A Hidden Gem Drug Discovery Platform with Endless Applications

Sometimes it might seem like big technological change is really only happening when we see it in the everyday consumer goods and services we buy. Amazon and Apple are very obvious manifestations of technological progress. But the truth is, there are at the moment some incredible technological advances happening behind the scenes in a number of industries and some of the biggest aren’t really apparent in our everyday lives.

In the same way Amazon modernised the brick-and-mortar shopping experience, there are some very exciting new companies modernising the entire drug discovery industry (which is called the biotech sector). You will have seen in my note on Frontier IP how biotech company Exsciencia is the first company in the world to launch a new drug without any human input, purely using artificial intelligence. This radical change in approach has been enabled by the convergence of A.I. and biotech and I believe has the power to completely revolutionise the way drug discovery works for the better. The reality is that A.I. can screen a vast database of molecules and even existing drugs far more quickly and cheaply to determine which ones have a reasonable probability of being effective against specific disease targets. That takes much of the guesswork out and saves millions of dollars and valuable years in the process.

Invest in Thesis Portfolio

Join NOW!

The crazy thing is that currently only 1 in 5,000 drugs that are discovered in the lab actually make it to market! This is according to Praxis, a clinical trial patient recruitment company. The process involves sorting through thousands of possible molecular combinations and trying to ‘guess’ the right sequence that could become an FDA-approved drug. It’s like trying to randomly guess a password or win at a slot machine game! Unbelievably, modern medicine as we know it - and the many miraculous cures we’ve found so far- has all been built on that “slot machine” probability of just 0.02%. So if now we are at a point where using A.I. technology, we can bump that success rate up to, say, just 1%, this would mean 50 times more drugs on the market (not to mention the enormous health benefit of 50 times as many answers to pain, suffering, and disease).

And what’s more, this is no longer a pipedream: more and more evidence is emerging that the use of A.I. in biotech is making the discovery and testing of new drugs far more effective and efficient as we can see for example with Exsciencia. Today, I’d like to tell you about a US company I hold in both my personal portfolio and the Thesis portfolio, which is in my opinion ideally positioned to benefit from this amazing technological convergence.

The Power of A.I. Biotech Convergence Explained

Traditionally, the drug development process involved scientists screening existing libraries of molecules to identify potential prospects which might be useful in interacting with disease. An experimental screening is typically done to find molecules with detectable activity, or “hit molecules”. After that, many rounds of chemical synthesis are done to attempt to optimise these hit molecules to a development candidate that can be advanced into clinical development.

With the advent of computers, the first improvement was to use machine learning to process existing libraries of molecules to identify potential prospects. Schrödinger’s glide product, which was released in the early 2000s, was one such computer software designed to augment and accelerate screening and hit discovery. At the heart of Schrödinger’s platform is the application of the ‘Schrödinger equation’ (proposed by Erwin Schrödinger in 1926) which aims to explain and predict the properties of atoms and molecules. Drug molecules are often large organic compounds and calculating their interaction with target proteins is still very difficult. To obtain results with any meaningful accuracy, a great deal of computational power was required so the advances we have seen in computing power have brought the potential of this platform to life.


Source: MSN Money (Price data as of 28.10.2021) Past performance is not indicative of future performance. Returns may increase or decrease as the result of currency fluctuations. Figures refer to price data between February 2020 and October 2021)

Business Description

“Schrödinger’s industry-leading computational platform facilitates the research efforts of biopharmaceutical and industrial companies, academic institutions, and government laboratories worldwide. Schrödinger also has wholly-owned and collaborative drug discovery programs in a broad range of therapeutic areas.”“Schrödinger is transforming the way therapeutics and materials are discovered. Schrödinger has pioneered a physics-based software platform that enables discovery of high-quality, novel molecules for drug development and materials applications more rapidly and at lower cost compared to traditional methods.” Company Presentation

US company Schrödinger’s platform truly represents a “new way” to discover drugs, offering pharma companies the service of finding drug molecules using artificial intelligence (i.e. without human input). According to the company, their technology can “evaluate billions of molecules per week”. Of course, this ‘new way’ isn’t just aiming at improving the odds from the current 0.02% to just 1% as in my example … it’s aiming for much, much higher. The ability to now evaluate billions of molecules a week is a seismic step change in the speed at which human beings can undertake this search for new drug targets. As validation of their approach, what better than the fact that each of the top 20 pharmaceutical companies in the world have now entered into contracts to use Schrödinger technology and platform.

But the potential doesn’t stop at drug discovery. Schrödinger has a goal to transform the way in which not only therapeutics but also materials are developed. The company aims to do this through its platform which allows for discovery of high-quality molecules at lower costs and improved chances of success. While there is clear potential for the platform within drug development, there are other huge potential applications including in aerospace, defence and semiconductor industries amongst others.

