Newsletter · 2026-09-27 · Ashutosh Agarwal

Pharma Tariffs Land Tuesday as China Takes Half of New Drug Candidates - Life-Science Tools Recovery - Week of September 27, 2026

A synthesis of what podcasts, journalists and adjacent operators said about the life-science tools group for the week of September 20-27, 2026, built around the up-to-100% pharma tariff that starts Tuesday, September 29, China's rise to 48% of licensed or acquired new drug candidates, and a sharper debate over whether AI in biology needs deeper sequencing data. No coverage-company executive was on tape this week, so the read leans on policy and structure, anchored to last week's operator guidance.

Life-Science Tools Recovery

Week of September 27, 2026: Pharma Tariffs Land Tuesday as China Takes Half of New Drug Candidates

TL;DR

• The biggest news came out of Washington, and it has a date: Tuesday, September 29. On BioCentury This Week, the show's Washington reporter said the government is "on track to impose tariffs of up to 100% on patented pharmaceuticals and their ingredients," starting September 29. (A tariff is a tax on imported goods.) He also said "nobody knows which products are going to be affected." Thirteen drugmakers that signed pricing and US-manufacturing deals with the government are exempt. Small and mid-sized biotechs are the most exposed, and those are the same "emerging biotech" customers Thermo Fisher credited last week for the pickup in demand. (BioCentury This Week)

• China now takes 48% of licensed or acquired new drug candidates, up from roughly zero five years ago. Ginkgo Bioworks' CEO put the shift at "130 billion and 48%" last year, a structural question for US lab demand. A STAT reporter filing from Shanghai described the sector as "basically a gold rush," with US venture investors setting up shop there. (Grow Everything Biotech Podcast · The Readout Loud)

• The AI-for-biology debate got sharper. Guardant Health's co-founder said "90% of that is probably overhyped," not because AI needs less data but because most cell-model data is "undersampled." His fix, deeper sequencing on more samples, points straight at sequencing volumes and sample-prep consumables. (Minus One)

• Ginkgo says its model is now "basically Thermo Fisher." After its partnership model collapsed in 2022 and it cut "half the company," Ginkgo now sells robotic hardware and "pay by the data point" CRO services, and is pushing fully automated "lights-out" labs. That is a new competitor-and-customer dynamic for the tools group. (Grow Everything Biotech Podcast)

• Biotech mood: indexes up, sentiment fragile. The XBI is up about 30% on the year, but investors say "the wrong stocks are working," an M&A chill has set in after Novartis's Avidity setback, and Moonwalk (run by Illumina's former CTO) closed an oversubscribed $70 million round. (Biotech Hangout)

• Honest flag: no sitting executive of the 15 coverage companies appeared on a podcast in the window, so there is no fresh bioprocessing order data, guidance update or sequencing-competition read. September's investor conferences are over and Q3 earnings do not start until late October. This week's signal is policy and structure, not orders.

What's new

September's investor conferences are over and Q3 earnings do not start until late October, so this week's signal came from policy, China and the view from inside biotech rather than from the tools companies themselves. No sitting executive of the 15 coverage names appeared on a podcast in the window, which means no fresh bioprocessing order data, no guidance update and no sequencing-competition read. What we did get was substantive. Ranked by what matters most for a portfolio:

1. The pharma tariff has a start date, and nobody knows the rules yet

On BioCentury This Week ("Ep. 388 - Biotech IPOs, CAR T safety & pharma tariffs," September 22), BioCentury's Washington reporter (called Steve on the show) described the policy plainly. This is expert journalist commentary, not an operator:

"Yeah, it's a mess. Look, the government is on track to impose tariffs of up to 100% on patented pharmaceuticals and their ingredients, APIs, starting on September 29th. That's what's supposed to happen. But nobody knows which products are going to be affected, which companies are exempt, which products are exempt, or how the process is going to work."

