When the Rockets and the Robots Go Public: What the SpaceX, Anthropic, and OpenAI IPOs Will Do to the Listing Market, and Which Healthcare and Life Sciences Companies Get to Ride the Next Hype Cycle
Video Preview
🎧 Part I Podcast free on Spotify.
🎧 Part II Podcast episode for paid subscribers only. Also available on Spotify.
To listen to paid episodes in Apple or Spotify, link your Substack subscription via the show settings on those platforms (instructions inside the Substack app under Subscriptions → Podcast).
Table of Contents
The setup: three filings that ate the entire tape
Why these three reset the table for everybody else
The plumbing: what a mega-IPO actually does to pricing and appetite
The healthcare lag, and why it always lags
The class of 2025 and what Hinge and Omada actually proved
TechBio, or what happens when the AI money finally notices biology
Who gets to ride, and who gets left at the gate
The stuff nobody wants on the roadshow slide
Abstract
This piece argues that the 2026 mega-listings (SpaceX at a reported ~$1.77T, Anthropic near $965B, OpenAI somewhere between $852B and $1T+) are not really IPOs in the normal sense. They are gravity wells that reset comps, soak up banker bandwidth, and reprice retail psychology. Key data anchors:
SpaceX: confidential filing April 1, S-1 circulated ~May 20, ~$18.67B 2025 revenue post-xAI merger, Starlink ~61% of that, ~$4.9B GAAP loss, super-voting Class B, up to 30% retail allocation, June 2026 target that could slip to 2027.
Anthropic: confidential June 1, ~$47B run-rate, projecting its first profitable quarter in Q2, ~31% LLM revenue share, October-ish listing window.
OpenAI: ~$25B run-rate, ~$2B/month, reportedly losing ~$1.22 per $1 of revenue, ~$27B 2026 burn, Polymarket pricing an actual listing at ~4% by end of June and ~40% by year-end.
Macro context: ~80% of global VC went to AI in Q1 2026; OpenAI, Anthropic, xAI, and Waymo alone took >$188B of a record $300B quarter.
The healthcare and life sciences thesis: the wave does not lift everyone, but it cracks the window. The likely riders are the “Health Tech 2.0” cohort (profitability, embedded AI, clinical evidence) behind Hinge and Omada, plus the vertically integrated TechBio names that get to wear the AI label at the exact moment public markets are paying up for it.
The setup: three filings that ate the entire tape
Start with the scoreboard, because the scoreboard is absurd. SpaceX filed confidentially on April 1, circulated a public S-1 around May 20, and as of early June was being marketed at a fixed price near $135 a share for a valuation of roughly $1.77 trillion. Earlier reporting had floated everything from $1.5 trillion with a $30 billion raise up past $2 trillion with a raise north of $75 billion, so the only safe statement is that whatever it prints will be the largest IPO in history by a wide margin, leaving Saudi Aramco’s 2019 record of about $29 billion looking quaint. The financials underneath are stranger than the headline. Consolidated 2025 revenue landed around $18.67 billion, but that already folds in xAI, which SpaceX absorbed in an all-stock deal in February, and the combined entity posted a GAAP net loss of roughly $4.9 billion against adjusted EBITDA of about $6.6 billion. Starlink is the actual business, throwing off something like 61 percent of revenue and most of the operating profit, while the rocket part remains a magnificent capital incinerator. Governance is its own punchline: Class B shares carry ten votes apiece, which means Musk cannot be removed without the consent of the people who answer to Musk.
Anthropic went confidential on June 1, days after closing a $65 billion Series H at a roughly $965 billion post-money led by Altimeter, Dragoneer, Greenoaks, and Sequoia. Run-rate revenue had crossed $47 billion by late May, up from something in the neighborhood of $10 billion a year prior, which is a growth curve that makes valuation committees giggle nervously. The interesting wrinkle is that Anthropic is projecting its first profitable quarter, with operating income in the hundreds of millions, and Counterpoint had it edging OpenAI on large-model revenue share at roughly 31 percent to 29. A fall listing, maybe October, is the working assumption.
