저자: 미상 · 외국인
요약 보기
ARK Invest의 바이오·유전체 보유 종목을 분석. Intellia의 HAE 치료 데이터(96% 발작 감소)는 과학적으로 긍정적이나 대규모 현금 소진과 평가손실(PacBio -83%, Intellia -54%)이 부담. 10x Genomics는 장비 매출 감소에도 소모품 성장과 특허 합의금에 의존. Recursion은 Nvidia 지분 매각 후 12% 급락했으나 사업 파트너십은 유지. Generate Biomedicines는 상장 후 밸류에이션 부담(매출 대비 약 45배). 결론적으로 과학 진전과 재무 취약성을 대비시키며 유보적 시각을 유지.
다듬은 글
We are currently witnessing either the greatest medical revolution in history or one of the most expensive financial collapses the stock market has ever seen. Right now, ARK Invest is sitting on staggering unrealized losses up to 83% while their core holdings are proving they can functionally cure incurable diseases.
Is this a visionary foundation or just a multi-billion dollar hole in the ground? Today, we're breaking down the high-stakes 2026 genomic report including the shocking 96% success rate of a new gene editing therapy and the real reason Nvidia just dumped 7 million shares of a major AI partner. The survival clock is ticking for these companies and the next few months will decide the fate of these stocks.
Let's break it down. When you build a cathedral, uh for the first few years, you're really just digging a massive hole in the ground. To a passerby, it looks like a tremendous waste of resources, right? Just a chaotic mess of dirt and scaffolding. But, uh to the architect, it is the completely necessary foundation.
Right. And that's exactly the dynamic we need to analyze today, uh through the lens of our source material, Textus Pegatus. The report outlines a really comprehensive strategic evaluation of ARK Invest's core genomic holdings right now in the second quarter of 2026. And the central disagreement we have to unravel is whether this 2026 biotechnology sector is finally witnessing the, you know, the long-awaited industrialization of drug discovery or or if it's just trapped in a cycle of speculative capital destruction, right?
Just masked as innovation. Yeah, well, look, at some point, you have to actually stop digging and lay a brick. Otherwise, you don't have a foundation, you just have a very, very expensive hole. And the macroeconomic backdrop that's detailed in Textus Pegatus makes this incredibly clear. We are operating in a persistent high interest rate environment.
Sure, the macro's tough. It's unforgiving. When capital was virtually free a few years ago, you could dig forever. Now, companies are being forced to prioritize immediate clinical execution simply because of their cash runway. And, you know, for anyone tracking the space, cash runway isn't just a basic financial metric anymore.
It's a survival clock. It is literally the number of months a company has left before its bank account hits zero, assuming they keep burning cash at their current rate. Mhm. In this market, if your survival clock is ticking down and you aren't showing immediate commercial viability, the market punishes you severely.
I don't care how beautiful your cathedral looks on the blueprints. I hear you, but, uh let's actually look at what is in those blueprints. Based directly on the evaluation in Textus Pegatus, we're looking at the realization of the genomic revolution thesis, the convergence of multi-omics, generative AI, and in vivo gene editing.
It represents a profound inflection point in human medicine. We aren't just tweaking symptoms anymore. Well, drug discovery is being industrialized and to cross that threshold from a lab insight to a functional cure for rare diseases, it requires high conviction long-term capital. That is exactly why ARK Invest's strategy of holding double-digit percentages of the total float of these companies is so vital.
But, holding that much of the float? Just to clarify for the audience, the float is the regular shares available for public trading. By locking up such a massive portion of that float, ARK effectively acts as an institutional anchor. They're providing the structural stability these foundational platforms need to survive exactly the short-term market drawdowns you're talking about.
An institutional anchor or a sinking weight? I mean, you have to look at the sheer financial reality of what holding that much of the float has actually cost them. Textus Pegatus gives us the brutal spread between ARK's cost basis and the current price in April 2026. The unrealized losses are They're staggering.