(APIs, or "active pharmaceutical ingredients," are the chemicals in a drug that actually do the work.) Here is what is known, according to the show:

• Thirteen companies are exempt. These are the companies that signed "MFN agreements and on-shoring agreements with HHS." MFN, or "most-favored nation," is a pricing arrangement that ties US drug prices to what other rich countries pay. Onshoring agreements are commitments to make drugs in the US.
• Generics and biosimilars are exempt "for now." Some countries face a 15% rate, some 20%, and others the 100% default.
• Several categories are exempt on paper: orphan drugs (for rare diseases), nuclear medicines, plasma therapies, fertility treatments, cell and gene therapies, and antibody-drug conjugates (antibodies that deliver a toxic payload straight to cancer cells). But the proclamation adds conditions. For example, a product may need to come from a country with a trade framework deal, or meet an "urgent U.S. health need." Lawyers disagree on whether existing products are exempt automatically.
• No guidance has been issued. The administration had promised rules before September 29. At the time of recording, the reporter said, "today is the 21st... who knows whether guidance will be issued between now and then?" He also raised President Xi's upcoming visit to Washington as another unknown.

He also relayed the industry's pushback. BIO (the biotech trade group) CEO John Crowley wrote to the Commerce Department warning that the tariffs "are going to disproportionately affect small and mid-sized companies," while "the big companies have gotten themselves exempt through the MFN and onshoring deals." The small companies that make their drugs "in China or in India and other places" will be hit hardest.

Why it moves numbers (our analysis): This cuts two ways for the tools group. In the long run it helps. Every onshoring deal behind those 13 exemptions is a pledge to build or expand US drug plants, and those plants have to be filled with bioreactors, single-use bags, filters and purification resins from Danaher's Cytiva, Thermo Fisher, Sartorius, Repligen and Avantor. In the short run it hurts. The companies left unprotected are the small and mid-sized biotechs. Last week, Thermo Fisher's CEO said demand from that group "really pick[ed] up." If they spend the next quarter sorting out tariff exposure instead of buying lab equipment, the recent biotech recovery could slow. Watch Tuesday, then watch whether exemptions get defined, delayed or dropped.

2. China's biotech boom, and the argument that America is exporting its lab work

Two podcasts from very different angles told the same story.

On The Readout Loud ("The FDA's Overton era, and dispatches from Shanghai," September 24), STAT's reporter Jason reported from a biotech conference in Shanghai. This is journalist commentary. He described a sector running flat out: "The competition for molecules here is as high as it's ever been... It's basically a gold rush out here." He quoted one Chinese biotech CEO: "U.S. companies cannot compete. We are too fast. We are too efficient." He also noted that US venture investors are setting up shop there, including RA Capital talking about "the firm's plans for China" from the 38th floor overlooking Shanghai. He explained the deal structure now in favor: Chinese biotechs license molecules into a new company set up by mostly US investors, which makes it "easier to then sell to a U.S. pharma company." He added one caution. Three patient deaths in fast-track "investigator-initiated" gene therapy trials led the Chinese government to "put some new controls on these kinds of trials." Even so, "it does not seem to be a major worry."

The sharper numbers came from an operator. On Grow Everything Biotech Podcast ("200. That's One Cell of a Machine: Jason Kelly of Gingko Automates Labs to Accelerate Discovery," September 25), Jason Kelly, CEO of Ginkgo Bioworks and former chair of the US National Security Commission on Emerging Biotechnology, said:

"We've gone from effectively 0% of acquired new drug assets and less than $2 billion worth back in five years ago, coming from China to last year, or 130 billion and 48%. Like crazy shift to moving the startup ecosystem to China."

His worry is that the US is offshoring the hands-on lab work behind biology: "the way we move those fields forward is empirical lab work, which is manual. It's hands. It's PhD hands today." China now has "a large number of low cost, technically trained scientific hands." He pointed to a proposed bill called BINSA, which would add biotech to the outbound-investment restrictions that already stop US investors from funding Chinese AI, quantum and drone companies. As it stands, "you can go fund biotech in China."