The company IPO’d last year at $17 per share, reaching a high of $113 in February 2021. In November 2020, the company announced a huge deal with Bristol-Myers Squibb (potentially worth over $2.7 billion). However, in March of this year the company reported what the market took to be a lower than expected 28% increase in fourth quarter sales to $33.0 million. This disappointing short-term result caused the shares to sell off sharply. However, in my mind this pullback represents an attractive buying opportunity for reasons that I will explain below.

The research branch of the infamous hedge fund, Citron Research, called Schrödinger the “most disruptive software platform to ever hit the pharmaceutical industry”. They also called the company “even more compelling than an early-stage Tesla”. A business scholar from the University of Michigan published an analysis in which he wrote, “the market is missing the disruptive potential of Schrödinger.”

Schrödinger’s Market leading Platform

“In a peer-reviewed study, our platform was tested against traditional methods for selecting tight-binding molecules and resulted in an eight-fold increase in the number of molecules with the desired affinity.” Company’s IPO Prospectus

Schrodinger Development
Source: Company Presentation

Revenue Model

The company generates revenue in a multiple of ways. First the company licenses its software to drug developers using a land and expand model. At the time of writing, this provides most of the revenue generated and is showing c.30% revenue growth. This is a stable and growing business that provides the company with a long-term source of recurring revenue. Also these deals generate precious data and I believe the company is entitled to use the data discovered by its customers to help it complete its database of known molecules.

Schrodinger Revenue Model
Past performance is not indicative of future performance. Source: Company Presentation

Schrödinger sells on subscription, and they had total software revenue of $92.5 million last year, with annual contract value growing at a 16% CAGR over the past seven years — which is not nosebleed growth for basically a SaaS company, but it has accelerated in recent years (up to 22% last year). They have 16 customers who each generate more than $1 million in annual contract value, up from 10 in 2019, and they had 99% customer retention. I don’t think we should underestimate the fact that essentially all the big pharma companies are using the platform, and that on average they’re using it more each year.

Second, the company earns revenue and milestone payments by drug development with its partners. This is a highly volatile portion of the business and is dependent on timing of milestones or upfront payments.The company has also been incorporating equity in its partners as a form of payment, which I believe is an extremely savvy move and shows the company’s confidence in its technology.

Schrodinger´s Research
Source: Company Presentation

Lastly, the company also owns and develops an internal drug discovery platform. The programme is currently in the discovery phase; however, in November of 2020 Schrödinger received a huge vote of confidence from Bristol-Myers Squibb which agreed to a $55 million upfront payment and over $2.7 billion in milestone payments in addition to a royalty agreement.

It is clear from the amount negotiated in the agreement with Bristol-Myers Squibb and with the aggressive nature of the move into drug development that the company has a very high degree of confidence in its approach.

Investment Case

The great thing from an investment point of view about Schrödinger is that although this is a nascent emerging sector, the company is effectively selling the ‘picks and shovels’ to every major pharma company to get involved in this AI/Biotech convergence! As I have noted before, this ‘picks and shovels’ business strategy is in my view an ideal way of gaining exposure to new emerging growth sectors for investors: it offers the upside from the high growth on offer but at far lower risk. What makes Schrödinger an appealing investment to me is the duality of the likely steady growth and recurring cash from subscription revenue alongside the many exciting partnerships and development programmes that could bear profitable fruit in the future.

We can see from the recent results that Schrödinger’s software sales create pretty steady cash flow, and this would be attractive enough on its own. However, the company also has an incredible number of collaborations with pharma companies that can lead to royalties on any future approved drugs, as well as a pipeline of its own (very early stage) molecules that they’ve discovered and are preparing for possible clinical trials.

The subscription business is of material importance in supporting the development of its drug pipeline and R&D efforts. Thankfully, this pillar of revenue is very stable, growing healthily and is quite predictable, thus providing a solid foundation enabling the company to shoot for the stars.

$2.7bn Collaboration with Bristol Meyers

The $2.7bn collaboration between Schrödinger and Bristol Myers announced back in November 2020 is a big deal for a company of Schrodinger’s size, with the potential for billions of dollars in milestone payments and some good royalties if drugs are successfully developed and commercialised. Investors reacted positively to the news and drove the shares up dramatically from where they subsequently sold off when it became a little clearer that this collaboration was not going to instantly double revenues. Nothing has changed in the upside of this deal and if anything, it further validates the company’s technology to be the chosen partner.

SDGR did not change their forward guidance for 2021 (which perhaps is another reason for recent weakness in the shares), so they’re anticipating $133 million in revenue this year — and have made it clear that milestone payments will not be as consistent or imminent as was originally anticipated by some analysts back in November. That was really the shock to the system that caused the share price to falter this year — analysts overstated the immediate impact of the BMY deal and sent the 2021 revenue estimates to $175 million or so, with expected profitability right away (47 cents/share in earnings forecast), and SDGR threw cold water on that in March. Now profitability is pushed out to 2023, most likely, though the business is very hard to predict, and analysts have probably over-corrected on the pessimistic side.