And just so we're all on the same page, your cost basis is just the original average price you paid per share. When we look at that spread, they are sitting on an 83% unrealized loss in Pacific Biosciences. They are carrying a nearly 54% loss in Intellia Therapeutics. This isn't visionary anchoring.
This is a complete failure to adapt to a market that now demands clinical execution and profitability over, you know, speculative platform building. Okay, I'm not debating that the unrealized losses look severe on a spreadsheet today. They do. But, let's test your thesis about clinical execution by actually looking at the clinical benchmark for CRISPR commercialization.
We'll start with Intellia Therapeutics. Because if you want execution, look at the phase 1/2 data presented at the AAX I 2026 conference for their lead candidate, Longbovisalci. Textus Pegatus heavily highlights this. Right, for HAE. Yes, for hereditary angioedema. For context, HAE causes severe, unpredictable, potentially life-threatening swelling in the body.
Intellia administered a single 50 mg dose and achieved a 96% mean reduction in those attacks. Wow. Yeah. 31 out of 32 patients were essentially free from attacks at the data cutoff. But, here is the critical part. The 3-year durability of this effect. Gene editing has always faced skepticism about waning efficacy, the fear that the edited cells would eventually die off and the disease would just return.
Sure. Intellia is proving that wrong. They are proving that you can alter the genetic code permanently. They aren't treating HAE, they are functionally curing it. Look, the science is miraculous. Let me be absolutely clear about that. Achieving a 96% reduction in a debilitating genetic disease is a triumph of modern medicine.
Exactly. But, and there is a massive but here. We're looking at this through the lens of institutional investing. We cannot ignore the financial mechanics required to actually get that drug to market. Textus Pegatus points out their Q4 2025 net loss of $95.8 million. And remember that is after they already implemented a 27% workforce reduction and killed several research stage programs to save money.
Trimming the fat to extend the runway. But, they're still bleeding cash. So, my question to you is about this strategy of aggressive accumulation. In January 2026 alone, ARK added over 139,000 shares. They are buying a stock trading at $14.26 when their average cost basis is $30.90. Why keep pouring capital into a company that is still burning nearly $100 million a quarter in a macro environment that is openly hostile to debt?
Because the risk profile fundamentally changes the moment you clear these clinical hurdles. Look, I get it. Funding late-stage biotech looks like a massive cash burn right up until the moment it isn't. But, Intellia's heavily de-risked right now. They have $605.1 million in liquidity. Which buys them some time, yes.
It gives them a clear cash runway straight through to the second half of 2027. And what happens between now and then? They are submitting their biologics license application, the final FDA hurdle, in the second half of 2026. This timeline perfectly positions them for a commercial launch in early 2027.
If you are a high-conviction investor, you accumulate shares while the market is punishing the cash burn right before the commercial revenue turns the spigot back on. Assuming the FDA review goes perfectly, sure. But, Intellia's gene editing doesn't exist in a vacuum. You can't edit what you can't map.
Right. To even find the genetic targets for these therapies, you need highly sophisticated mapping tools. And that mapping bottleneck brings us right to 10x Genomics, the company generating the biological data in the first place. Exactly. And 10x Genomics is the undisputed leader in single-cell and spatial biology.
To put that in perspective, uh older sequencing methods were like putting a piece of tissue in a blender. You get a smoothie of genetic data, but you lose all the context of where those cells were in relation to each other. Right. Spatial biology is like a highly detailed GPS map. It lets researchers see exactly which genes are turned on or off in their exact original location within a tumor.
Textus Pegatus specifically highlights their Xenium platform's central role in the Bio Optimist Stellan initiative. The 100,000 patient specimens project. Yes. They are profiling 100,000 specimens to create the world's largest clinically linked spatial biology atlas. This is the data engine powering the generative biology you see down the line.
And structurally, ARK has played this one beautifully from a market perspective. We should definitely talk about how they've handled tactical rebalancing here. Okay. When a stock has a massive run-up, smart portfolio management dictates that you trim your position to secure gains and manage your risk weighting.