Why it moves numbers: This complicates the China story. Last week, Thermo Fisher's CEO came back from China saying "there's so much investment going on in pharma and biotech." This week confirms that Chinese biotech is booming. But the more drug discovery moves to Chinese labs, the more lab spending depends on whether Western tool companies win in China against local suppliers. That question got zero operator commentary this week. There is also a mirror-image risk for US academic and startup lab demand, which is the core market for Illumina, 10x Genomics, PacBio, Bio-Rad and Bruker. The BINSA proposal is the policy item to watch. If it passes, capital could flow back toward US labs.

3. Ginkgo says its business model is now "basically Thermo Fisher," and it is betting on robots

The same Ginkgo episode had a second point that matters directly to our group. Kelly described how Ginkgo's original model fell apart after 2022. That model was research partnerships paid in fees and royalties. When interest rates rose, startup customers "were not going to get another round," and big pharma companies "put their arms around their own R&D orgs" and cut outside partnering. Ginkgo "did a big layoff... half the company." His description of what it sells now:

"Increasingly the business model is basically Thermo Fisher. So life science tools, we sell our robotic hardware and we sell... our CRO services. Like we will just generate data for you. No royalty, no nothing. Pay by the data point."

(A CRO, or contract research organization, runs experiments for other companies for a fee.) He is pushing fully automated "lights-out" labs and argued the US has to automate its labs, because otherwise "I just don't see in the long run how we compete" with China's pool of low-cost scientists. He admitted he is "talking my book."

Why it moves numbers: Here, straight from an operator, is the trend that has quietly squeezed tools demand since 2022: startups couldn't raise money and big pharma pulled research back in-house. It also sets up a new competitor-and-customer dynamic. Automated labs still use reagents and consumables, probably more efficiently. But a "pay by the data point" service model competes with selling instruments to individual labs. It fits last week's theme, that AI in biology is short on data and data takes lab work. This week added a twist: more of that lab work may be done by robots.

4. Guardant's co-founder: most AI-for-biology hype is "overhyped" because the data is too thin

On Minus One ("The Blood Test That Could Detect Thousands of Diseases | Helmy Eltoukhy, Guardant Health," September 24), Helmy Eltoukhy, co-founder of Guardant Health, gave operator-level numbers on one of the largest sequencing-based diagnostics businesses. Guardant has tested about 1.3 million cancer patients, holds 700,000 epigenomic profiles "which I don't think anyone has," tracks about 500 labeled diseases in its database, and works with 200 pharma companies. (Epigenomics means reading the chemical marks that switch genes on and off. He compared DNA to "the hardware" of a cell and the epigenome to "a software program.")

His pushback on AI hype is worth quoting:

"I would say that 90% of that is probably overhyped in some ways... the underlying data that's collected to essentially look at those models is very undersampled."

His point is that many "foundation models for cells" rely on shallow data. "60% of a cell is like mostly housekeeping genes, and if you undersample it, all you're seeing is essentially this standard stuff." Guardant took the opposite approach, collecting very deep data, and says AI now gives it the bandwidth to analyze it. "Every single one of our R&D individuals is now a manager of multiple agents." The next step is using one blood test to screen for "hundreds, if not thousands of diseases" beyond cancer. He also described pharma work as "a leading indicator for us": buying signals from drug-company trials tend to come before demand from doctors.

Why it moves numbers: This is a helpful counterweight to last week's AI-needs-data optimism, and it points the same way. Eltoukhy's complaint is not that AI needs less data. It is that most AI-biology data is too shallow, and the answer is deeper sequencing on more samples. That is good news for sequencing volumes and sample-prep consumables, which is where Illumina and its challengers compete. It says nothing about which instrument vendor wins.

