History's largest IPO. The world's first fusion plant licences. A BCI enters US trials.

SpaceX listed June 12 on Nasdaq as SPCX at $135 per share, raising $75B through the sale of 555.56 million shares — the largest IPO in history, surpassing Saudi Aramco's 2019 record. The stock hit an intraday high of $176.52 before closing at $160.95 (+19.2%), lifting market cap to ~$2.1 trillion.

The S-1 identified a market opportunity exceeding $28 trillion across SpaceX's business lines, with 90% attributed to xAI alone — an AI company SpaceX acquired in February 2026 and whose valuation now anchors SpaceX's long-term growth thesis.

Under new Nasdaq rules effective May 2026, SPCX can join the Nasdaq-100 in as little as 15 trading days, which Bloomberg Intelligence estimates could pull approximately $600B in passive capital into the stock automatically.

The IPO creates the first public benchmark for the space economy: SpaceX's $2.1T debut valuation is now the reference price against which every private space company — Rocket Lab, Axiom, Impulse Space — is implicitly repriced. Investors in private space assets should reassess their portfolio marks.

Marc's analytical frame from the CNBC investor returns story: the IPO validates the deep-tech venture model at the highest scale ever achieved. Early backers who held through the full private cycle generated returns that rank among the largest in venture capital history — a signal with direct implications for how frontier tech venture will be capitalised going forward.

The lockup structure matters for timing: employees can sell 20% of shares after first quarterly results on September 2, 2026, with full lockup ending around 180 days. A wave of insider selling at that point could materially pressure the price.

The IPO provides the first public price discovery for Starlink, which generated $11.4B of SpaceX's $18.7B in 2025 revenue — the company's only profitable business unit. How the market prices Starlink's subscriber trajectory versus the xAI bet is the central valuation debate for the next several quarters.

SpaceX is GAAP loss-making: the company recorded $8.7B in net losses between January 2025 and March 2026. The $2.1T valuation requires conviction on xAI's commercial trajectory, Starship cadence, and Starlink growth — three assumptions that are structurally unproven at commercial scale.

Risk caveat: the staggered lockup and ongoing losses create a visible near-term price pressure calendar. The September 2 earnings date and early lockup release will be the first real test of whether the public market will hold the IPO price thesis at scale.

The Washington DOH granted Helion two world-first licences — a Radioactive Materials Licence and a Radioactive Air Emissions Licence — on June 16, enabling the company to begin construction of Orion's generator building in Malaga, WA and advance toward its 2028 delivery target.

Helion also confirmed a transmission interconnection agreement with Chelan County PUD — the first interconnection agreement ever granted for a fusion power plant — providing the grid connection required to deliver electricity to Microsoft under the sector's only live commercial PPA.

Fusion is regulated under the byproduct material framework, not as nuclear fission, per NRC rulemaking and the ADVANCE Act of 2024 — placing it alongside particle accelerators and hospitals rather than fission reactors, which represents a materially shorter and simpler regulatory path than most investors assume.

These are the first fusion plant operating licences issued anywhere in the world. No fusion developer has ever advanced this far in the regulatory process; Helion has now mapped and cleared the licensing path that every other fusion company must eventually follow.

The combination of licences, transmission interconnection, a 2028 Microsoft PPA, and $1.5B raised means Helion has more of the hard commercial prerequisites for fusion delivery in place simultaneously than any company in the sector's history — a qualitatively different position from a company that has merely raised capital.

The regulatory framework matters beyond Helion: the Washington DOH's engagement with Helion has produced the first worked example of fusion power plant licencing under US law, which will reduce regulatory uncertainty cost for every subsequent developer seeking state-level approvals.

The transmission interconnection agreement is the underappreciated asset in this announcement: grid connection before a plant is operational is the prerequisite step that most fusion companies have not yet begun. Chelan County PUD's agreement provides validated route-to-market infrastructure.

The licences validate Helion's site selection and safety programme as meeting real regulatory standards — not just investor-facing milestones. This is the type of third-party verification that meaningfully changes how institutional capital prices execution risk.

Risk caveat: no fusion company has produced net commercial electricity. Helion's 2028 date remains the most ambitious timeline in the sector, and the distance between first operating licences and first commercial power still requires engineering milestones that have not been cleared.

