: Gline finds it “a little bit surprising” if Laura Loomer, the WSJ op-ed page and patient groups aligned and “Sarepta had nothing to do with it,” given Doug Ingram’s Botox-litigation and Elevidys history. Another panelist counters that Sarepta’s regulatory wins span three CEOs and points instead to MDA, PPMD, CureDuchenne, and individual mothers who have influenced ultimate decision-makers. A further panelist asks, “Between Peter Marks and Vinay Prasad, who even wants that job?” — prompting “George does, maybe.”
Gline attended Makary’s fifth listening session (~60 CEOs, New York, the day after the departure): “benign, helpful, favorable” — “if a third of what he said came to pass, it would be all good for the biotech industry.” Prasad’s name went unmentioned in three hours, and Makary explicitly pushed CDER–CBER alignment with Tidmarsh, freshly interim CBER head, sitting right there — a sequence that “almost seemed intentional.”

5. Policy grab bag: tariffs manageable, RFK churning, NIH spared

Sam on tariffs: August 1 brought 15% on the EU with pharmaceuticals ultimately included, and 35% — though he had also seen 37% cited — on Switzerland, hitting Roche and Novartis. Pharmaceutical inclusion for Switzerland remained uncertain, and the measures were not due to hit until the following week, so “there’s still time for these to be corrected.” The practical caveat is that small and mid-cap biotechs rely on CDMOs and can’t reshore, but 15% “is manageable.”
Josh’s RFK watch: restructuring the U.S. Preventive Services Task Force as too woke (“admittedly, they are quite woke. I don’t know what too woke means”) with fewer doctors and nurses, more dietitians and therapists; plus overhauling vaccine-injury compensation — a litigation bonanza that could bring the vaccines-autism fringe view into Supreme Court litigation.
One clean positive: Senate Appropriations rejected the administration’s proposed NIH funding cut and advanced roughly a $400 million increase — “good news for the biotech ecosystem if it plays out.”

6. Deal flow: GSK buys options, Bristol spins, AbbVie validates psychedelics

Sam sizes GSK–Hengrui as the third-largest deal ever by buyer dollars, behind Daiichi–Merck at $22 billion and the $12 billion, 40-target Roche deal he cited: GSK bought rights to assets Hengrui takes through Phase 1 — “good human data including non-Chinese patients for a fraction of the cost.” The structure highlights how quickly and cheaply Hengrui can perform early-stage development.
Another panelist’s counter-kicker: GSK has invented medicines “for literally over a hundred years”; Hengrui, a traditional generics maker that he would have ranked “in the bottom five or 10” of Chinese biotechs five years ago, has done it for four. That GSK outsources its science to them “really goes to show what a structural problem this is.”
Gline on BMS–Bain: a Phase 3 TLR7/8 program, mostly in SLE, goes into a SpinCo — Bristol keeps ~20%, Bain commits $300 million plus royalties back — the same structure Roivant ran with Pfizer on Telavant and Priovant. Even recovering pharmas face P&L pressure and “tough decisions”; creative structures that keep programs alive beat shelving them. A prior Bain deal was described as very successful.
Paul on the rumored ~$1B AbbVie–Gilgamesh deal: a second mainstream player going “all in on psychedelics” after J&J’s Spravato — tracking toward $2B+ despite cumbersome every-other-week dosing — validating Compass, GH Research and MindMed. Compass has a Phase 2 and a completed Phase 3 but not a second Phase 3; filing without completing that second study “would traditionally be a nonstarter,” though “if it’s ever going to change, it could be now.” Separately, Gline calls Galapagos “a box of cash that has some really complicated encumbrances on it” until the Gilead entanglements sunset.

7. Shorting launches gets more dangerous — and the mood has turned

Paul’s launch survey: small and mid-caps are better resourced, more sophisticated, and better at “managing down expectations” — Alnylam blew out numbers on vutrisiran in ATTR cardiomyopathy despite sky-high expectations, Neurocrine’s Crenessity in CAH has over a thousand patients close to therapy start versus “fairly small product” forecasts, plus Verona and Madrigal. Shorting launches “has become just a lot more dangerous,” and commercial stories now look stronger in his coverage than they historically did.
Josh says the sector is in a different era and points to the pipeline programs that still receive little value recognition after successful launches. In response to the suggestion that analysts and management have simply become more conservative, Brian says the change is multifactorial and includes altered operational execution, not just expectation-setting.
Brad’s closing mood check: “we’ve all been in a bad mood for a long time… things are going really well lately” — Alnylam through $50 billion in market cap, Madrigal launching well enough to buy assets, the XBI weighed down by “cleanup from the COVID bubble.” Sam’s pharma counterpoint: Novo cut sales-growth guidance from 14% to 7%, its lunch “eaten by compounders” and a very good Lilly drug. Gline’s close: commoditization will create interesting developments in new parts of the industry — “we’re watching the early or middle stages of some of that transformation.”

8. Alzheimer’s: blood biomarkers may ease diagnosis, AHEAD is the wildcard

Brian on Biogen: one of the first quarters he remembers in which Leqembi outpaced expectations — even unwinding a one-time China build — against a class that has been “a huge disappointment” versus five-years-ago hopes, weighed down by the Aduhelm controversy. The bottleneck is diagnosis and PET confirmation, and blood-based biomarkers — up three-fold over the past year per Biogen — may become the primary diagnostic mode, “getting around PET altogether.”
Sam adds Roche is “sticking its colours to the mast” on the trontinemab Brainshuttle and moving quickly into much earlier-stage Alzheimer’s — telling, from a company that “knows all about diagnostics.”
Sam’s view is that growth remains fairly linear, gated by getting patients to see a neurologist at all — “conceptually wild.” Lilly’s AHEAD study in presymptomatic patients could shift motivation psychology, but Daphne notes its 85%–90% screen-failure rate and questions real-world scaling: “maybe data from that study could double the size of this class, but I’m just not sure if it could 5x it in the way that some investors are starting to talk about.”

Did ETH Just Signal Alt Season?Episode 149 - July 25, 2025