Bio Strategy Overview

Happy Monday

The past week or two was more or less entirely consumed with preparing for our Annual Limited Partners Meeting, where I presented the bio strategy and portfolio. What I told them, more or less, is below, for your entertainment. But before you read it, some context:

Our LP meeting is an unusual one, as AGMs go. It’s quite small, about 30-35 attendees total, half from the institutional funds that make up the LP group and half from Sutter Hill. Given how concentrated our portfolio is, we have time to walk through it almost in its entirety, interspersed with market and strategy overviews for each vertical (software, bio, and hardware). The small group setting allows for remarkable candor, which the LPs appreciate; they enjoy it so much that they routinely send the heads of the endowments to represent them, along with up to one additional representative.

That means that the audience is made up of maximum generalists, to whom we must explain our approach to investing in highly technical domains. So, if the following seems overly distilled or high level, just know that was intentional. Also, in case you are wondering: I wrote it, the emdashes are representative of correctly deployed spoken pauses.

Bio Strategy Overview at SHV’s 2026 Annual LP Meeting

Our Approach

We invest in early stage technology platforms that aim to remove bottlenecks in the biopharma market, because we believe that is the path to generating outsize returns. To understand why, let’s talk briefly about this market and how it works.

Structurally Unlimited Demand

For one, it’s huge: about $1.8T today and projected to grow to over $3T in the next decade.

The vast majority of this revenue belongs to Big Pharma, who commercialize most new drugs, and the vast majority of their revenue comes from products they acquired externally.

They spend between $200-350B a year through M&A and licensing on product opportunities acquired for their pipelines.

This is structurally growing demand: every successful drug launch creates a future revenue gap that needs to be filled by the time they lose exclusivity and generic competitors enter. The bigger the product, the bigger the gap they need to fill and the more they need to spend to try to fill it.

You can think of this collection of 25 or so economic buyers as analogous to ‘hyperscalers’, with large, diverse product portfolios of medicines that they are always looking to replace and expand in order to grow, and product opportunities across this landscape are, in the hyperscaler analogy, like ‘sockets’ they need to fill.

The startup opportunity in this space is to create and develop new assets — potential drugs — for Big Pharma to acquire.

But importantly not all startup opportunities are created equally, particularly when it comes to risk and return profile.

Tale of two startup opportunities

An asset that a startup develops has a quantifiable net-present value at every stage, adjusted for the risk remaining in advancing it to the next stage.

The upside is capped at the value to the acquirer, which is roughly the market size of the ‘socket’ you’re aiming for.

There are two strategies to capture this value as a startup.

There is the traditional biotech playbook, where you develop an asset along this value chain: start with a biological insight, build and develop a drug around it with existing tools, derisking it systematically with increasing capital intensity, all while aimed at a particular “socket”.

Historically, the tradeoffs between risk, capital efficiency, and binary outcomes of this model offer asymmetric but fundamentally capped upside for investors: IRRs of 20-25% are possible, but it’s very hard to do better than that with this strategy.

The other strategy is to build new tools for making medicines: technology platforms that unlock previously inaccessible asset opportunities, which give you an unfair advantage in making better products that can aim for as many ‘sockets’ as possible.

This gives you the most strategic optionality and a path to uncapped upside: you can decide how far you want to take each asset and how much value to try to capture based on your cost and availability of capital, and there is no limit to the number of asset opportunities you can pursue. This is how you generate IRRs well north of 40%, and that’s why we focus here.

There are tradeoffs: this is a much higher risk strategy, the time to value is longer, and there is no playbook to follow.

But this is the perfect strategy for us: it is technical risk, not demand risk, and we can be patient knowing that if we build it, the demand will be there.

And we excel at identifying the best technologists and entrepreneurs who can build things no one else can build and have the creativity to write their own playbooks.

More importantly, those people prefer to work with us over anyone else; you’re going to meet one soon and can ask him yourself¹.

The forces disrupting the market are undermining the traditional biotech playbook and advantaging ours. Let’s talk about China.

China is a technology commoditization machine

China is a technology commoditization machine. That is just as true in drug discovery: if you are running a playbook with existing technology, China can run it better, faster, and cheaper to make an equivalent asset at a lower price.

The numbers bear this out: About a third of pharma licensing deals now involve Chinese assets, with projections suggesting that could rise to 40–50% and surpass US biotech. Big Pharma doesn’t care where assets come from; they want the best ones at the lowest price.

But China is also risk averse. They focus on mature therapeutic technologies and crowd into proven disease targets and drug types — because that’s where the playbooks are.

This moves the bottleneck in the market in our favor, placing a premium on new technologies that unlock better products, and where there’s no playbook for China to out-execute us on.

There is a lot more nuance to the China story, but let’s move on to the thing everyone is talking about, including in bio: AI.

AI makes data useful as a tool

If physics and engineering have math, biology has data. There are no equations of biology we can write down and solve — if we want to learn something about a living system, we need to measure it.

Modern AI has done many things, but in biopharma its transformative impact has been to make that data useful for making predictions about biology. That makes it a tool to design better drugs than competitors, but only as long as you have a proprietary data advantage.

In bio, small models can be very useful; AlphaFold is a few orders of magnitude smaller than a frontier LLM. It’s achievable for a startup with proprietary datasets and modest compute to build custom AI tools as part of their platform. The bottleneck is talent and knowhow.

To address this, as part of the SHV platform on the Bio side, we have an in-house Bio AI R&D group led by our partner Alex Peysakhovich — building internal data assets, models, and ML infrastructure, and working across the portfolio to help them train and deploy custom AI tools on their proprietary data. Some of the AI work you’ll see from Ridge shortly is a direct outgrowth of this effort.

Our Opportunity

To sum up: as the conventional biotech playbook gets commoditized and AI reshapes what’s possible, the competitive advantage belongs to technology platforms that can make medicines no one else can. That’s what we invest in.

It was almost 50 years ago to the day that Sutter Hill passed on Genentech — a generational business that kickstarted an industry. The next generational biopharma companies are being built today. I know for a fact that we’re not going to miss out this time, because we’re going to talk about some of them next.

Binding Sites

1. PINCHs - A New Pharmacological Modality to Sequester Homomeric Proteins Another day, another induced proximity mechanism of action. The core idea is to trap and neutralize problem proteins by forcing them to clump together into insoluble polymers and removing them from the cell’s soluble phase. Creative!

Footnotes

1. Weston ably presented on Ridge after I presented my overview and reviewed a selection of the portfolio. ↩