As 10x Genomics surged 200% from its 2025 lows, ARK actively sold over 150,000 shares in March 2026. That is taking profits on a picks and shovels play exactly the way a disciplined institutional investor should. I'm all for securing gains, but I mean, I have serious questions about the underlying quality of the surge that allowed for that profit taking in the first place.
Just look at the 2026 revenue guidance detailed in Textus Pegatus. They are guiding for a completely stagnant 0% to 4% growth. And why? Because underneath the hood, there is a massive 39% year-over-year decline in instrument revenue. Customers are shifting away from buying this expensive capital-intensive hardware because academic and research budgets are incredibly constrained right now.
But, their margins improved. Right, I know the counterargument is to point to their improved 69% gross margin and their massive $523.4 million dollar pile as proof of corporate resilience. But Text Topia notes explicitly notes that a critical component of that cash pile was a 44.1 million dollar non-recurring patent litigation settlement.
On top of that, there was another litigation gain that artificially reduced their operating expenses by 17%. It's still cash on the balance sheet. But it's a legal windfall. It's lawyers winning in court, not organic business growth. It's protecting their intellectual property, which is the core asset of literally any picks and shovels company.
But even if you strip out the litigation settlement, the consumables revenue, the chemical reagents labs have to buy every single time they run a test, that still grew. The ecosystem is expanding even if hardware sales temporarily slow down due to macro budget cycles. Fair enough. But let's follow the data trail here.
If 10X Genomics is providing the spatial biology hardware to map these cells, you then have to deal with the overwhelming amount of data they produce. You need massive AI supercomputing infrastructure to actually process it and find the drug targets. That brings us to Recursion Pharmaceuticals. Right.
The integration of technology and biology. The tech bio play. Exactly. Recursion is championing this integration with their BioHive 2 system. They have built a 23 petabyte data set of gene compound relationships. They're essentially building a biological foundation model, similar to how ChatGPT learned language, but this AI is learning the language of human cells.
It's fascinating tech. It really is. They are computationally testing how millions of potential drug compounds interact with disease cells before ever touching a physical petri dish. According to Text Topia, this computational approach is yielding 30% to 60% improvements in clinical trial enrollment speed because they are targeting the exact right patient populations from day one.
And unlike the stagnant hardware growth at 10X, Recursion's top line is exploding. They reported 35.5 million dollars in Q4 2025 revenue, which is a massive 671% year-over-year increase. Driven largely by milestone payments from their partnership with Roche. Let's pause on that 671% increase for a second because yeah, it sounds phenomenal in a headline, but you just said it yourself.
It's driven by milestone payments. Those are lumpy, unpredictable injections of cash dependent on hitting very specific research targets, not a steady stream of recurring product revenue. But honestly, the bigger issue with Recursion isn't just the burn rate. It's the market sentiment and a massive institutional red flag highlighted in Text Topia.
We have to talk about divestment. I had a feeling you'd bring this up. I have to. When a major strategic stakeholder suddenly liquidates their entire position, the market notices. On February 19th, 2026, Nvidia completely exited its 7.7 million share position in Recursion. That immediately triggered a 12% drop in the stock.
Nvidia isn't just any investor. They are the world's undisputed leader in AI hardware. If the architects of the AI revolution are fully divesting from an AI-driven drug discovery platform, how can you not see that as a glaring warning sign? Look, I will concede that the Nvidia divestment looks absolutely terrifying on paper.
If I see the king of AI hardware dump over 7 million shares of an AI bio stock, my heart skips a beat, too. But you have to actually look under the hood at the operations because the market dramatically misinterpreted that move. It was financial noise. Noise? 7 million shares is noise? Text Topia clarifies that the operational partnership between Recursion and Nvidia remains completely intact.
Recursion is still actively using the Nvidia-powered BioHive 2 system to train their foundation models. The multi-year collaboration they established in 2023 to distribute AI models via Nvidia's cloud is still completely operational. Their actual business hasn't skipped a beat, and their overall cash runway extends all the way into 2028.
