The 2026 IPO Boom Is Real. The Aftermarket Is the Test.

Global IPO proceeds more than tripled year-over-year to roughly $187–191 billion across 483–530 deals in the first half of 2026, already exceeding the full-year total for 2025^[1][2]. U.S. issuers alone raised approximately $251 billion across 86 offerings^[2]. Total equity capital markets activity reached $297.1 billion in proceeds^[2]. By any issuance metric, 2026 is a boom year.

Yet the year’s two largest debuts — both AI-linked — are trading below their offer prices. The aftermarkets tell a different story than the headline calendar. That divergence is the market-structure story of the moment.

The SpaceX Stress Test: Lockup Expiry as the Real IPO

SpaceX priced its IPO at $135 per share on June 12, raising approximately $75–86 billion in the largest offering on record^[3]. The stock popped 19% on its first day^[4]. Five weeks later it was below its IPO price^[4]. By August 1, the stock had lost more than $500 billion in market capitalization and was down over 50% from its intraday high^[5]. Short sellers had accumulated roughly $8.3 billion in paper profits^[5].

The proximate cause is not fundamentals — SpaceX reported a 92% year-over-year revenue jump in its first quarterly report, with management guiding to $100 billion in annualized revenue by year-end^[6]. The pressure is structural: a staggered lockup schedule that begins freeing shares just weeks after the IPO.

On August 6, the first lockup expired, releasing 911.5 million shares into the public float — roughly 43% more than the 638.9 million shares sold in the IPO itself^[7]. That single expiry more than doubled the company’s public float^[6]. The staggered schedule will free an additional 12.9 billion shares by mid-2027^[6]. CEO Elon Musk, who owns roughly 42% of the company, is barred from selling until June 2027 under a separate agreement^[6].

The stock actually rose about 6% on the day of the first unlock^[7], and by August 10 had recovered to its $135 IPO price^[3]. The initial selling pressure was lighter than some feared. But this is only the first tranche. The real question is not whether the lockup expiries will pressure the stock — the staggered schedule guarantees they will, episodically, through mid-2027. The question is whether early holders view each unlock as an exit or as a non-event.

I would put the probability that subsequent tranches produce smoother trading at roughly 55/45. The 55 case: the first expiry’s orderly reception signals committed long-term holders, and the hosted-compute revenue stream (SpaceX has deals with Google at $920 million per month and Anthropic for its Colossus 1 data center capacity^[5]) gives the story a fundamental floor. The 45 case: each successive tranche draws from earlier, lower-cost holders with progressively stronger incentives to diversify, and options implied volatility remains at levels one advisor called “sheer insanity”^[6], indicating the market is pricing genuine uncertainty about the path.

The AI IPO Pipeline: Anthropic and OpenAI in the Wings

Behind SpaceX, the two most anticipated offerings are both AI companies that have already filed confidentially with the SEC.

Anthropic filed its prospectus in June^[8]. CFO Krishna Rao is leading early investor meetings that remain high-level — covering Claude models, Claude Code, enterprise positioning, and management, but notably not discussing financials or valuation^[8]. The company’s run-rate revenue crossed $47 billion in May, up from approximately $10 billion in all of 2025^[8]. Its last private funding round valued it at $965 billion^[8], and on secondary markets the valuation has reportedly reached $1.5 trillion^[9]. Some investors are floating a potential IPO valuation of $2 trillion or more, though that figure originates from investor analysis, not from Anthropic itself^[8].

OpenAI has also confidentially filed but, as of late June, had not held pre-IPO investor meetings or outlined a public timeline^[8]. The sequencing matters: Anthropic appears to be moving first among the two, which could allow it to set the AI-IPO valuation benchmark — or to be the one that discovers whether public-market investors will pay private-market multiples.

The valuation gap between private rounds and secondary-market trading is a leading indicator worth tracking. When secondary-market prices run ahead of the last primary round by a wide margin — as Anthropic’s $1.5 trillion secondary valuation versus its $965 billion primary round suggests — it typically signals either that the last primary round was conservatively priced or that secondary-market liquidity is too thin to be price-dispositive. The IPO will resolve which interpretation is correct.

The S-1 Pipeline: Breadth Beyond the Megacaps

The IPO pipeline is not just two names. At least six companies filed S-1 registration statements with the SEC in a single day on August 12, spanning AI infrastructure, biopharma, fintech, and minerals^[9]. The prior week saw at least ten S-1 and S-1/A filings^[10]. The SEC’s EDGAR system has logged 1,288 S-1 filings since March 2026, according to one tracking service^[10].

