Academics will spend the next decade or two unpacking the Rona years, trying to understand the complex socio-eco-cultural ramifications of the pandemic. Or something like that. For retail investors, the lessons are much simpler: zeroes became heroes overnight. In other words, relatively obscure companies suddenly gained billions in market cap as they filled a newly created niche, like video conferencing or covid vaccines. However, we’ve always viewed stocks like Zoom (ZM) and Moderna (MRNA) as pandemic one-hit wonders like In-A-Gadda-Da-Vida. Go ahead; try to name another Iron Butterfly song.
Abcellera (ABCL), an AI drug discovery company out of Vancouver, BC, joined the biotech big leagues in 2020 when it developed an antibody treatment for covid in collaboration with Eli Lilly and Company. The little-known Canadian firm topped $485 million in revenue last year, with more than 90% of that from royalties associated with its covid drugs. The Rona gravy train has run out of track, as we noted in our recent video on Abcellera stock in which we tried to figure out where the company is headed next. Theoretically, if its AI-powered platform for discovering antibody therapies is the real deal, then more success should be coming its way.
Meanwhile, another AI drug discovery company that we also like, Recursion Pharmaceuticals (RXRX), has yet to score a big hit with its Recursion OperatingSystem (OS). The platform reportedly maps “trillions of biological and chemical relationships’” using the company’s ginormous datasets to develop novel therapies for hard-to-treat diseases. We first took a hard look at Recursion shortly before it IPO’d in April 2021 and concluded that the breadth of its pipeline, partnerships, and proprietary data were signs that it could be a legitimate contender in an increasingly competitive drug-discovery sector. Now, more than two years later, is that still the case?
About Recursion Stock
Like just about every other biotech IPO in 2021, Recursion stock got off to a hot start, practically doubling from its initial $18 per share price on the first day of trading. But the party pretty much ended after the Fourth of July 2021, and Recursion stock has been on a pretty steady slide ever since. Today, at a market cap of about $1.6 billion, the company is down more than 70% in value. However, that may not last long. As Wall Street blindly grasps at anything AI, it’s possible that AI drug discovery firms may soon be hyped along with all other AI-related names. Therefore, it’s critically important to know which names are the most promising so we can avoid the rest. Does Recursion still hold the same appeal since our last piece on Four AI Drug Discovery Stocks On Sale – 50% Off ?
How Does Recursion Develop Drug Candidates?
Before we push through that analysis, it would be useful to briefly (and very simply) recap what Recursion does. Recall that the Salt Lake City-based biotech (which calls itself a techbio company) has developed a unique approach to drug discovery. It basically uses gene-editing technologies like CRISPR to make thousands of cellular models of different genetic diseases in parallel. The platform then uses high-tech imaging to screen each diseased cell, followed by some slick computer vision to extract relevant details. Machine-learning algorithms crunch away on all the data, looking for patterns between healthy and diseased cells that would suggest a potential therapy.
The concept is to “industrialize” drug discovery, which presumably makes it faster and cheaper to discover treatments, which traditionally can take upwards of $2 billion and a decade to bring to market. To date, the Recursion OS has conducted more than 175 million experiments, generated more than 21 petabytes of data, and mapped more than three trillion biological and chemical relationships.
How Does Recursion Make Money?
So far, all those big numbers have not added up to much in the Recursion bank account. In theory, the company’s business model consists of three pillars:
- An internal pipeline of drug candidates
- Partnerships with big pharma companies like Bayer and Roche
- Leveraging its big datasets for big bucks by licensing subsets of its data universe. For instance, the company’s recent collaboration agreement with Roche/Genentech includes something like $500 million in milestones based on creating and optioning collaboration data generated by Recursion.
In practice, most of the money that Recursion has made to date has come from its partnerships with Bayer and Roche/Genentech. In 2021, the majority of about $10 million in revenue came from the Bayer agreement, which involves R&D programs focused on fibrosis. The deal included upfront payments of $30 million and a $50 million direct investment into Recursion. In 2022, most of the $40 million in revenue flowed out of the $150 million Roche/Genentech deal, which has programs in neuroscience and one in oncology. While both partnerships have lucrative milestones and royalties attached to them, there are no guarantees.
Meanwhile, it actually cost Recursion about $8 million more to make $40 million in 2022 – meaning that the cost of revenue exceeded the actual revenue. And that’s before piling on about $155 million in additional R&D expenses and nearly $82 million for just running the company (for a total loss of about $240 million in 2022).
