The Borlaug Report #1: Twist Bioscience
Thanks for reading. I am a public equity investor posting his ideas here to stay sharp while he looks for his next seat. If you like my writing, please feel free to subscribe or message me with feedback (@borlaug_ on Twitter is fastest).
TL;DR (Thesis Introduction)
Twist Bioscience is the leader in gene synthesis, or the “writing” of genes using its proprietary silicon-based technology to use traditional chemistry methods at high throughput. It also markets a competitive offering in NGS library prep kits. Finally, it is growing an end-to-end antibody discovery engine leveraging its technology, offered as a CRO service.
Since its IPO in 2018, Twist has been subject to a lot of skepticism - it originally entered the low-margin DNA synthesis industry with pricing as its competitive advantage, drawing the criticism that the business could never turn a profit, and if it did, a small one from a margin standpoint. Twist was also hit with a short report that builds on this skepticism, alleging that Twist’s accounting for investment in capacity expansion was hiding just how low these margins are.
Since then, Twist has introduced adjacent products that are highly margin-accretive as well as expanded its offering into higher-value parts of the gene synthesis workflow (such as producing proteins) and amassed a portfolio of call options in its antibody drug discovery business.
Twist has become a crowded short at 21% of float sold short (~teens DTE) It appears clear that the short report has led many investors to miss the forest for the trees, creating a massive opportunity to own Twist shares at a discount. The segment covered by the Scorpion Capital report will account for only 40% of revenue, and the largest segment is well on its way to being the most profitable. Furthermore, Twist’s antibody discovery business has amassed a royalty stack for it that could be worth as much as the whole company is today. Given its status as a consensus short, I believe that if Twist shows margin upside from any of a variety of sources as it ramps its new facility, it will cause a significant re-rating event, which I believe is worth 50-115% in upside.
Company & Technology Background
Twist Bioscience was founded by CEO Emily Leproust and 2 of her peers, Bill Banyai and Bill Peck in 2013. She had the idea for a next-generation gene synthesis company while at Agilent and decided to start a company around it, which led to several legal disputes, including a trade secret theft case as the company went public in 2018. These cases have been settled and Twist operates its platform free from its past.
Shortly after is IPO, Twist launched its NGS tools business, using the same chemistry know-how it needed to innovate in synthesis to launch library preparation kits, panels, and reagents for next-gen sequencing. The kits win at the enrichment step of preparation, which increases read quality and uniformity while lowering sequencing costs. Its technology works on all sequencers today. Since launch, Twist has released a large variety of new products for specific applications, which has enabled it to enter the diagnostics market as a picks & shovels provider. It is specified into every major liquid biopsy company filing (meaning the users can’t switch). This alone represents a large opportunity in outer years, though I am less bullish on pan-cancer liquid biopsy. Regardless, several indication-specific tests are heading to market shortly.
Subsequently, Twist began an antibody discovery business, using its “Library of Libraries” to help customers screen libraries of antibodies that they didn’t think could work and identify targets that might. Using this tactic, Twist was able to win deals in an otherwise challenging market. It now has enough credibility to compete with the other 3-4 major players consistently.
Twist’s synthesis products win primarily on 2 things: 1) Price, and 2) predictability of errors. While synthesis is challenging to do accurately, Twist has been able to prove an error rate of 1/1,000 bases with lower overall variability.
Twist differentiates itself in NGS using a proprietary “double-stranded DNA” probe, which effectively doubles the amount of target-seeking probes that capture target sequences in enrichment. Below is from a paper written in 2018 showing how Twist outperforms peers in library enrichment. As you could guess, top-right is good.
It also wins in uniformity, where customers will get less useless data from runs that have under-sequenced sample or off-target reads. Below is an illustration:
Industry Background
Twist’s gene synthesis business serves the R&D market - a combination of academic, biopharma, industrial, and other customers using synthetic biology approaches to create new products. Similarly, its NGS products are used in R&D end markets where sequencing is done, which includes a lot of the same customer base. However, NGS also extends to clinical end-markets, where companies that market diagnostic tests using next-gen sequencing need sample and library preparation kits to enrich their samples to enable better data collection.
Both of these markets are highly competitive but under-innovated. Gene synthesis services have been offered by the likes of Genewiz (AZTA), IDT(DHR) and Agilent for over a decade. Several companies offering benchtop DNA printers have arrived at the market such as Telesis (TBIO), DNA Script, and Molecular Assemblies, though the value proposition is somewhat different as compared to services (cost vs turnaround time). The service portion of this market is a low-margin, low-TAM business, preventing larger players from investing in it.