OpenAI filed its own confidential S-1 in the same window with Goldman, Morgan Stanley, and JPMorgan steering, targeting a valuation between about $852 billion and a clean trillion, off a run-rate near $25 billion and roughly $2 billion a month coming in the door. The catch, and it is a large catch, is that the company was reportedly losing about $1.22 for every dollar it earned in the first quarter, with 2026 cash burn estimated around $27 billion. Traders noticed. On Polymarket, the probability of an actual OpenAI listing sat near 4 percent by end of June and only around 40 percent by year-end, which is the market’s polite way of saying a confidential filing and a printed prospectus are different animals. Zoom out and the concentration is the real story: something like 80 percent of global venture dollars went to AI in the first quarter, and four names (OpenAI, Anthropic, xAI, Waymo) hoovered up more than $188 billion of a record $300 billion quarter. Everything else, including all of healthcare, was fighting over the scraps.
Why these three reset the table for everybody else
Here is the thing nobody in a pitch meeting says out loud: a trillion-dollar IPO does not compete with a billion-dollar IPO, it warps the space around it. When the benchmark deal of the year is priced in trillions, a $1 billion digital health float stops registering as a number a generalist portfolio manager has to think about. That cuts two ways. On the bad days it means a perfectly good health tech company gets ignored because the entire buy-side is busy modeling Starlink ARPU and arguing about whether Anthropic’s gross margins are real or a function of subsidized compute. On the good days it means the window everybody has been waiting on finally gets pried open, and the spillover liquidity sloshes downhill into smaller deals that would otherwise never have gotten a roadshow.
The bandwidth problem is underrated and very concrete. There are only so many lead-left bankers, so many crossover funds with real check-writing authority, and so many weeks on the calendar that are not blacked out by earnings or holidays. Three monster offerings will tie up Goldman, Morgan Stanley, and JPMorgan for much of the back half of the year, which means a kidney-care analytics company or a value-based oncology platform is queuing behind the rockets whether it likes it or not. The flip side is the halo. The reason anyone in life sciences cares about a satellite company has nothing to do with satellites. It has to do with whether a string of successful large listings restores the thing that has been missing since 2021, which is the basic willingness of public investors to underwrite a story that has not fully matured. Confidence is the commodity in shortest supply, and mega-IPOs that trade well manufacture it. Mega-IPOs that break manufacture the opposite, faster.
The plumbing: what a mega-IPO actually does to pricing and appetite
Mechanically, the wave moves through a few channels that matter to anyone planning a 2026 or 2027 exit. The first is comps. Aftermarket performance of the marquee deals becomes the mood ring for everything that follows, and the order of operations is brutal: if SpaceX prices and pops, the bankers shopping the next twenty deals get to point at a green chart, and if it prices and sags, every CFO behind it suddenly discovers reasons to wait one more quarter. The second is allocation. SpaceX reportedly floated giving up to 30 percent of its shares to retail, three or four times the usual sliver, which is partly a Musk-flavored populist flex and partly a recognition that ten million Starlink customers are a captive retail bid. That normalizes a structure smaller issuers will copy, and a hot retail tape is exactly the environment in which a consumer-facing health company can punch above its institutional support.
The third channel is the one most people miss, which is that public-market confidence flows backward into private financings before it shows up in any listing. Biotech told that story in late 2025, when the S&P Biotech index ran up roughly 36 percent against the broad market’s 16 and the Nasdaq’s 20, and that rally showed up first as a record wave of follow-on equity raises rather than fresh IPOs. Crossover investors do not wait for the bell; they pre-position in the last private round, and a healthy IPO tape is what gives them permission to do it. So the practical signal for a healthcare operator is not the IPO calendar at all. It is whether the late-stage private rounds are clearing at flat-to-up valuations, because that is the leading indicator and the listing is the lagging one. Crunchbase’s Gené Teare had already flagged that early-2026 listings came in slower than expected precisely because the oxygen was being consumed by AI names that had not yet listed, which is a tidy summary of the whole dynamic: the giants suck in the air on the way up, then exhale it on the way out.
The healthcare lag, and why it always lags
Healthcare is congenitally late to every cycle, and it is worth being honest about why instead of pretending it is bad luck. Reimbursement is the original sin. A SaaS company books a contract and recognizes revenue; a care-delivery company books a contract and then spends eighteen months proving outcomes to a payer who would prefer to pay nothing. Layer on regulatory timelines, FDA clearances, the perennial overhang of drug pricing policy, Medicare Advantage rate uncertainty, and the threat of tariffs on imported devices, and you get an asset class that public investors approach with a fire extinguisher already in hand. The 2021 cohort did not help. A parade of digital health names went public on growth-at-all-costs math and pandemic tailwinds, the tailwinds reversed, and a depressing number of them ended up below cash or got swallowed in cleanup M&A. The hangover was extreme. Twenty digital health companies listed in 2021, then two in 2022, then a clean zero in 2023, which is not a slowdown so much as a sector deciding to play dead.
