Dave Ricks, CEO of Eli Lilly, on GLP-1s and the business of pharma

1. Key Themes

The AI Revolution in Drug Discovery is Real But Constrained by Data

Eli Lilly is building the largest biologically-focused supercomputer in pharma with Nvidia B300 chips, focused on proprietary drug discovery models, particularly in chemistry. However, the impact is limited by incomplete biological knowledge.

"Probably we need to create the equivalent of what got created with human language, which is a more complete repository of biological knowledge to train against before the machines get a lot better. Today, I don't know, I would estimate we might know 10% to 15% of human biology. So the machine is not going to be good at all until we get way above 50%." [00:04:25] - Dave Ricks

This suggests a massive opportunity for systematic biological data generation - something Ricks believes NIH should be funding but isn't.

The Clinical Trial System is Broken and Half of Drug Development Time is Just Waiting

The most striking operational insight is that half of the 7-year clinical development timeline at Lilly is spent simply waiting for patients to enroll, not conducting actual science.

"Drug development time in the industry is about 10 years in the clinical, little less right now. We're running a little less than 7 at Lilly. But actually, half of that 7 is, we have a protocol open. That means it's an experiment we want to run. We have sites trained. They're waiting for patients to walk in their door and to propose, would you like to be in the study? But we don't have enough people in the study." [00:16:36] - Dave Ricks

The cost structure is equally broken: the median clinical trial enrollee costs $40,000 - two-thirds of the median US wage. This is driven by three factors:

1. 20-30% is "leveling up" care for patients
2. Premium payments to institutions (not physicians)
3. Harvard-style overhead markups capturing costs unrelated to the trial

The US only has 4% of cancer patients in clinical trials versus 25% in Spain and Australia, largely due to systemic friction.

GLP-1s Represent a Fundamental Shift in Medicine's Business Model

The GLP-1 story reveals how a 20-year iterative process led to a category-defining breakthrough, with implications far beyond diabetes and obesity.

"This overnight phenomenon of ozempic and everything else, old story. We launched the first GLP one twice a day injection in 2006. People say, Oh, Dave, when did you know, like, this is going to be huge? My God, I don't know, 2016, 2017." [01:03:55] - Dave Ricks

The broader impact extends to inflammation reduction, which occurs "precipitously early" - CRP (heart attack risk marker) drops 60-70% in weeks, independent of weight loss. This opens applications in chronic knee pain, psoriasis, Crohn's disease, and potentially Alzheimer's.

Critically, GLP-1s are enabling a direct-to-consumer model that bypasses traditional healthcare intermediaries, with Lilly Direct becoming "the largest prescription platform online in terms of revenue" and the "#1 prescribed form" of these medications through self-pay.

2. Contrarian Perspectives

The Patent System Should Be Reformed to Protect First Clinical Data, Not Just First Filing

Ricks argues the 2011 shift from "first to invent" to "first to file" inadvertently created a "shadow generic industry" by forcing rapid patent publication, giving Chinese competitors a roadmap to develop molecules with "one or two atom differences" that work the same but don't violate patents.

"If the monopoly is debased by 30 Chinese biotechs who feed that patent into a computer, the computer then can imagine chemical structures that have one or two atom differences that don't fit within the patent and then make that substance tested works just the same. You've created basically a shadow generic industry and undermine the patent system itself." [01:12:08] - Dave Ricks

His solution: Give 12 years of exclusivity to anyone producing primary Phase 3 data, independent of patents, and extend patent confidentiality from 12 months to 6 years for countries that respect IP.

America's "Free Rider" Problem in Drug Pricing is Bad for Innovation AND Politics

While conventional wisdom says the US subsidizing global drug development is generous, Ricks argues it's actually harming American interests by:

1. Creating political backlash that undermines innovation
2. Skewing R&D toward US-specific problems rather than global health needs
3. Being unsustainable as Americans "wake up" to the disparity

"80 to 100% of the revenue they'll pitch you on is the US. Meaning there is no return outside the US if you started the point of origin of the idea. Now, once we get to the market with a product, it's not sensical to not market it in these countries, at whatever price you can get, because your R&D is paid for on the US launch." [01:45:52] - Dave Ricks

His "One Fair Price" solution: Manufacturers set one US price and must offer it in other developed economies at a GDP-adjusted discount (e.g., 70% in UK if GDP per capita is 30% lower), eliminating rebates and discounts entirely.

