Comparing Four AI-Powered Drug Discovery Stocks

A technology like artificial intelligence (AI) is extraordinarily complex to begin with. When you start to apply AI to complex domains like drug discovery, there’s a snowball’s chance in hell your average retail investor will understand what’s happening under the hood. If, as risk-averse investors, we adhere to Warren Buffett’s adage of only investing in what we understand, this puts us in a difficult position.

One way to break down the complexity of a company is by examining their business model. Without revenues, it’s practically impossible to properly examine a business model, which is why we don’t invest pre-revenue. Today, we’re going to look at four different publicly traded companies that are using AI to improve the drug discovery process.

Four AI Drug Discovery Stocks

Company NameTickerMarket Cap
(USD billions)
Recursion PharmaceuticalsRXRX4.52

Artificial intelligence algorithms are great for solving complex problems using lots of data. If an AI algorithm can master the game of Go, it can probably help us discover better drugs. In past articles we’ve written extensively about AI drug discovery companies using techniques like computational drug discovery techniques. Earlier this year, we wrote about 7 Companies Using AI for Drug Discovery, one of which was a U.K. startup called Exscientia which recently filed for an initial public offering (IPO).

Exscientia Stock

Click for company website

Founded in 2012, Oxford’s own Exscientia raised $374.4 million in disclosed funding from a slew of investors including names like Softbank, BlackRock, Celgene, Bristol Myers Squibb, and Sir William of Gates. All that money was used to build an AI-powered design platform known as Centaur Chemist which combines the power of machine learning with the knowledge of human chemists to discover drugs faster. As seen below, Exscientia reduces the time it takes to go from target to candidate by 70%.

Chart showing how Exscientia's platform speeds up the drug development process
Credit: Exscientia S-1 Filing

Once there’s a drug candidate, it then needs to proceed through the FDA drug approval process like any other. That’s an easy value proposition to understand, but the accompanying business model is anything but.

If a software-as-aservice (SaaS) business model is to be rewarded for consistency and predictability, then a business model with unpredictable revenue streams should be penalized. While Exscientia’s revenues may appear to be starting out stable, there’s loads of volatility bubbling under the surface. There are two revenue streams – service fees and licensing fees – from which there can be four types of payments; upfront payments, research funding, milestone payments, and opt-in payments. Each relationship Exscientia has comes with its own terms. From collaborations to joint ventures, the business model quickly becomes so complex that it’s hard to fathom how anyone can keep track of what’s going on. So far, most of their revenues are coming from their relationship with Celgene.

During the periods ending December 31, 2019 and 2020, 69% and 83% of our revenue, respectively, related to the recognition of the Celgene up-front payments in line with our progress towards delivering up to three clinical candidate compounds.

Credit: Exscientia S-1 Filing

Then there’s the $4.6 million in revenues recorded for the first half of 2021 which consisted of 13% share ownership in a Chinese firm called GT Apeiron Therapeutics. While it’s counted as revenue, it sits on the balance sheet as an asset, albeit one that’s not likely to be very liquid.

While Exscientia has originated “the first three AI-designed precision drug candidates to enter human clinical trials,” their business model is far too complex for our liking. Complex business models = uncertain cash flows = stock price volatility. The same can be said for our next company.

AbCellera Stock

Click for company website

AbCellera went public in December of 2020 reaching nearly $72 per share on the first day of trading. Waiting for the dust to settle would have been a smart move. Today, shares of AbCellera trade at around $20 a share – what the IPO was originally priced at. Investors who want a piece of the company should consider the business model which suggests irregular revenue streams that depend solely on their partners’ ability to commercialize drugs they help develop.

The vast majority of the potential value for each program under contract is represented by potential future milestone payments and royalties rather than research fees.

Credit: AbCellera

Earlier this year, we did a deep-dive on AbCellera titled, AbCellera Stock – Turning Antibodies Into Drugs. In that piece, we talked about how their AI-powered antibody discovery engine has attracted some of the biggest names out there. One of those names, Eli Lilly, has already been filling AbCellera’s coffers with cash because of a COVID drug that came to market a whole lot quicker than usual. Aside from that, here is the total sum of molecules currently in the clinic from AbCellera’s many partnerships.

A list of molecules in the clinic for AbCellera
Credit: AbCellera 10-Q

Note that the molecule being worked on with NovaRock Biotherapeutics results from AbCellera’s acquisition of humanized rodent platform, Trianni, last year. Given the unpredictability of revenue streams stemming from their business model, AbCellera is a stock we believe represents far too much risk for our liking. If only their business model looked a bit more like Schrodinger’s.

Schrodinger Stock

Click for company website

In last year’s piece titled A Computational Chemistry IPO From Schrodinger, we liked what we saw enough to add the company to our own tech stock portfolio. While revenue growth has recently stalled, we’re assured in their Q2-2021 earnings call that’s only a temporary problem. Says the company:

Software revenue growth is expected to be higher in the second half of the year with the majority of second half growth in the fourth quarter of 2021.

Let’s hope that’s the case.

What makes Schrodinger stand out from the other companies we’ve talked about today is their business model which offers upside across three areas:

  • Access to their platform via recurring contracts
  • Equity positions in collaborators and co-founded companies
  • Broad pipeline of collaborative and internal drug discovery programs

The result is a relatively smooth stream of revenues over time, albeit one that’s presently stalled.