NGS library prep kits is a market largely dominated by Roche’s KAPA kits, as well as QIAGEN’s solutions (~15% of the prep market, which includes sample prep as well). Twist has been rapidly taking share, but the larger players still optically own the market. Much like synthesis, little product innovation has been seen from either company, as the smaller historical TAM has kept companies from investing. These companies also compete with Twist’s array of panels and reagents. Illumina has some of its own products as well, but are seldom used.
Antibody discovery is competitive as well. Several competitors have a head-start on Twist including Adimab, Abcellera (ABCL), and Omniab (OABI). These companies all have more programs, active and completed than Twist, but focus solely on the discovery piece whereas Twist is competing as an end-to-end contract research organization following its acquisition of Abveris.
Overall, while Twist competes in 3 competitive markets, it is differentiated in each, and each has tailwinds from end market expansion as well, leaving large amounts of room for such a player.
Financial Overview
Addressable market: Twist believes that its addressable market is just shy of $6B today, comprised of the following:
Liquid biopsy/MRD will be a much larger market in the future as well, as the first products the industry have seen have either just launched or launches this year.
Revenue: From FY18-FY22, Twist’s Synbio business has grown at a CAGR of 36% from $18M to $62M, giving it 15% market share assuming it only currently reaches “DNA buyers” today. From FY19-FY22, its NGS tools business has grown 68% from $21M to $99M and becoming the largest segment at Twist. The company also earned $24M in revenue in FY22 from biopharma services, about $10M of which was from a small CRO acquisition.
Gross margin: Twist’s gross margins look messy today, but have arguably trended in the right direction. Gross margins were negative from 2016-2018, turning positive in 2019 and reaching 41% in FY22. Gross margins are guided at 39-40% for this fiscal year as new capacity impacts margins, but management also guided 49% for FY24 as utilization approaches the right levels. This line is arguably the key point of controversy for Twist, but it is driven by a lot of moving factors, including product mix, growth, and Twist’s capacity utilization of its new Factory of the Future.
Cash burn: Twist is guiding to a cash burn of ~$200M, of which $46M is opex related to DNA data storage and $50M is capex. This would leave the company at $300M in cash at year end. Twist’s guidance also set forth its expectation that it would still have another $170M in cash by YE24. Investors are skeptical here given the above.
Conclusion: Overall, Twist looks very much like a startup, but investors are treating it like a structurally-broken business that can’t make money. I believe the alpha here is in figuring out where margins ultimately will land.
Peer Group and Relative Valuation
High-growth life science tools companies command a premium to many other sectors. Twist used to participate in this valuation regime until its profitability became a focal point. However, despite street estimating above-group growth rates, profitability has created a massive difference in valuation tied to this profitability, optically 70%. If Twist shows that its margins are real, contrary to what the short report against it alleges, I believe a re-rating will be triggered (not all the way to group, as much of group is more profitable.
The Opportunity - A Sum-of-the Parts That Can be Realized with One of Many Catalysts over the Next Year
The short version of this section is as follows: TWST would still be cheap even if you assigned a value of $0 to the controversial gene synthesis business. Several things can happen to unlock this value.
TWST already looks optically cheap at its current trading multiple of 2.6x NTM sales (Price/Sales - assuming all cash is used getting to profitability and thus not included in valuation). However, I believe it is even cheaper than meets the eye given the call options that antibody discovery business has generated.
How much is the average antibody discovery program worth?
As of its latest earnings call, Twist had 63 programs, active or completed, in which it is owed milestones and/or royalties on the products it helped discover. These have some level of value that is hard to triangulate, but putting a floor valuation should be sufficient for valuation.
Drug discovery, where one company does discovery and a partner does the rest is a relatively novel concept. In the antibody space, however, several companies have done a large number of deals in the antibody space, including Omniab, who has presented investors with illustrations on what deals usually look like:
Using the first example, use macro-level drug stats to arrive at a floor NPV value of each program. For simplicity’s sake, I used 5 milestones and assumed an equal fraction of the value for each.
Assumptions below:
$400M is roughly what the average forecasted peak sales per drug was across all drugs in 2022 according to Deloitte, but most of the NPV value is in milestones for this exercise given the high discount rate. The milestone probabilities are an aggregate as well.
Plugging in these basic assumptions into a DCF, each program is worth at least $3.2M. This is what I believe the lowest value of a single program could be. However, it also could be much higher - the milestones that Omniab has in its deal models look extremely low, and there is a chance TWST gets much larger ones today on later levels of success. Most of the time, early milestones are $5-$10M each with bigger ones for harder-to-hit milestones. A good illustration of this is TWST’s deal with Boehringer in which it claims $710M in cumulative milestone payments. The other lever one could use is discount rate, where I used 15%, a standard in single-drug valuation due to single-drug risks, but since we are evaluating on a portfolio level, there is an argument for a lower rate. Below is the sensitivity for those two axes:
The above shows that AT LEAST 25% of TWST’s market capitalization is covered by the NPV of programs. If the market uses or shifts to using a lower rate as the portfolio grows/rates decline in 2024+, or TWST is winning bigger milestones, this number could plausibly be 30-50%. This doesn’t leave much for the base business, which already trades at an extremely low multiple.