The "Obesity Bias" is Preventing Healthcare Economics from Working

Ricks reveals a cultural bias preventing rational healthcare resource allocation around obesity treatments.

"The idea of just treating someone who's overweight or obese, without any other illness, to many people, I think, exposes a bias we have about that particular condition. That if it wasn't something you could see, you might not have. But I think we are conditioned to think of someone who's overweight as someone who is not disciplined. The data does not show that, actually." [00:27:41] - Dave Ricks

He notes that ICER (funded by insurance companies to prove drugs aren't worth it) found GLP-1s are cost-effective within 2 years, saving twice the $100,000/person/year threshold at current pricing - yet coverage remains limited due to this bias.

PBMs Created the Insulin Pricing Crisis, Not Pharma

The insulin pricing controversy was actually caused by the rebate system, not manufacturer greed. Lilly's net price remained around $30-40 while list prices hit $275 because they were competing on the "spread" - the difference between list and net that PBMs captured.

"I have like at least one or two AIs running every minute of every meeting I'm in. And I'm just asking it, science questions. So which one do you use for science? I tend to use either Claude or the XAI one. I find it more terse and more, the references actually check out more often." [00:06:03] - Dave Ricks

When Lilly launched their own biosimilar at one-third the price, "all these insurance companies and middle people called me and said, why did you do this? As well, because we're trying to lower insulin prices, they said, don't. This is a threat to our model." [00:44:46] - Dave Ricks

Scale Has Diseconomies in Drug Discovery

Contradicting conventional business wisdom about scale advantages, Ricks argues strongly for keeping discovery units small (400 people) while scaling clinical trials, manufacturing, and distribution.

"Active invention has diseconomies in scale. So we started in San Diego, we built a monoclonal antibody biotech hub there. It's produced like a third of the medicines we've made since we started hugely successful. There's 400 employees." [01:02:01] - Dave Ricks

He implements an "allowable variation" model where sites aren't micromanaged on budgets, allowing scientists to pursue unauthorized projects - which have produced "a long list of pretty compelling inventions that came out of Lily that were not sanctioned projects."

3. Companies Identified

Nvidia

Building the largest biologically-focused supercomputer in pharma with Lilly using B300 chipsets. Part of an exclusive "rule of 80" club (growth + margin) that includes only three large-cap companies, with Nvidia trading "at multiples above" Lilly despite potentially shorter cycle sustainability. [00:02:01, 00:49:57]

Stripe

Powers Lilly Direct, the direct-to-consumer platform that has become "the largest prescription platform online in terms of revenue" and the #1 prescribed form of GLP-1 medications. Represents the broader trend of established businesses "building direct digital relationships with their customers" across industries. [01:06:12, 01:44:22]

Adobe

Ricks serves on the board. He was "fascinated when I first arrived that 90% of the revenue was off a website that they ran" - exemplifying the power of direct-to-consumer models that Lilly is now replicating in pharma. [01:45:06]

Novo Nordisk

Direct GLP-1 competitor currently growing "in the teens" versus Lilly's 40% year-to-date growth. In the US, Lilly captures 70-75% of new GLP-1 patients and has a 60-40 advantage overall. Running an oral semaglutide study in early dementia patients. [00:49:58, 01:11:11]

CVS, Express Scripts, United Healthcare (PBMs)

Identified as value-extracting intermediaries in the insulin pricing crisis. "These PBMs take their returns on the percent off list. So the higher the list, the better they do. And I think that's a terrible incentive." Called Lilly to ask them to stop offering lower-priced insulin because "this is a threat to our model." [00:44:46]

ICER (Institute for Clinical and Economic Review)

"Funded by someone who hates our industry and the insurance companies" to analyze new drugs, "usually seeking to prove that they're not worth it." Nevertheless found GLP-1s cost-effective at current pricing, with tirzepatide twice as effective as their threshold. [00:26:57]

Genentech

Mentioned as the originator of monoclonal antibody technology that led to a "30 year run of spectacular new medicines" - exemplifying how platform innovations can create entire new drug classes. [01:59:47]

Gilead Sciences

Cited as example of company built on "verology chemistry and small molecules" platform innovation - another case of catching the wave early in a new modality. [01:59:59]

4. Operating Insights

The "Never Decide in One Meeting" Rule

Lilly has institutionalized a discipline where major decisions require multiple meetings to allow for reflection and persuasion.