5. Biotech mood: indexes are up, sentiment is fragile, and the M&A pause is weighing

On Biotech Hangout ("Episode 196 - September 18, 2026," published September 25), a panel of biotech investors and analysts took the sector's temperature. This is pundit and investor commentary. The host noted "the XBI is still up about 30% on the year" (the XBI is the main biotech stock fund). But buy-side investors complain "the wrong stocks are working" while "the good names are languishing." One panelist, Paul, put it down to "some negative data readouts, some rate stuff... quiet on the catalyst front." Another, Josh, pointed to Novartis's setback after its large Avidity acquisition, which is "putting a chill in terms of appetite for large deals for assets that aren't fully de-risked." He also flagged that aggressive launch pricing ahead of the midterms could bring drug pricing back as an issue: "it just takes one tweet to roll the sector over." Adam's reply: "all it will take is one... big deal, M&A deal to kind of get people excited again."

On the next episode of Biotech Hangout ("Episode 197 - September 25, 2026"), a panelist identified only as Matt, who speaks for a company, said "we don't have an MFN agreement with the government today" and suspects "it's sort of coming for the mid-caps." Panelist Brian read this week's Senate hearing for FDA nominee Heidi Overton as broadly positive for industry. She talked about "speeding through innovation," raised competition with China, and called every US-marketed vaccine "safe and effective." His takeaway: the administration "is really trying to calm the chaos of the FDA."

There was one fresh funding data point. On Business Of Biotech ("Finding New Opportunities In Obesity With Moonwalk Bioscience's Alex Aravanis, M.D., Ph.D.," September 21), host Ben Comer noted that Moonwalk, run by Alex Aravanis, Illumina's former Chief Technology Officer and scientific founder of Grail, "just closed and oversubscribed $70 million Series B." The money will take its lead obesity drug through early human trials, with a second program close behind.

Why it moves numbers: Biotech funding flows into tools demand with a lag. Last week Thermo Fisher called it the "funding-to-revenue lag." A stock index up about 30% and oversubscribed rounds are still supportive. But a pause in M&A, rising interest rates and pricing worries ahead of the midterms are the kind of things that slow the next round of funding, and the lab spending that follows.

The debate

The bull (bioprocessing recovery / sequencing and multi-omics re-acceleration). Nothing this week undid last week's operator evidence: Thermo Fisher guiding to 4% growth in the second half, U.S. academic back to growth, China stabilizing, and "AI, more spending, not less." This week added some structural support. The tariff regime is pushing large drugmakers into US-manufacturing commitments, and every new US plant needs bioprocessing equipment and consumables. Biotech funding is still open, with the XBI up about 30% and Moonwalk's oversubscribed $70 million round. The AI-data argument got more precise: Guardant's co-founder says the problem is shallow data, which means more and deeper sequencing. Ginkgo's CEO calls his company "basically Thermo Fisher" and wants to automate US labs, which still consume reagents. And the FDA looks calmer, with a nominee talking about speed.

The bear (China / academic / tariff / lumpy-capex). The bear had the better week, and none of it came from our companies. (1) Tariffs: a 100% default rate arrives Tuesday with no guidance, and the pain lands on small and mid-sized biotechs, the customer group that drove the recent pickup. (2) Academic funding: a reported draft executive order would create a committee with power to "veto awards in the National Institutes of Health's vast research portfolio." Even secondhand, that reopens the funding fear that eased last week. (3) China: 48% of licensed or acquired new drug candidates now come from China, according to Ginkgo's CEO. The new-drug engine is moving to Chinese labs, where Western tool makers face local rivals and politics. (4) Still no order data: another week without a bioprocessing book-to-bill from anyone who reports one. (Book-to-bill is new orders divided by shipments. Above 1 means the order book is growing.)

Our read. After last week's bullish tilt, this week leans modestly bearish on the news, but it is a quiet week, not a turn. The new information is all policy and structure: tariffs, NIH governance, China. None of it is an operator telling us orders have changed. The right stance is to hold last week's operator-based view and treat September 29 as the next real test. Q3 earnings, starting in late October, are where the book-to-bill question finally gets answered.

Stocks in play

None of our 15 coverage companies had an executive on a podcast this week. Only two were named: Thermo Fisher (by Ginkgo's CEO, as the business model he is copying) and Illumina (in the background of Moonwalk CEO Alex Aravanis, its former CTO). Everything else below is read-through from this week's policy and industry news, labeled as such.