Valuation Upside

I expect the company to consistently surprise with upselling their customers to higher usage of their software platform (from what is a relatively low base), and I think this steady growth in usage is very likely to happen. There are no obvious other near-term share price catalysts in terms of newsflow (their own three drugs to be launched are years off) but with partnerships with basically every major global pharma in place, the scope for new positive announcements is high. The other positive is they have plenty of cash on hand ($600m) to ride out any difficult moments in the fundraising markets.

The reason there is more potential growth on the software side (which Schrödinger makes clear in their presentations and investor calls) is that their large users are still not really “power users” and can dramatically increase usage of the system from this low as they begin to see the benefits it provides. Aside from subscriptions, there is big potential from the other ways in which Schrödinger can get a piece of the profits from drugs that are discovered on its platform. On top of this, there is actually no reason that the company cannot use AI for its own push into drug discovery. The upside from these two other revenue streams could clearly be far bigger than software sales in the medium term.

The share price is still 50% off its highs in March though still significantly above its IPO price. The reason the shares got clobbered, other than the general panic about richly valued stocks, is that they issued guidance for 2021 revenues that was wildly lower than analysts had been forecasting. I expect that’s partly because of a misunderstanding about how much revenue would be recognised from their big Bristol-Myers Squibb deal announced last year and partly management conservatism in the face of what is likely to be a lumpy revenue profile. In my opinion, the long-term growth picture remains unchanged, and this pullback therefore represents an attractive entry point.

In terms of valuation metrics, these types of businesses are often valued on a multiple of sales and Schrodinger is currently trading on 30x sales, so may look fully valued. But the problem is the sales multiple underestimates some of the assets within the business (as demonstrated by the Bristol Myers deal). Besides the software revenues and revenues from collaborative programmes, there is an internal pipeline developed by the company as well. Furthermore, the company has equity stakes in numerous companies and collaborations as well, which provide further potential upside as these positions are held at cost on the balance sheet.

The Blue-Sky Scenario: Unlimited Applications!

It’s difficult not to get excited about the application of Schrödinger’s approach in other fields. Schrödinger is, at its core, a software and computing company. The company essentially takes billions of known molecules and uses a physics based computational system along with AI to predict which molecules in combination should achieve a specific desired result. The obvious implication is that this can be applied to numerous different materials outside of drug development. Although the company has chosen to focus on drug development for its opening act as a public company, significant upside may be provided by newer applications. So they have the just-barely-tapped market outside of pharmaceuticals to sell their software platform into - this is obviously all manner of materials companies who are looking for the best molecular interactions for a myriad of things like, for example, new battery materials.

The company can further use its physics-based approach to determine potential side effects of specific medicines and potential interactions with other drugs, theoretically limiting adverse events during the clinical phase. Not only does the company’s approach simply make sense, but it currently is also able to save roughly 50% of a drug company’s most precious resource, time!

And this technology is truly just in its infancy. Schrödinger’s technology improves incrementally with each compound it simulates (even without greater computational advancements) due to the fact that with each molecule modelled, the better the end data becomes.

The pot of gold at the end of this rainbow could simply therefore be the data. Once Schrödinger is able to predict and model each of the countless billions of molecules available with precise accuracy, it will be able to basically design anything in a much shorter period of time.

The company platform could very likely in the future be used in diverse fields such as new battery design, solar panel design and high-grade composites for the aerospace industry among countless other applications like shampoos, plastics, and fuels. Unlike other companies defined to one area, the total addressable market for this company could therefore be basically endless. Schrödinger appears to be a fantastic company with unbelievable potential. If the company can continue to build its database of defined and modelled molecules, it can literally be a part of every industry that uses molecules. Valuing the potential of other applications is nearly impossible. The market for molecular modelling and simulation I feel may be one of the most important markets for all of humanity in the future and he who controls the most data will be in an enviable position!

Risks

Schrödinger is loss making and operates in a brand new industry so there are a number of big risks to be aware of:

  • We cannot discount the risk of potential competition from a similar platform - whether it comes from academia, industry, or pharmaceutical companies themselves - which will limit how much revenue the company’s software business can derive from its customers.

  • The upside to pricing may be limited and self-defeating: if Schrödinger tries to charge more and generate more revenues from its clients, they will likely look for other options, including developing their own inhouse software, funding an open source solution or start to give business to Schrödinger’s competitors.

  • The company’s customers are big pharmaceuticals, and they aren’t very easy to deal with. So, it may be over optimistic to assume that the company can ever command meaningful pricing power in its software business.

