Summary

Monday’s MFN executive order was, in John Maraganore’s words, “not a most-favored-nation executive order, but a mostly fluff-and-nonsense executive order.” It sets a 30-day voluntary period while HHS studies prices, followed by 180 days of negotiations. Several pharma shares recovered above their beginning-of-week levels. John still sees Trump pursuing MFN, potentially through an IRA-related mechanism or by applying it only to drugs not yet launched.
Sam Fazeli’s structural diagnosis is the keeper: the US is “the only healthcare system in the world where basically no participant who has any control also wants prices to be low.” Providers prefer high prices, ACA medical-loss-ratio rules limit insurers to about 15% margin, and the government is heavily restricted in negotiating drug prices. Matt warns that MFN done wrong could produce higher US prices if companies abandon ex-US pricing and try to recover the difference; he also worries biotech, focused on future drugs, will be underrepresented.
No implemented government-scale MFN is expected by end-2025. John sees no price pressure within the year; Matt expects superficial negotiations, especially where pricing pressure is already coming in obesity. A CMMI demonstration is the only executive-order route Matt identified, but the 2020 version was challenged and failed largely on process grounds. Sam’s colleague Dwayne Wright suggested the IRA’s price ceiling/no-floor structure could let CMS claim a win in November; Nina is skeptical because the administration is loath to use a Biden-originated pathway. John floated bilateral trade commitments, while Sam’s European colleagues see little enforceable mechanism.
The roughly $234B of US investment pledges — J&J’s $55B down to Merck’s $9B — are bargaining chips with escape hatches. Sanofi’s $20B through 2030 includes substantially increased ongoing R&D and unspecified manufacturing expansion, while saying investments may be adjusted as conditions change. Roche said that if the EO takes effect, its announced US investments “will be in question.”
Nina Kjellson floats an “AI reckoning?” — question mark deliberate — off Recursion discontinuing five drugs amid a pipeline reorganization, Insilico’s renewed Hong Kong listing effort, and C4’s 60-person, 22% reduction in force. Matt compares pharma’s AI messaging with its US-investment messaging; Roivant does see promise in tools such as protein–protein-interaction models. John says the industry is still “far away” from typing in a disease and getting a curative small molecule from a computer. Sam compares the moment with the late-1990s SNPs hype around Genset’s 100,000 sequences.
CMS’s IRA Part B guidance treating hyaluronidase subQ reformulations as the same drug hit Halozyme, J&J’s Darzalex and Merck’s Keytruda subQ. Matt and John agree that the same active ingredient is scientifically the same drug; John proposes a “super-generics” category that earns a convenience and system-savings premium without NCE economics. The outcome will become clearer in 2028.
Below-cash names can resurrect: CytomX rose from $0.43 on April 7 to about $2 on CX-2051’s late-line CRC data and immediately raised $100M, while Galapagos is Matt’s “very beautiful picture of the laws of unintended consequences” — cash trapped between a public biotech and a Gilead affiliate after its attempted cell-therapy restructuring. Deal tape included GSK’s $1.2B-upfront MASH purchase from Boston Pharmaceuticals, AbbVie’s first RNAi deal with ADARx for $335M upfront, and a less-disclosed, apparently RNA-editing deal between Lilly and South Korea’s Rznomics.

Deep dive

1. The MFN order: “mostly fluff and nonsense” — but the negotiation is real

John’s verdict on Monday: “not a most-favored-nation executive order, but a mostly fluff-and-nonsense executive order” — vague, process-heavy and nonspecific, a “nothing burger” after a weekend of anxious chatter. That is why several pharma shares recovered above their beginning-of-week levels. He still thinks Trump will pursue “some type of MFN mechanism,” but potentially through industry-tolerable routes: inside the IRA once negotiations are recalibrated to 13 years, or applied only to drugs not yet launched.
Sam’s parsing of the order was a 30-day period for voluntary proposals while HHS studies some drug prices, followed by 180 days of negotiations or conversations. He linked that timing closely to the next IRA negotiations and also flagged reimportation and direct-to-consumer sales.
Nina’s grief-stages framing: the industry is responding “with a mix of denial, bargaining and distraction,” and MFN has to be read alongside tariff threats, biosecurity, Section 232, IRA renegotiation and the Big Beautiful Bill’s health-care provisions. She thinks the administration is forcing pharma to the table while currying populist support; roughly $100B of committed US manufacturing and R&D from six of the top ten pharmas suggests that pressure is working.
Matt’s uncomfortable concession: “at some fundamental level the administration is right” — and the week’s gyrations are themselves evidence: “we are at the whims of one person’s tweets, mostly because US drug pricing is so disconnected from the rest of the world.” A painful reset could eventually buy a more “equitable and evenly distributed” growth path, but biotech’s longer-term interests may be underrepresented in negotiations led by big-pharma CEOs.