Also discussed but not investable here: Moonwalk Biosciences (private; $70M Series B), Amela Biosciences (private), Genezen (private, PE-backed CGT CDMO).

Read-throughs

Bioprocessing peers (Sartorius, Repligen, Avantor, Maravai, Danaher/Cytiva). No order data for yet another week. The main new factor is the tariff. The 13 exempt drugmakers got there through onshoring commitments, which means US biologics capacity that will need equipment and consumables. That helps in the long run, but the timing is uncertain. On the CDMO side, the Genezen finance leader (below) said clients who reserve dedicated capacity pay "some sort of an upfront payment." That structure protects the CDMO and can smooth demand for consumables.

Sequencing (Illumina, PacBio, 10x). No read on the competitive fight. The structural argument got stronger: Guardant's co-founder says AI in biology needs deeper data, and that means more sequencing per sample. The risk comes from Ginkgo's China figures. If new drug discovery keeps moving to China, US academic and startup sequencing demand, a core market for PacBio and 10x, faces a long-term headwind.

CDMOs and outsourcing. Several operators confirmed that outsourcing is how small biotechs work now. On Smart Biotech Scientist ("290: Batch-to-Batch Variability Is a Design Failure, Not the Norm with Sandra Núñez - Part 2," September 24), Sandra Núñez, CEO of Amela Biosciences, described a fully outsourced model. Her company "owns no factory, no cleanroom, no fermenters," as the host put it. She called outsourcing "a very agile model" for "a startup with like little money." She said the biggest tech-transfer risk is "not... the technical part" but "the human part." She also noted that some purification steps "are just not scalable because of the bed size" once you plan for larger volumes. On FP&A Today ("The Real Job of FP&A: 40% Numbers, 60% Influence," September 22), a finance leader at Genezen, a PE-backed cell and gene therapy CDMO, said clients are "essentially buying our capacity." The firm plans for "a 20%, 30% threshold of failures," and a client funding collapse "really sucks" for revenue. On policy, Take as Directed ("Monisha Ashok, Director of Health Investments, U.S. International Development Finance Corporation (DFC)...," September 22) featured Monisha Ashok of the DFC, a US government development-finance agency. She said reshoring drug-ingredient production "is now a top U.S. foreign policy and national security priority." She noted "China dominates a lot of the kind of CDMO market," and said DFC has "spoken to a number of the leading CDMOs" about financing expansion outside China. The host mentioned the BIOSECURE Act, passed last December, which will produce a list of Chinese contract manufacturers US companies should avoid. That is a tailwind for non-Chinese CDMOs, and for the Western tool makers that supply them.

Cell and gene therapy (CGT) demand. Still no operator read on reagent volumes, but a clear sign that manufacturing now drives deal value. On Cell & Gene: The Podcast ("Structuring CGT Deals for the Long Term with Troutman Pepper Locke's EA Weeks, J.D., Ph.D.," September 24), EA Weeks, counsel at Troutman Pepper Locke (a deals lawyer, so an expert, not an operator), said "we're seeing major milestone payments tied to CMC events, not just traditional clinical phase one, two readouts." CMC means chemistry, manufacturing and controls: proving you can make the product consistently. Moving a CGT process to a new site "can take years." On payment, on Business Group on Health ("Can We Afford the Future of Medicine? The Promise and Price of Cell and Gene Therapies," September 21), an executive from ICER (an independent drug-price watchdog) said about 70% of the 13 cell and gene therapies it reviewed were fairly priced for their value. That is the reverse of drugs overall. The executive added that the pipeline "is still pretty robust" and "there is at least more predictability" at the FDA. Cell and gene therapies being on the tariff exemption list (with conditions) is a small plus.