What replaced that mania is a framing the bankers have started calling Health Tech 2.0, and underneath the buzzword is a genuinely different standard. The new bar is profitability or a credible line to it, AI that is actually embedded in the workflow rather than bolted on for the deck, and clinical-grade evidence that earns a place in a regulated care pathway instead of a wellness app store. It is a less fun era and a more durable one. The companies that survived the freeze did it by getting unit economics that look like software and outcomes data that look like medicine, which is a hard combination to fake and an even harder one to fund through a drought. The ones that could not raise a big private round spent the lean years scaling back trials, chasing pharma partnerships, or quietly accepting an early acquisition, and that sorting of the haves from the have-nots is exactly the population that now lines up at the door when the door opens.
The class of 2025 and what Hinge and Omada actually proved
The drought broke in the spring of 2025, and it broke with two names that were almost suspiciously well-behaved. Hinge Health priced at $32 on May 22, jumped about 17 percent on debut, and crossed a $3 billion market cap, which put it somewhere around five and a half to six times forward sales, a premium to most listed digital health peers that the analysts grudgingly admitted was earned. The reason it was earned is in the metrics, and the metrics are the entire point of this section. Hinge posted second-quarter revenue around $139 million, up roughly 55 percent, with non-GAAP gross margins near 83 percent, net dollar retention around 117 percent, and positive free cash flow, and by late in the year the stock had appreciated something like 63 percent off the offer. That is not a digital health profile. That is a software profile wearing a lab coat, achieved by using sensors and computer vision to strip the human physical-therapist hours out of the cost line. Omada followed on June 6 at $19, raising about $158 million at roughly a $1.1 billion valuation, off 2024 revenue near $170 million that had grown 38 percent and a first quarter that grew 57. HeartFlow joined the parade mid-year, and the broader recent cohort (Waystar in 2024, Tempus AI, Caris Life Sciences, Kestra Medical) all carried the same fingerprint of real revenue and a path to the black.
Baird’s Sasha Kelemen put it well at HLTH, saying the two debuts essentially redefined the blueprint for what a public digital health company has to look like, and Silicon Valley Bank’s Megan Scheffel noted there is a real backlog of companies with the financial profile to follow if the market cooperates. The cautionary footnote, courtesy of some grumpy reporting out of Modern Healthcare, is that the IPOs did not fully convert the skeptics: plenty of digital health investors still quietly prefer a clean sale to a strategic or a private equity buyer over the quarterly indignities of public life. So the class of 2025 proved the window exists and proved what gets you through it, but it did not prove that the average venture-backed care company actually wants to walk through, which is a distinction that matters a great deal for predicting volume.
TechBio, or what happens when the AI money finally notices biology
The most interesting place the AI wave touches healthcare is not in another telehealth app, it is in drug discovery, where a relabeling exercise is quietly doing real work. Call it TechBio, the term Recursion has been pushing for years, the conceit being that these are AI and computing companies that happen to make drugs rather than biotechs that happen to use software. The capital backing the thesis got loud in early 2026. Isomorphic Labs, the DeepMind spinout, raised $2.1 billion in a Series B and signed a research collaboration with Johnson and Johnson where Isomorphic does the in-silico design and J and J does the wet work, which is the cleanest possible expression of the full-stack ambition. Xaira launched with a reported $1 billion in a single round and no prior history, the largest debut financing the space had seen, betting entirely on the founding team. On the public side, Eikon Therapeutics and Generate:Biomedicines both completed IPOs in February, raising roughly $381 million and $400 million respectively, and Alamar Biosciences priced an upsized proteomics offering in April near $220 million at $17 a share. Kailera, stacked with China-sourced obesity assets and a $400 million Series A behind it, became the name everyone watched as a bellwether for whether late-stage data plus differentiated science could still command a crowd.