2. Nobody in US healthcare actually wants prices low

Sam’s signature framing is that the US is “the only health care system in the world where basically no participant who has any control also wants prices to be low.” Pharmaceutical companies, hospitals, physicians and providers prefer higher-price environments; ACA medical-loss-ratio rules cap insurers’ margins at about 15%, so their incentive is not to reduce underlying costs; and the government is heavily restricted in its ability to negotiate drug prices.
John widens the lens: the US/ex-US gap runs “up and down the health care system” — MRIs, CT scans, mammograms and essentially any procedure. Sam’s anecdote makes it visceral: a relative hospitalized for nine days with flu in San Francisco received a $250,000 bill — “you will never see that here” — while some European countries devote a higher share of their smaller health-care pie to prescription drugs.
Matt raised the possibility of Europe and other markets becoming somewhat like Medicaid, with confidential contracts and effective rebates that maximize gross-to-net discounts. Sam is dubious that companies could hide a large enough pricing loophole from the US government, particularly when the manufacturer is US-domiciled.

3. Handicapping 2025: gestures, not law

Sam’s direct poll — any price pressure by year-end? John: “I don’t see it happening.” Matt identified a CMMI demonstration as the only executive-order route, similar to the one proposed in November 2020; that effort was challenged and failed largely on process grounds, and a new version would face lengthy implementation and legal challenges. Nina likewise thinks outright price reform is unlikely, though broader business pressure will continue.
Sam relayed colleague Dwayne Wright’s view that the IRA sets a price ceiling but no floor, potentially giving CMS latitude to reference foreign prices or other factors in the November, 180-day process and declare some victory. Nina noted the administration may resist using a Biden-originated pathway.
Matt expects “superficial negotiations by big pharma companies trying to forestall worse policy changes.” Sam thinks the most likely grand gestures will come where “the writing is on the wall” anyway, especially obesity, where competition is already bringing pricing pressure. John floated bilateral trade commitments around drug pricing; Sam’s European colleagues see no clear mechanism to force Europe to raise prices.
Reimportation: John calls it “very low risk”; Matt agrees it is more likely to be a negotiating cudgel than a solution, but pushed back on the idea that manufacturers mechanically control all ex-US volume. If supply is available and price differences are large, enterprising people can move drugs across borders, subject to supply and country-level restrictions.
Direct-to-consumer pharma sales remain unlikely at scale. Nina said GLP-1 sales still face supply-chain and prescribing-referral issues, though companies such as Hims and Ro have corporate-practice-of-medicine structures already serving that need. Pharma may occasionally use DTC for ultra-specialty, orphan or rare-disease products and vaccines, but generally does not want the conflict risk of prescribing its own drugs.

4. Reconciliation-bill fine print: ad taxes, PBMs, orphan carve-outs

Matt’s caveat first — “anybody who tells you they can predict inclusion in current government legislative proceedings is lying.” The change from current-year R&D deductions to three-year amortization will be heavily lobbied. Eliminating the DTC advertising deduction may raise revenue, but increasing advertising costs by roughly 25% is unlikely to dramatically change behavior.
Nina cited Chris Murphy’s No Handouts for Drug Ads Act and what she believed was a Brookings analysis estimating that $1.5B–$1.7B could be recouped from roughly $6B in annual DTC advertising. Matt’s response: it may be a smart way to raise $1.5B, but much of the spending would continue.
Nina’s tell on PBMs: for the first time in her investing career, friends and family are directing sound bites at PBMs rather than treating pharma as “the only devil in the mix.” Possible reforms include banning spread pricing, requiring more rebate and contract reporting, and restricting integrated payer-PBMs from steering patients to their own mail-order pharmacies when those are not the lowest-cost option.
Exempting orphan drugs from IRA negotiations drew a “big win” from Sam and John; it is not yet law but appears to have bipartisan support. John also wants the orphan-drug tax credit for clinical expenses restored from 25% to its historical 50%. He flagged uncertainty over how Makary and Prasad will implement regulatory policy, since they do not necessarily see eye to eye.