"We actually have a rule to never decide in one meeting. So you're asking about the day. But we come back to it, think about what others said, and kind of push it again." [00:08:18] - Dave Ricks

Aggressive Cycle Time Reduction as Core Competitive Advantage

Lilly runs clinical trials in under 7 years versus the industry average of 10 years, primarily through systems optimization rather than better science.

"If you can make software faster than someone else, you're going to win. And the same in the drug business." [01:51:54] - Dave Ricks

This is achieved through:

• A "more complete repository of biological knowledge" for training
• 24/7 robotic experiments to create training datasets
• More rigorous decision processes with "bumpers on the bowling alley"
• Portfolio management across 5-10 late-phase projects simultaneously

Keep Headcount Flat While Growing Revenue 40%+

Ricks is "very unpopular" for insisting on flat headcount despite hypergrowth, with exceptions only for manufacturing (unit operations) and modest single-digit growth in R&D headcount despite high-teens R&D spend growth.

"The money goes to projects... clinical trials, new equipment, new laboratories, super computers from Nvidia that's expensive." [02:04:41] - Dave Ricks

Create Physical Proximity Ecosystems for Scale-Ups

The Gateway Labs and Catalyst 360 model hosts scale-up companies (not startups) by offering active Lilly scientists and resources rather than retired consultants, creating "loosely coupled" relationships that may lead to acquisitions. This is a capital-efficient way to expand R&D reach. [01:57:02]

Use AI Constantly for Learning, But Choose Carefully

"I have like at least one or two AIs running every minute of every meeting I'm in. And I'm just asking it science questions." For science, Ricks uses Claude or XAI rather than ChatGPT because "the references actually check out more often. Sometimes the AIs produce references and they're actually not the thing that it said." [00:06:03] - Dave Ricks

5. Overlooked Insights

The 3,600 Calorie Problem and GLP-1's 800 Calorie Solution

Almost casually mentioned but profound: average daily caloric consumption in America is 3,600 calories, and GLP-1s reduce consumption by 800 calories - nearly a full meal. This isn't just about weight loss; it's addressing a fundamental mismatch between human biology (evolved for starvation) and modern abundance. Food companies are "so worried" for good reason - this could reshape significant portions of the economy. [00:28:29, 00:28:39]

The "Therapeutic Index" is Expanding Non-Linearly

Ricks briefly mentions that "the techniques we make drugs with, that therapeutic index, the TI we call it, is expanding and expanding non-linearly." This widening gap between helpful and harmful doses, driven by RNA-blocking medicines and genetic approaches, is what enables the shift to preventative, self-administered medicine with minimal healthcare system involvement. This is the technical unlock for a fundamentally different healthcare business model. [00:56:49]

The LP(a) cholesterol example - a once or twice yearly treatment with "totally benign" side effects - exemplifies this. "That's a really wide therapeutic index. So now when you have that, you can think, well, my trials, I can run faster because I don't have worries about treatment. They can be cheaper. I can charge less and get it to more people at scale. And I actually don't really need a healthcare system." [00:57:30]

This is the technical foundation enabling the entire direct-to-consumer transformation in preventative medicine - arguably more important than any single drug.

---

Investment Implications: Lilly is trading at 50x forward P/E versus ~12x for the sector, capturing 70-75% of new GLP-1 patients with 40% YoY revenue growth. The company is one of only three large-cap companies (with Nvidia and one other) in the "rule of 80 club" (growth + margin >80). The shift to direct-to-consumer channels and expanding therapeutic index suggests a potential escape from traditional pharma patent cliff dynamics. Clinical trial acceleration and decentralization represent significant operating leverage opportunities across the sector. The Chinese biotech threat is real - they now represent ~30% of the global pipeline versus single digits a decade ago - but could be addressed through data exclusivity reforms rather than patent term extensions.