NGS Business - Increasingly Less-Hidden Gem
The NGS tools business is really the hidden gem investors should pay attention to. The reason is simple - Twist is innovating in a sleepy segment of the NGS tools business with high-margin products that provide significant value to customers. As mentioned in the product overview, Twist’s NGS products save money and time, and have clinical utility. Each of the major liquid biopsy players, for example, have filed their tests using Twist NGS products, meaning Twist gets ~$65 every time one of those companies sells a test. The customer base is rapidly expanding, and Twist just announced a new whole-genome product at AGBT this year which could propel growth even further.
Just how profitable the NGS business is is hidden in the messy gross margin we’re seeing today, but management has repeatedly given useful background. Jim Thorburn (CFO) said this at the Barclays Conference in a fireside chat with Luke Sergott last month:
“As we scale the revenue, we get increased contribution margin. The overall contribution margin for the company we're targeting at 78%. We have 2 businesses, NGS and Synbio. NGS contribution margin, which is the difference between price and material cost is about 80%. The Synbio contribution margin today is about 65% to 70%. And as we scale the factory of the future and launch the fast genes, the contribution margin on the fast genes is higher, because we're getting a price premium.”
Most investors write this comment off every time it is said because contribution margin is extremely misleading for gene synthesis, given its service-like nature that relies on a headcount and capital-constrained overhead footprint. HOWEVER, NGS kits are much more widget-like - they scale better, they are less capital-intensive to make, and offer much greater margin potential. Look at almost any large-cap tools company for a comparable. I’d estimate that the kits business could scale to a gross margin of 65-70% if it were a standalone business. QIAGEN has a lot of similar assets Twist (NGS tools, PCR/amplification technologies), with a very high consumable mix - it sports a mid-60s gross margin and a high-20s EBIT margin at scale. Diagnostics is a big part of their business, but given the level of competition in diagnostics, the tools piece of the business is likely margin-accretive.
DNA Synthesis - Probably Not as Low-Margin in Potential as It Appears Today
The biggest debate on Twist is whether or not its DNA synthesis business will burn cash forever. I would assert that it won’t and that it could actually show upside to what investors think future profitability looks like.
The best comparable here is Azenta Life Sciences. Azenta owns GeneWiz, a competitor in gene synthesis with an array of adjacent service offerings. The company provides segment information that breaks out what it makes on services. 70% of this business is sequencing and synthesis services. This comparable segment as a whole had a 48%47% gross margin in FY21/22 along with a 7%/5% operating margin. This illustrates the above point about contribution margins - services place a lot of the cost in overhead. 5% is where I’d conservatively place the peak profitability potential of Twists DNA synthesis business.
However, there is a lot of upside potential here - TWST is finding ways to make more dollars per head/square foot of production capacity. There are as follows:
Longer Genes (Today): Twist has regularly released service products that produce longer and longer genes.
Faster Genes (Ramping): With the Factory of the Future, Twist will offer “fast genes” for which it will charge a premium to offer its existing services with faster turnaround times.
Antibody Production (Ramping): Twist has developed an end-to-end workflow that moves from producing the genes to expressing them into the antibodies of interest on behalf of customers and characterizing them. Twist can’t do all kinds of antibodies yet, but could easily do so in the future.
Longer genes has been playing out since IPO - TWST gives the number of genes it ships in its 10-k. Revenue per gene has gone from $65 in 2017 to $110 in FY2022. With the latter 2 initiatives ramping over time, I expect these numbers to go even higher. Antibody production could be the most exciting, as selling proteins can be lucrative - Bio-Techne is the leader in production production and makes 45% operating margins doing it at scale. Even moving a short distance in that direction would go a long way for TWST. In the linked video above, the team shows that single-domain antibodies start at $270 and the more complex IgG antibodies start at $310 (could be higher but can’t see second digit), 145%-182% higher than genes.
Things that can happen that will cause a re-evaluation of the above arguments (catalysts):
Upside from TWST’s NGS business, which is gross margin-accretive, could drive margins up over the next couple of quarters.
Margin upside could come from Twist’s SynBio business, which would come from fast uptake of Fast Genes and Antibody production.
Milestone awards from Twist’s discovery business, which optically come in at 100% margin and extend Twist’s runway.
Exiting DNA data storage: If Emily threw in the towel on this business, the stock would immediately gain interest from a more pragmatic investor base.