Regular readers will point to our traditional avoidance of negative gross margins, but these business models are exceptions to the rule. To assess firms like Recursion, we need to deviate from our traditional investing methodology because drug discovery outfits operate a lot differently. They spend massive amounts of money upfront for payoffs down the road. They’re of the same ilk as gene editing companies that don’t have revenues yet. While we never invest pre-revenue, gene editing is one of a few domains where we might make an exception (and AI drug discovery if more traction is made). Again, these domains involve discovering and developing drugs which requires loads of capital upfront for future large payoffs. These business models are extremely risky which is why we typically avoid traditional drug discovery companies.
What we’re most interested in is if Recursion can survive long enough to start realizing enough success stories to pay the bills. With $473 million in cash at the end of Q1-2023 (and only about $1 million in debt), it reported about $65 million in losses for the most recent quarter which means they’ll be out of runway in less than two years. This makes us wonder what progress the company is making.
What Progress is Recursion Making?
Recursion is making progress with its internal drug pipeline, initiating five clinical trials in 2022, including three Phase 2 programs. However, most of these more advanced-stage therapies target rare conditions and are still years away from commercialization. For example, take Neurofibromatosis Type 2 (NF2), a genetic disorder that causes tumors to grow on the nerves of the brain and the spine, but most commonly affects auditory nerves. About 1 in 25,000 to 1 in 40,000 people are born with it, representing a global patient population of between 200,000 and 375,000. The current Phase 2/3 clinical trial involves 90 people, with an interim safety analysis due in 2024. None of these therapies are likely to add to the bottom line before the war chest well runs dry.
Recursion did add some new AI capabilities to its OS platform through two recent acquisitions, worth a total of $87.5 million, payable in company stock. Toronto-based Cyclica is a specialist in digital chemistry, with a couple of AI tools that will be integrated into the Recursion toolbox. For example, one is an AI-enabled deep learning engine that predicts the polypharmacology of small molecule drugs, so that one molecule can bind to multiple targets. The other acquisition, Valence, is also a Canadian biotech based in Montréal at Mila, reputedly the world’s largest deep learning research institute. Valence has developed an AI application for drug discovery that can design small-molecule drugs based on sparse or otherwise crappy datasets that would defeat most deep-learning methods.
Can Recursion Remain Relevant in the AI Drug Discovery Race?
While those acquisitions eliminate a couple of small-time players, the AI drug discovery space is still filled with competition, from both publicly traded companies and startups. Failure for some of these firms is inevitable. London-based BenevolentAI (BAI.AS), a drug discovery company that was one of the biggest AI startups back in the day, is pivoting after a big flop midway through a clinical trial. Today, the company sports a market cap of just $240 million after losing more than 90% of its value just 14 months after merging with a special purpose acquisition company (SPAC).
Recursion is not a SPAC but the company could still suffer the same fate. Its novel approach to drug discovery remains unproven. The Bayer and Roche/Genentech deals are relatively modest in terms of upfront payments. The current total addressable market (TAM) for its most advanced-stage therapeutics also appears to be pretty modest. The company will need a big win sooner than later to remain relevant in the AI drug discovery race.
But the appeal of AI drug discovery isn’t a single drug, nor a dozen drugs, it’s a platform that churns out drugs at a consistent pace at a speed and success rate that completely disrupts the current methods of drug development. That means we’ll be looking for some milestones:
- First drug approved
- Second drug to be approved
- A somewhat reliable cadence of drugs being approved
- First drug to achieve blockbuster status
The first milestone is one to be watched closely. That’s because such an event will make all the news headlines and this rising tide will affect all boats. In a coming piece, we’ll look at which companies are furthest along when it comes to pipeline progress.
Conclusion
Drug discovery remains a risky and expensive proposition. One analysis a few years ago said that over a five-year period, big biopharma companies managed to convert just 24% pipeline candidates to approved drugs, spending an average of just under $1 billion on R&D for each in development and $5.3 billion for each approval. The value proposition is obvious but whether AI can significantly move the needle in this process remains a big black box. For retail investors, AI drug discovery stocks could turn into a black hole. That’s why the first AI drug success story is a critical milestone towards proving the concept.
Tech investing is extremely risky. Minimize your risk with our stock research, investment tools, and portfolios, and find out which tech stocks you should avoid. Become a Nanalyze Premium member and find out today!