Paradromics and University of Michigan Health completed the first implant in the FDA-approved Connect-One EFSon June 17, performed by neurosurgeon Dr. Matthew Willsey — the first fully wireless intracortical BCI to enter a permanent human clinical study in the US.

The Connexus BCI uses a high-density microelectrode array in the brain, a chest transceiver, and wireless skin transmission — enabling fully implantable operation without any external connectors through the scalp, the design risk that has complicated Neuralink's early device iterations.

The six-year follow-up protocol across three sites — University of Michigan, UC Davis, and Massachusetts General Hospital — will generate the longest continuous intracortical BCI safety and performance dataset in clinical research, directly informing the FDA's path to a pivotal trial.

Three distinct intracortical BCI architectures are now simultaneously in human trials: Neuralink (wired percutaneous), Neuracle NEO (epidural), and Paradromics Connexus (fully wireless intracortical). The clinical data emerging from all three over the next 2–4 years will define which architecture the FDA ultimately favours for a cleared indication.

Paradromics targets speech restoration — the highest-value BCI indication after severe motor impairment from ALS, stroke, or injury. A successful pivotal trial would create an FDA clearance pathway that benefits the entire intracortical BCI category, not just Connexus.

The FDA IDE in November 2025 followed a temporary implant during epilepsy surgery in June 2025 that confirmed Connexus' safety and signal quality in a human brain. The regulatory path reflects a methodical de-risking strategy that institutional investors in the space will recognise as structurally sounder than Neuralink's faster-but-riskier approach.

Paradromics is privately held and has not disclosed its current valuation. Connect-One is the company's first pivotal clinical milestone — the next funding round will likely price off this study's early safety data, expected within 12–18 months.

The three-site study structure at U of Michigan, UC Davis, and Mass General simultaneously validates the device across multiple institutional neuroscience settings, accelerating data accumulation compared to a single-site study and reducing the risk of site-specific bias in FDA review.

Risk caveat: Connect-One is an early feasibility study, not a pivotal trial. FDA clearance requires a successful pivotal study that has not yet been initiated; commercial data will not be available until 2031–2032 at the earliest.

Frontier announced $915M in new Growth AMC pledges on June 17, bringing total commitments to $1.8B, with Anthropic and Salesforce joining founding members Stripe, Google, Shopify, and H&M Group.

The Growth AMC targets 10–15 focused 8-to-10-year offtake partnerships at companies where Frontier has high conviction of gigaton-scale potential — a deliberate shift from the original model's broad early-stage discovery approach toward concentrated long-term bets.

Frontier has so far contracted ~$700M across 53+ projects, removing 1.8M tonnes of carbon, with seven companies breaking ground on 1.4M tonnes of new annual removal capacity in 2025 and more than 50,000 tonnes forecast for delivery in 2026.

The shift to 8–10 year offtake agreements is the structural maturation signal the CDR sector needed: contracts of this length and scale function as infrastructure-grade revenue guarantees that unlock project finance, rather than voluntary market credits that do not. This is the mechanism that converts CDR from an experiment into a bankable asset class.

Anthropic's participation signals that AI companies are beginning to formally price their energy and emissions footprint — Anthropic has not previously published a sustainability report and has favoured an "all of the above" energy approach. The first AI lab commitment to structured carbon removal purchasing will put pressure on OpenAI, Microsoft, and others to follow.

The total carbon removal market has now reached approximately $12 billion in credit sales to date, with Microsoft accounting for roughly three-quarters of all engineered removal purchases. Frontier's Growth AMC is explicitly designed to diversify that demand base — which reduces Microsoft's outsized influence on pricing and project selection.

The 8–10 year offtake structure is the credit instrument that CDR project developers need to unlock debt financing. Each Frontier offtake agreement functions as a revenue guarantee that supports project finance at commercial scale — a structural shift from grant-dependent CDR to contracted infrastructure cash flows.

Frontier's requirement that every funded project demonstrate a clear line of sight to government subsidy programmesis a signal about where long-term CDR demand is expected to come from: public procurement and carbon pricing policy, not voluntary corporate markets alone.

Risk caveat: the Growth AMC's success depends on government support frameworks materialising at the scale required for gigaton CDR — a policy assumption that has not been validated in any major jurisdiction. The 8–10 year contracts are only as durable as the regulatory environments in which projects operate.