M&A: I believe Illumina, Bio-Techne, and Abcam would all be extremely interested in owning an asset like Twist. Others such as Danaher would become interested in margin potential became more obvious. Illumina strikes me as the most strategic AND the most motivated acquirer given its capital allocation track record of late.
If Re-Rated, Where Will Valuation Go?
Lots of investors get lazy and miss a key nuance with sales multiples- not all sales are worth the same multiple. A smart hedge fund I know asks junior candidates the following in an interview: “Two companies trade at the same P/E, but different sales multiple. Why?”
The answer is margins (and tax rates if you don’t think that falls under the umbrella). Taken a step further, you shouldn’t be using a sales multiple as an analytical tool at all if you don’t have an opinion on how profitable the company can/will be.
To create a sum of the parts, I’ll put forth a base-case operating margin profile for each segment and apply a 25% tax rate to get to a look-through earnings number and apply a PEG to this number based on sustainable LT growth rates for each. I used 1.5x PEG to be conservative as this reflects a 30%-40% discount to the median/average tools company with out-year estimates (2.2x-2.4). This is more in-line with CROs, which as a group trades at long-term lows on valuation.
My base case is below - there should be few surprises if you’ve been following up until now. It implies TWST should trade at at least 5x forward sales.
My bear case implies that SynBio is worth $0 and that all other segments have lower margins. It also assumes much lower medium-term growth for NGS, which together imply Twist should trade at 2.4x forward sales for a downside risk of 20%.
The pie-in-the-sky/bull case requires the following:
Twist’s Synbio segment starts looking a lot more like Techne’s protein production operation
Royalty milestone portfolio gets revalued at 10% discount rate and assumes more milestones per deal than Omniab
If these conditions were met, the valuation would imply a 114% return from its current market cap, which would largely be realized in price given the short time frame.
Key Debates & Risks
Most of this case has been a rebuttal of widely-circulated risk but I will review in bullet-point form:
Margins: The market believes Twist has extremely low margin potential, which makes its stated path to profitability hard to believe, rendering the company almost worthless until one can price dilution.
As shown above, I believe comparable businesses for each segment suggest that not only WILL Twist become profitable (when not if), it could be innovating its way to much HIGHER margin potential than many realize.
SynBio’s competitive advantage: The short report alleges that Twist has no advantages against benchtop synthesis companies, and that eventually it will have to raise its pricing against other service providers, rendering it meatless there as well.
I do not believe that benchtop providers will be a serious risk, and Twist is working on its own enzymatic synthesis approach, which would allow it to make its own superior instrument if necessary
SynBio is not important to the thesis - in my most bullish case it is 21% of a valuation showing >100% upside.
What if NGS growth slows down?: One of the bigger valuation drivers is the growth rate of NGS over the next several years. The clinical NGS market itself is growing 15-20%, per Illumina, which drives its longer-term consumable growth estimates. My assumptions are at a premium to that/the high end.
This is because Twist is levered to nascent products and is taking share from incumbents. Most incumbents will never enjoy the benefits of liquid biopsy growth the way TWST will from a library prep perspective, for example.
Twist also routinely launches new products that a clearly stimulating growth in this segment. For example, the WGS product got a lot of attention at AGBT and should contribute to growth later this year.
Stock-based compensation: In FY2022, Twist recorded $80M in stock-based compensation, amounting to 39% of sales, more than double the prior year.
This is an issue with most life science companies, but the magnitude was smaller than meets the eye - $20M of this was due to acquisition-related stock comp, and half of this was forfeit as a key deal condition required to earn it was not met.
In 1Q23, all of this was reversed - if you make this adjustment, SBC still came down meaningfully in SG&A.
Within R&D, SBC only grew 10% during the same quarter.
Cash Burn: Beyond margins, Twist is spending too much money - it is over-investing in DNA data storage, a highly speculative product to spend a lot of money on.
The good news is that the short report placed this front-and-center, Twist’s transparency on its intended path to profitability has grown dramatically since. Given the above commentary on margins, I think the story for the base business is compelling enough, especially at these levels, to offset investment in DNA data storage.
I’d also argue DNA data storage as a bottomless pit for zero-return spend is pricing into the stock. If Emily announced that the program was being discontinued, the stock would start to run.
Conclusion
Overall, I think that TWST at $14 per share represents a de-risked growth company with enormous potential. I also believe that the potential is and will become more obvious over time as Twist matures, given the similarity of its products and services to existing, mature businesses that are visible in the market today.
With 21% of its shares sold short, and a lot of things that could surprise to the upside, both on earnings calls and in between, I expect shares to re-rate over the next 6-18 months in a significant way. Some of these things, such as exiting DNA data storage, are firmly the company’s control, whereas others are purely factors of clinical end-market growth, which is recession-resistant, as is much of the R&D market.
Disclosure: I own shares of TWST in my personal account.