Still, the optics So we use ARK to purchase 1.25 million shares of Recursion shortly after that Nvidia disclosure. That was a denial. That was classic bargain hunting. They understood that the fundamental mechanics of the partnership hadn't changed, only the share price had artificially plummeted. Bargain hunting or just catching another falling knife.
They now carry a 63% unrealized loss on a 341 million dollar total investment in Recursion. You can have all the petabytes of data and Nvidia chips in the world, but if the market is re-rating your valuation downward because you're burning almost 400 million dollars a year, that hurts. Well, but let's move from that macro level cloud computing AI processing back down to the granular physical challenge of actually reading DNA accurately because this is where the financial realities we've been debating are perhaps the harshest.
Let's talk about Pacific Biosciences. Okay. PacBio is definitely navigating a challenging transition. I'll give you that. But their operational progress at the ground level is undeniable. They are the leader in highly accurate long read sequencing. For the audience to grasp why that matters, think of traditional short read sequencing like trying to read a novel that's been put through a paper shredder.
You can piece together small words, but you miss the overarching context. Right. Long read sequencing lets you read whole paragraphs of DNA intact. This is absolutely crucial for catching complex structural variants like large insertions or deletions of DNA that are often the root cause of undiagnosed rare diseases and complex cancers.
To understand their business value, you really have to look at the razor and blade model. Where you sell the machine cheap to lock them into buying the expensive chemicals. Exactly. Think of their flagship Revio sequencing system like an incredibly sophisticated high-end Nespresso machine. The initial machine sale is the razor, often sold at a lower margin just to get it into the lab.
But the real value is in the metric we call pull-through. Pull-through is the recurring revenue generated by the chemical consumables, the biological coffee pods, that the lab has to buy every time they run a test on that machine over a year. And the numbers there are decent. Better than decent. Text Topia reports a staggering 252,000 annualized pull-through per Revio system.
That proves robust, growing utilization of their installed base. Labs aren't just buying the machines, they are running them constantly, which is why PacBio hit record Q4 consumables revenue. I understand the Nespresso analogy, and the pull-through numbers are solid, but at what cost? What does it actually cost them to generate that 252,000 per machine?
Let's look at the fundamental ranking provided in Text Topia, which gives Pacific Biosciences a C grade. Well, grades are subjective. But the math isn't. This is where ARK's capital destruction is most evident. They invested an estimated 353 million dollars into this company, and today it is worth approximately 60 million dollars.
That is an 83% loss. PacBio has a trailing operating margin of -134.5%. The transition to the Revio system is expensive. But that means for every dollar they make on those biological coffee pods, they are spending significantly more than two dollars just to run the company. Furthermore, Text Topia notes their gross margins are at risk of falling below 40% due to demand weakness.
Yet, ARK continued to accumulate 423,000 shares in April 2026 alone. To maintain a nearly 15% ownership stake in a company burning cash this rapidly while operating at a negative 134% margin frankly defies financial logic. It defies short-term financial logic, absolutely. I won't argue that point. But it does not defy long-term technological logic.
The entire shift toward precision medicine relies on long read sequencing. You simply cannot develop targeted gene therapies if you can't see the structural variations in the genome clearly. It's a multi-year growth flywheel, and right now we are in the painful trough of that cycle. Very painful. But if we're going to contrast the struggles of a legacy hardware sequencer like PacBio, we have to look at the brand new AI-native approach to biology.
This brings us to the final core holding highlighted in the report, Generate Biomedicines. The generative frontier. Exactly. Generate Biomedicines is pioneering an entirely new field. They just completed their initial public offering in February 2026. What makes them so unique is their technological moat.
They use a tight design, build, test, learn loop to create novel proteins de novo. From scratch. Right. Mechanics-wise, this means instead of scouring nature to discover a protein that happens to bind to a disease target, which takes years of trial and error, they use AI to custom design a bespoke protein from scratch that fits the disease target perfectly.