The arbitrage here is almost too obvious to say, so it should be said. Public markets are paying a steep premium for anything that can credibly attach the letters A and I to its story, and a vertically integrated discovery platform with proprietary data, owned wet-lab infrastructure, and an actual clinical pipeline gets to wear that label honestly. That is a meaningfully better position than a pure software vendor selling shovels into the gold rush, because the platform owns the asset that eventually reads out. The danger is symmetric and unforgiving. The market has watched enough AI-discovery promises evaporate that it now wants the readout, not the press release, and the moment a flagship TechBio program misses a clinical endpoint, the whole cohort gets repriced as biotech, which is to say binary and brutal. The wave can lift these names higher and faster than ordinary biology deserves, right up until biology sends the bill.
Who gets to ride, and who gets left at the gate
The honest forecast is that the wave does not lift the sector, it lifts a list. On the digital health side, the names that keep coming up as 2026 and 2027 candidates are Sword Health, Quantum Health, Transcarent, Maven, and Thyme Care, and they all map onto the post-Hinge checklist: real recurring revenue, a credible profitability arc, AI doing something load-bearing rather than decorative, and clinical or financial outcomes a payer will actually pay for. The framing that gets a company through the door now is infrastructure, not engagement. The winners of this phase look like utilities, the trusted pipes and models and reimbursement rails that sit underneath the system and are annoying to rip out, which is exactly why the claims-processing, payer-contract-intelligence, value-based-enablement, and revenue-cycle layers tend to age better in public markets than the consumer-facing apps. A company that can describe itself as embedded plumbing with switching costs has a far easier roadshow than one that has to explain its monthly active users.
Who gets left at the gate is the more useful question. Anything still running on 2021 math gets nothing, because the buy-side learned that lesson with its own money. Anything that needs the market to take a flyer on a story that has not produced margins yet competes directly against the AI giants for attention and loses, because a generalist will always take the trillion-dollar narrative over the unproven one. And there is a quieter drag worth naming: Rock Health’s early-2026 read showed digital health funding concentrating into fewer, larger startups, which means the long tail is being starved at the private stage and will never reach an IPO conversation at all. M&A remains the gravitational default for everyone else, both because investors are tired of waiting and because a strategic buyer does not require a perfect tape. The wave, in other words, rewards the companies that least needed it, which is the most reliable rule in capital markets and the one founders most consistently refuse to believe applies to them.
The stuff nobody wants on the roadshow slide
The risks are not exotic, they are just inconvenient. The macro backdrop in the first half of 2026 included a war involving Iran, oil-price jitters, sticky inflation, and rate uncertainty, the exact cocktail that makes exit timing feel, in Rock Health’s phrasing, especially hard to predict. That matters disproportionately for healthcare because healthcare needs a calm tape to get its more complicated stories across, and a calm tape is precisely what three trillion-dollar offerings and a geopolitical flare-up do not produce. Then there is the circularity problem hanging over the AI complex, the uncomfortable pattern of model companies, chipmakers, and cloud providers funding one another in a loop that flatters everyone’s revenue until someone asks where the end customer is. That critique is going to feature in every AI prospectus risk section, and OpenAI’s own unit economics (losing more than a dollar for every dollar earned, burning roughly $27 billion in a year) are the kind of figures that look visionary in a bull market and like a smoking crater in a bear one.
The asymmetric risk for healthcare is the one to actually worry about, and it is a borrowed risk. If even one of the giants prices badly or breaks in the aftermarket, the halo inverts on contact, the window everyone was counting on slams, and the kidney-care platform that did everything right gets punished for the sins of a satellite company it has never met. Layer the sector’s own overhangs on top, the drug-pricing policy that never resolves, the Medicare Advantage rate whiplash, the device tariffs, the lockup expirations that turn insiders into sellers right when sentiment is fragile, and the picture is one of borrowed momentum that can be recalled without notice. Somebody at CNBC framed the whole 2026 setup as either the most consequential IPO cycle since the dot-com era or the most expensive lesson in narrative versus fundamentals the public markets have ever taught, and the irritating truth is that both can be right in sequence. The smart money in healthcare is not trying to call which one it is. It is watching the late-stage private rounds for the leading signal, keeping its prospectus warm, and quietly assuming the door stays open for about a quarter and a half, because that is roughly how long these things ever stay open before the rockets come back down and take the air with them
.o