  • Right now, this field is still developing and while Schrödinger has emerged as a leader, I don’t think it will always be the only player in the industry. The real test will be to see how much better the company’s results are versus those calculated from computational methods generally available in the public domain. That will be a real measure of the company’s competitive advantage.

  • The company has no real competitive advantage as of yet in discovering its own drugs compared to other biotech players. Any biotech company can buy and use the company’s software. So, the predictive power or lower cost advantage vs. traditional methods, is available for every competitor as well. Having said that, as the data bank grows, so does Schrödinger’s competitive advantage.

  • Milestone and development revenues are notoriously difficult to predict and frankly, the revenue number at this stage in the drug development and molecular modelling market is not material to the company’s short-term prospects. However, the collection of data and the progression of its drug pipeline remains important.

  • The molecular modelling market does have some competition for Schrödinger, with the likes of Simulation Plus, Dassault Systèmes, Thermo Fisher Scientific Company LLC., Genedata AG, Chemical Computing Group ULC, among others looking to gain traction. I am confident however in naming Schrödinger as a leader in the field due to the long and impressive history of the company leading to an edge in its database of molecules, the marquee list of partners, which include the aforementioned Bristol-Myers Squibb, along with Takeda Pharmaceuticals and Sanofi and high-profile investors such as David Shaw, Bill Gates, and many others.

  • The primary risk that analysts and investors seem to be worried about is whether or not they’ve “used up” their target market, since they already work with most of the large pharmaceutical companies, but that doesn’t particularly worry me — the platform can be used lightly or heavily, with variable costs, and their customers mostly use it lightly so far… so there’s potential growth as they become familiar with it and begin to use it more, assuming that it continues to work well for them (retention is very high).

  • Schrödinger is mainly focused on the earliest stages of drug discovery called “IND-enabling” - that means you’re doing the lab research that would provide enough data for you to apply for IND status (that stands for Investigational New Drug) with the FDA, and if they approve that, then you can put that drug into a human being for the first time as you launch Phase 1 clinical trials. Drug discovery is speeded up by the ability to use AI and their massive processing power to select molecules for their precise interactions with other molecules, but that doesn’t mean the rest of the process of actually getting regulatory approval and testing the drugs in humans is going to be any faster than usual. Against this I would say that presumably with better molecule selection, there should be more successes in the clinic.

  • Failure in early-stage drug trials is common. Maybe it will be less common with a molecule that is vetted for its physical properties and interactions as well as those discovered using Schrödinger’s technology, perhaps there’s lower probability of surprising side effects or higher probability of efficacy, but the company’s three lead products could still fail.

Conclusion

Anyone who knows me knows that I am a big believer in Impact Investing (using our capital and the power of capitalism to make a real difference to the world). Schrödinger is doing its bit to make the world a better place by helping improve the drug discovery process to help eradicate and treat diseases faster for the benefit of us all. I believe the company is hitting its stride with high profile partnerships and a balance sheet stocked full of nearly $600 million in cash. The company’s software revenue also seems to be gathering steam with growth rates expanding in 2021 and beyond.

Schrödinger is an exciting, next generation drug development and software player. However, at its core, it is a molecular modelling company, leading to vast opportunities outside of biotech. The very nature of the business, the potential hidden gems and the disruptive potential and nearly unlimited end markets makes for a compelling investment. The model they use in drug development allows them to capture upfront, recurring revenue and equity investments in their partners. The pullback in the stock price is an attractive buying opportunity.

Invest in Thesis Portfolio

Join NOW!

This material is not investment research in accordance with the legal requirements designed to promote investment research independence and is also not subject to any prohibition on dealing ahead of the dissemination of investment research; and as such is considered to be a marketing communication.

This is not financial advice, please seek professional financial advice from a certified & regulated independent financial advisor.

All investments have the potential for profit and loss and your capital is at risk.

Returns may increase or decrease as a result of currency fluctuations.

Past performance is not indicative of future results.

3 Likes

Wow - How do you find these (I was going to say little then) gems. Please don’t tell me!
Thanks again @pb1 for such a detailed weekly focus. I really appreciate the transparency with Pynk - It definitely breeds confidence :slight_smile:

Best wishes,
Paul

2 Likes

I love learning about new things! I’ve written about a few more really unusual situations that I’m really excited and that you might find interesting …the posts are making their way through the compliance etc process so stay tuned!

Your capital is at risk.

2 Likes
Disclaimer: By accessing the community forum you accept the Terms of Use in its entirety. Information, opinions and content posted or uploaded to the community does not constitute investment or financial advice given or endorsed by Pynk One Ltd and its affiliated companies. Information of this community is not intended to be relied upon for investment purposes. Investing carries a high degree of risk. Please make sure you invest aware, understand the risks of investing and seek independent advice if needed.
Terms & Conditions - Privacy & Cookies Policy
Copyright © 2022 Crowdsense Ltd.. All rights reserved.