Annual and quarterly revenue trends for Schrodinger
Credit: Yahoo Finance

Schrödinger expects 2021 total revenue to range from $124 million to $142 million, with software revenue expected to range from $102 million to $110 million. Along with this revenue predictability comes reduced volatility, provided they actually hit those targets. One important metric to watch on the software side of the business is annual contract value (ACV):

Annual contract value metrics for Schrodinger
Credit: Schrodinger

This brings us to the last name on today’s list – Recursion Pharmaceuticals.

Recursion Pharmaceuticals Stock

Click for company website

Earlier this year, we published a piece titled Recursion Stock – Revolutionizing Drug Discovery With AI which talked about how Recursion’s claim to have “one of the largest, broadest and deepest pipelines of any technology-enabled drug discovery company.” After a public market debut this past spring, shares of Recursion Pharmaceuticals are trading right around the same price – $26 a share. While we find their technology platform to nothing short of incredible, we’re concerned with how far down the road revenues might be.

To date, they’ve received limited grant revenues, milestone payments from Takeda, and a technology access fee from Bayer, a company they jumped into bed with last year. In Recursion’s words:

We do not expect to generate significant revenue unless and until we progress our drug candidates through clinical trials and obtain marketing approval of, and begin to sell, one or more of our drug candidates, or we otherwise receive substantial licensing or other payments.

In other words, the capabilities of the platform will only be evident when the first drug is approved and marketed, at which time the stock price can be expected to react accordingly. This makes us wonder, just when would an appropriate time be to invest in this company if they emerge as the leader in AI drug discovery?

One key milestone would be an internally developed program that’s successfully commercialized. Once the dust settles following that major event, it may make sense to enter a position because the concept has been proven. Right now, Recursion has four candidates that have advanced the furthest – Phase 1 – in their pipeline.

Recursion Pharmaceuticals pipeline of four most advanced drug candidates
Credit: Recursion Pharmaceuticals

In their latest earnings report, Recursion lists Phase 2 progression for all four of the above candidates to be “3 to 4 quarters away.” In other words, proving the concept will take a while. Fortunately, they have about $600 million in cash to hold them over until they successfully commercialize one of the 48 drugs they’re developing.

Recursion has a pipeline of 48 internally developed programs focused on areas of significant unmet need, an impressive partnership with Bayer (ten programs with possible development and commercial milestone payments of more than USD 100 million per program plus royalties on future sales), probably the world’s biggest supercomputer dedicated to drug discovery, and “one of the world’s largest and fastest-growing proprietary biological and chemical datasets.” One can imagine a point in time when Recursion has commercialized a handful of drugs and a blended stream of royalties start to provide some revenue consistency. Of the companies we’ve looked at today, Recursion seems to have the most promising future, at least based on the scale at which they appear to be operating.

Other AI Drug Discovery Companies

We also need to consider companies that haven’t gone public yet. One such company is Atomwise, a startup which is arguably The Leader in Artificial Intelligence For Drug Discovery. There’s Insilico, a company that’s become the face of AI drug discovery, and BigHat, a company that’s using AI to design biologics. Plenty of other firms are dabbling, and it’s not just startups. Novartis announced a partnership with Microsoft several years back as its “strategic AI and data-science partner.” It’s entirely likely that pharmaceutical companies will bring some of those efforts in-house once they realize the benefits of using AI for drug discovery.


Eventually, using AI for drug discovery will be the norm, not the exception. Given the size and scope of the $1.27 trillion pharmaceuticals industry, there’s plenty of room for more than one winner. For all these companies, success will only come down the road when the real upside manifests itself in terms of milestone payments and royalties. Until then, the unpredictability of revenues will likely lead to one volatile ride.

There are many AI healthcare stocks out there, but we're holding just one. Their software is used by all of the top 20 pharmaceutical companies, and they're also developing their own drugs. Become a Nanalyze Premium annual subscriber today and find out all five AI stocks we're holding.

6 thoughts on “Comparing Four AI-Powered Drug Discovery Stocks
  1. Need more contrast on your graphs. White on pale yellow and gray on black make them very difficult to read for me with a bit of glaucoma.

    1. This is a fair point Harold. I have 20/20 vision and I can barely read some of the graphics in this piece.

      We need to rethink including graphics that are hard to read. The reason we include these is typically because that’s all we can find. We’ll send a note around to the writers/researchers advising them to pay closer attention to this going forward.

  2. Cellarity – a first-of-its-kind therapeutics company that is discovering and developing medicines by studying and altering cell behaviors. Cellarity has developed generalizable platform harnessing single-cell technologies and machine learning to unveil the network state of a given cell, defining the cell’s behavior. Cellarity’s platform digitizes and quantifies cellular behaviors, unravels the network dynamics that govern those behaviors, and generates medicines that can direct them. These medicines are rationally designed to alter the cell behaviors that drive disease, obviating the need for lengthy and costly high throughput screening.

    1. Interesting company that’s raised $173 million so far to doing something similar to what Hungarian firm Turbine AI does. Thank you for raising it.

Leave a Reply

Your email address will not be published. Required fields are marked *