Academic / NIH funding. This is the most important reversal from last week, with a big caveat about the source. On Paging America ("RFK Jr.'s Chemtrails, Overton's FDA Bid, and Dr. Vin Gupta on What Comes Next," September 24), a political advocacy podcast (clearly partisan, not primary research), the hosts relayed a New York Times report. It said the White House "has begun drafting a new executive order" to "create a new external committee with the power to veto awards in the National Institutes of Health's vast research portfolio." Per Politico, the board would include NIH director Jay Bhattacharya and budget director Russ Vought and "would be required to reach unanimous decisions." The hosts also relayed that Bhattacharya refused a directive to cut Harvard's funding. This is secondhand, and an order is not a funding cut. But it adds a new layer of uncertainty over grant approvals, and grant uncertainty tends to delay lab purchases. Relevant to every academic-exposed name: Illumina, 10x, PacBio, Bruker, Agilent, Bio-Rad, Avantor, Thermo Fisher.

China-exposed instrument names (Thermo Fisher, Danaher, Agilent, Bruker, Mettler-Toledo, Waters). Chinese biotech activity is plainly booming: a "gold rush," US venture investors opening offices, and 48% of licensed or acquired new drug candidates. But no Western operator told us whether they are winning that spending. Add President Xi's upcoming Washington visit, raised in the tariff discussion, as a wildcard.

Biotech funding sentiment. Mixed. The XBI is up about 30%, and there are oversubscribed rounds like Moonwalk's $70M. Against that: an M&A chill after Novartis's Avidity setback, rising rates, midterm drug-pricing risk, and an MFN wave "coming for the mid-caps." On Capital Allocators ("André Perold – Investing Through Exponential Change," September 21), André Perold, CIO of HighVista Strategies ($14 billion), called today "the most exciting investment environment that I've ever seen" and "the most uncertain." He described "venture investments in the many billions. $20 billion value, $5 billion before," and said HighVista does "a lot in biotech" through narrow specialists. Supportive for funding, but not a tools-specific read.

FX. No coverage at all this week.

What changed vs last week

Last week (September 20) was the busiest, most bullish week in over a month, driven by operators at our two largest names. This week the news turned to policy, China and structure, mostly less friendly to the bull case. Specifically:

1. From operators on the record to operator silence. Last week: Thermo Fisher's CEO with hard guidance, and Illumina's CMO on strategy. This week: no executives from any of our 15 companies. Nothing contradicts their guidance, but nothing adds to it either.
2. NIH: last week's softening now has a new risk. Last week Thermo Fisher's CEO said the NIH cuts "ended up being a lot less severe" and expected "modest growth in NIH budgets over time." This week brought reports (relayed by a partisan podcast) of a draft executive order creating a grant-veto committee. The budget picture may not have changed, but control over how the money is awarded has become a new concern. Treat this as unconfirmed but material.
3. Tariffs and reshoring: from slow tailwind to hard deadline. Last week, a supply-chain consultant described gradual "regionalization of supply chains." This week the policy has a date (September 29), a rate (up to 100%), a list of 13 exempt companies, and no guidance. The long-term onshoring case holds. The short-term disruption risk for small biotechs is new.
4. China: a more complicated picture. Last week Thermo Fisher's CEO said China was back to low-single-digit growth and "a good growth market over time." This week confirmed the Chinese biotech boom, but framed it as competition. A Chinese CEO said "U.S. companies cannot compete," and Ginkgo's CEO put China at 48% of licensed or acquired new drug candidates. Whether that helps or hurts Western tool makers is still open.
5. The AI-data thesis: refined, not reversed. Last week, ex-Illumina executive Omid Ostadan said "the rate-limiting step... is the data." This week Guardant's co-founder agreed on data and pushed back on the hype: "90% of that is probably overhyped," because data is "undersampled." Ginkgo's CEO added that the data may increasingly come from robots. That still points to more sequencing and more lab consumption, with automation as a new variable.

What did not change, and it is still the crux: still no bioprocessing book-to-bill from any listed supplier, still no read on the sequencing competition, still nothing on FX, and no company guidance. Next up: the September 29 tariff start and the MAHA summit the same day, then Q3 earnings in late October.