On the international front, SK hynix is exploring a Nasdaq listing of its U.S. subsidiary Solidigm, with a potential pre-IPO fundraising of $3.5–7 billion and Morgan Stanley and Goldman Sachs under consideration as underwriters^[11]. Korean governance advocates have pushed back, arguing the listing would create a five-tier ownership structure that erodes SK hynix shareholder value^[11]. The Korean government has simultaneously begun implementing new guidelines to curb dual listings^[11]. The Solidigm case is worth watching as a test of whether cross-border subsidiary listings can proceed under tightening governance regimes.

Hong Kong delivered its strongest first-half IPO performance in five years, raising HK$209.9 billion across 85 new listings^[2]. India’s IPO calendar is similarly crowded, with 24+ companies targeting approximately Rs 35,000 crore in August alone^[1].

Selective Aftermarkets: Not Everything Pops

The aftermarket weakness is not uniform. It is selective, and the selection pattern is informative.

Consumer IPOs have struggled. Jersey Mike’s (JMKE) priced its $1 billion offering at $23 per share and closed its first day at $21.63 — roughly 6% below the offer price^[4]. Reformation (REF) priced at the low end of its range and closed essentially flat^[4]. These are recognizable brands with institutional backing, and they still failed to generate sustained first-day demand.

Meanwhile, some smaller, sector-specific listings have produced enormous first-day moves. Chengdu Ultra Pure Applied Materials gained 662% on its debut^[9]. CIQTEK rose 419%^[9]. These are Chinese listings on domestic exchanges, where retail-driven first-day dynamics differ structurally from U.S. and European markets, but they illustrate the bifurcation: the aftermarket is rewarding scarcity and thematic specificity, not brand familiarity or deal size.

The pattern is consistent with a market where capital is abundant but conviction is narrow. IPOs that fit the dominant AI-infrastructure narrative attract disproportionate demand. Everything else — consumer, financial, industrial — faces a colder reception.

The Buyback Counter-Narrative

While the IPO market is raising capital at a record pace, the buyback market is returning capital at an equally notable clip. Several large new authorizations landed in the first two weeks of August:

^[12]

The simultaneous acceleration of both issuance and buybacks is not contradictory — it is a sign of a market functioning at both ends. Companies that can access public capital are doing so; companies with excess cash are returning it. The net effect on aggregate share count and market liquidity is the under-discussed variable. If IPO issuance adds shares faster than buybacks remove them, the net supply of equity is expanding — a headwind for per-share metrics across the market.

What to Watch Next

1. SpaceX lockup tranche dates. The next staggered releases will arrive over the coming months through mid-2027. Each is a discrete liquidity event. Watch whether the stock holds above $135 on subsequent expiry days — that is the cleanest signal of whether early holders are exiting or holding.

2. Anthropic’s valuation discussion. The early meetings are not covering financials. When they do, the number that gets floated will anchor the AI-IPO valuation range. If it is near $1.5 trillion (the secondary-market level), the aftermarket risk is elevated. If it is closer to $965 billion (the last primary round), the deal is being priced with more cushion.

3. OpenAI’s timeline. OpenAI has filed but not yet begun investor meetings. The gap between Anthropic’s and OpenAI’s public debuts will determine whether the second deal benefits from or is punished by the first deal’s aftermarket performance.

4. Net equity supply. Track the aggregate dollar value of IPO issuance versus aggregate buyback execution. If issuance is running ahead of buybacks by a wide margin, the market is creating shares faster than it is retiring them — a structural supply pressure that does not show up in index-level analysis.

5. The Solidigm listing decision. SK hynix’s September 4 update could resolve whether the Nasdaq listing proceeds. A yes would establish a template for cross-border subsidiary listings under heightened governance scrutiny. A no would signal that political constraints are binding on corporate-structure decisions.

6. Consumer IPO reception. If the next batch of non-AI IPOs (retail, financial, industrial) continues to price below or at the low end and trade flat, the market is telling issuers that the window is open only for specific narratives. That narrowing would eventually constrain the pipeline itself.

FN2 Research provides market commentary and analysis for educational purposes. Nothing in this article constitutes investment advice or a recommendation to buy, sell, or hold any security.