It's like a custom machined key for a specific lock. And it works. Their phase three asset, GB0895, utilizes a differentiated 6-month dosing advantage for severe asthma. Compare that to current monoclonal antibody therapies on the market that require patients to come in for dosing every two to eight weeks.
That is a massive quality of life upgrade for patients. The custom key analogy is great, and the science is undeniably cool, but let's look at the reality of how the market treated that IPO, that initial public offering. Going public is supposed to be a triumphant moment of raising massive capital. Generate raised 400 million dollars, giving it an implied valuation of over two billion dollars.
A solid raise. But the market reception documented in Text to Pegado was ice cold. The stock opened at $15.11 and plummeted 21% on day one. By mid-April, it was trading near $12.01. To understand why, you have to look at the valuation, specifically the price-to-sales multiple. In simple terms, this metric shows how much investors are willing to pay for every single dollar of sales a company actually makes.
They're clinical-stage biotech, though. But they carry an estimated 44.8x price-to-sales multiple. In a high-rate environment, paying nearly 45 times sales for a company that has a trailing net loss of $223 million reflects exactly the type of speculative bet that investors are currently punishing. It's brilliant science, but the valuation is priced for absolute perfection in a market that is fundamentally imperfect right now.
I'll push back on that because you are trying to price a generative biology pioneer like a traditional mature software company. A 45x multiple looks high if you're selling enterprise software. But Generate has $672 million in combined liquidity post-IPO. That cash is the necessary fuel for this exact stage of clinical development.
They don't need to be profitable today. They need to finish their phase three trials and prove that AI-designed proteins can beat naturally discovered ones in the clinic. I I understand the vision. I really do. So, taking a step back and looking at the entire portfolio outlined in Text to Pegado, my conclusion remains incredibly firm.
The big biotech bet in Q2 2026 is fundamentally about crossing the threshold from laboratory insight to commercial reality. The 96% efficacy data from Intellia and the massive spatial mapping scale of 10x Genomics, they prove that the underlying science is sound and moving rapidly. Yes, the financials look strained across the board, but bringing functional one-time cures to the market requires bold, long-biased capital that won't retreat at the first sign of macroeconomic headwinds.
We are still building the cathedral. And I will counter by saying that while I completely agree the science is advancing, sometimes miraculously so as we saw with the Intellia data, the financial structures supporting that science remain highly precarious. Text to Pega paints a deeply concerning picture of massive unrealized losses, negative operating margins, and a dangerous reliance on sporadic milestone payments just to keep the lights on.
It's the nature of the industry. But technological success simply does not automatically translate to sustainable public market valuations. You can successfully design a novel protein de novo from scratch, and you can map out every cell in a tumor, but you will still destroy shareholder value if your cost basis and your operating margins are completely misaligned with the reality of interest rates.
The industrialization of drug discovery is absolutely happening, but the cost of funding it has been devastating for those who bought into the hype at the peak. It's a fair warning. We both clearly acknowledge the profound complexity of investing in generative biology and multi-omics. You are trying to perfectly time the intersection of the most difficult engineering and the most difficult biology in human history.
Exactly. And looking at all the companies we discussed today, I think we also agree that the upcoming mid-2026 phase three readout for Intellia Therapeutics is going to be the ultimate bellwether for this entire sector. Absolutely. It is the clearest test case. If they succeed and get that BLA submitted, it validates the long-term thesis that these massive upfront investments pay off.
But if they stumble, the financial realities of cash burn and short runways are going to crush whatever optimism is left in the space. I couldn't agree more. I strongly encourage everyone listening to pull up the Text to Pegado analysis for themselves. Weigh the clinical milestones and the biological breakthroughs against the financial fundamentals and the cash burn rates we've debated today.
Look closely at how these companies are navigating the current macroeconomic environment. And as you review the data, keep asking yourself that core question. Are you looking at the necessary, albeit messy, foundation of a medical revolution, or are you just staring down into a very expensive hole in the ground?
Because ultimately, the only difference between a visionary masterpiece and an abandoned ruin is time, capital, and execution.
Comments 0