5. $234B of pledges are bargaining chips, not capex

Sam’s tally runs from J&J’s $55B to Merck’s $9B, but the fine print deflates it. Sanofi’s $20B through 2030 includes substantially increased ongoing US R&D, an unspecified expansion of US manufacturing capacity, and the caveat that investment decisions will be adjusted as the external environment evolves. Cited job figures are often only 1,000–3,000 over five years.
Roche made the quid pro quo explicit: “Should the proposed executive order go into effect, Roche’s ability to fund the significant investments previously announced in the US will be in question.” Sam noted that funding would be difficult if policy reduced the cash flow supporting it. He also said tax relief may have been part of the lobbying behind some investment announcements; Nina added tariff concessions.
Matt grouped the announcements with what he called largely superficial guarantees of domestic investment. Productivity gains, including through AI and automation, could also allow companies to manage margins without adding as many US jobs as the topline pledge figures imply.

6. AI drug discovery’s reckoning — or its SNPs moment

Nina’s three datapoints: Recursion’s pipeline reorganization on its first-quarter call, including five discontinued drugs and several programs from the Exscientia acquisition six or seven months earlier; Insilico seeking another Hong Kong listing after a $100M Series E; and C4’s 60-person, 22% reduction in force. She still sees “huge, huge power” in the models, but asks whether companies can sustain organizations of that size and burn rate in this market.
Matt compared pharma companies talking about AI with pharma companies talking about US investment — a comparison Sam glossed as something companies feel obligated to say. Roivant has real efforts, including “a really good model for protein–protein interactions,” but Matt argues that the industry should stop treating capital-A, capital-I “AI” as a single thing. It is both a buzzword and a set of tools, and the two ideas get in each other’s way.
John agrees that computation and machine learning can improve discovery, clinical trials, simulations and other operations, but “we’re far away from being able to type in a disease of interest and push a button and have a small-molecule structure emerge from the computer that is the cure.”
Sam’s historical rhyme is the late-1990s SNPs moment: Genset had 100,000 sequences and hoped they would solve everything. The technology may eventually become an ordinary part of the system rather than a standalone story. Matt’s deadpan: “someday every person working at a biotech company is going to have a computer on their desk… and it will be plugged into the internet.”

7. CMS calls subQ reformulations the same drug — and John agrees

The IRA Part B guidance hit Halozyme hard and affected J&J and Merck: hyaluronidase subQ conversions such as Darzalex and Keytruda subQ would not receive a new negotiation clock. Sam questioned whether these products should instead count as new drugs; John and Matt said the same active ingredient makes that scientifically difficult to defend. As John put it, “call a spade a spade.”
John proposed a “super-generics” category: if a product is treated as substantially the same as the IV version, it could still earn a premium for convenience, reduced system burden and potential cost savings, but “probably shouldn’t be valued like an NCE.” Sam expects the practical test in 2028, when Darzalex or Keytruda may be affected.

8. Below-cash redemptions and a real deal tape

Galapagos — now reconsidering its prior plan to place the cell-therapy program under former CEO Paul Stoffels while spinning the rest into a newco, and bringing in Henry, most recently of Numora — is Matt’s “very beautiful picture of the laws of unintended consequences.” Onno van de Stolpe’s roughly $5B Gilead deal was designed to “forever secure” independence through US-rights obligations and standstills, and “one thing we can confidently say is that succeeded.” The result is cash trapped in an interstitial zone between public biotech and Gilead affiliate. Matt sees the tougher CBER stance on cell therapy as potentially contributory, but not the original cause.
CytomX is the counter-case: after three years below cash, it went from $0.43 on April 7 to about $2, touching $2.50, on Phase 1 data for the EpCAM-directed, topoisomerase-I-payload ADC CX-2051 in late-line CRC — a roughly $1.1B US third-line market — then immediately raised $100M. John gave credit to Sean McCarthy for sticking with the program. Matt’s generalization: the industry treats “investors as a class [as] the ultimate tastemakers,” yet many winners endured long periods of being violently out of favor before proving the market wrong.
BioMarin also announced an M&A deal involving a company close to or below cash. John did not know much about the target but saw it as James Sabry executing BioMarin’s business-development objectives and fighting for assets.
GSK paid $1.2B upfront plus $800M in milestones for Boston Pharmaceuticals’ FGF-21 drug efimosfermin alfa in MASH. Matt said the FGF-21 story should be important, with read-across to 89bio’s $514M enterprise value and Phase 3 drug versus Akero’s $2.4B valuation and efruxifermin; he called MASH “the new black.” He also described GSK’s HSD17B13 RNAi target as a phenomenal genetic target.
AbbVie made its first RNAi move with a $335M-upfront, multitarget partnership with ADARx across immunology, oncology and neurology. A separate, financially undisclosed Rznomics deal in South Korea appears to involve RNA editing for hearing loss, although John said the modality and economics were unclear; Lilly has pursued the area through gene therapy.

Bitcoin Above $100k, Is Alt Season Here?Episode 141 - May 9, 2025