SK hynix Tests the Record IPO Window With a $28 Billion Nasdaq Listing

The U.S. IPO market just posted the largest first half in history, and the second half opens with a deal that will tell us whether the window stays open on its own weight — or only when a generational listing props it up.

Through June 26, 2026, U.S. issuers raised $251 billion in IPO proceeds, surpassing the prior midyear record set during the 2021 listing boom, according to Renaissance Capital data reported by Reuters.{{cite:fcde628e3f6a}} Two transactions — SpaceX’s $86.2 billion IPO and Alphabet’s $85 billion equity raise to fund its AI buildout — account for roughly 68% of that total. Together, they are larger than all U.S. IPO proceeds from 2023, 2024, and 2025 combined.{{cite:fcde628e3f6a}}

The headline is the number. The more interesting signal is what comes next.

SK hynix: $28.1 billion, pricing this week

SK hynix, the South Korean DRAM and flash memory giant currently listed on the KRX KOSPI under symbol “000660,” announced terms on July 6 for a U.S. IPO of 177.9 million American Depositary Shares at $158.14, the as-converted July 3 close of its Korea-listed shares.{{cite:a6626b900819}} The deal would raise $28.1 billion and value the company at approximately $1.2 trillion market capitalization.{{cite:a6626b900819}}

Cornerstone investors Baillie Gifford, Coatue Management, and Situational Awareness Partners have indicated on $7 billion of the offering — 24.9% of the deal — providing a substantial demand floor.{{cite:a6626b900819}} BofA Securities, Citi, Goldman Sachs, and J.P. Morgan anchor a syndicate of 15 bookrunners. The company plans to list on the Nasdaq under the symbol SKHY, with pricing expected the week of July 6, 2026.{{cite:a6626b900819}}

SK hynix booked $84.9 billion in revenue for the 12 months ended March 31, 2026, and ranked first or second globally in DRAM, high-bandwidth memory (HBM), and NAND flash by revenue in Q1 2026.{{cite:a6626b900819}} The listing is explicitly framed as a play for AI-investor access: the company is the dominant supplier of HBM chips used in AI accelerator packages, and the offering rides the same demand current that powered Cerebras Systems’ 89% first-day pop in May and SpaceX’s oversubscribed book in June.{{cite:fcde628e3f6a}}

This is the first IPO since SpaceX that is large enough to stress-test the market’s absorption capacity on its own. If the book is oversubscribed and the aftermarket holds, it validates the thesis that H1’s record was driven by genuine demand for AI-infrastructure exposure — not just one-off euphoria around a single generational listing. If it struggles, the concentration risk that Goldman Sachs flagged becomes harder to dismiss.

SpaceX: the cautionary tale underneath the record

SpaceX priced its IPO on June 12 at $135, opened at $150, and closed its first session at $161 — a 19% first-day gain that made it the largest IPO in history, eclipsing Saudi Aramco’s 2019 record by more than three times.{{cite:fcde628e3f6a}}

The follow-through has been rougher. Across a four-session stretch in late June, SpaceX shed roughly $400 billion in market capitalization, including a single-day decline of about 16%.{{cite:fcde628e3f6a}} The stock remains above its IPO price but has been volatile, with analyst coverage initiated at mostly bullish price targets even as the shares swung.{{cite:8fc9c9c2a887}}

The next structural test is lockup expiration. Multiple lockup-period expiration dates are approaching, after which insiders will be free to sell restricted shares — a flow event that could inject additional volatility into a stock that has already shown it can move 16% in a session.{{cite:8fc9c9c2a887}} The locked-up share overhang means the freely tradable float that supported the debut run is a fraction of the total share count; when those restrictions lift, the supply-demand balance that produced the first-day pop will reset.

This matters beyond SpaceX. Lockup mechanics are a feature of every 2026 debutant. If the pattern holds — strong bookbuild, healthy first-day pop, then a volatility air pocket as lockups approach — the H2 calendar will need to price that risk into new deals, potentially compressing valuations for issuers coming after the mega-cap wave.

The structural backdrop: the market itself has changed

Citadel Securities’ Scott Rubner published his 1H 2026 Market Structure & Flows review on June 30, 2026, with a thesis that cuts deeper than the IPO numbers: “the defining story of 2026 has not been a single macro event, it has been the structural transformation of equity markets.”{{cite:ee5e6bf23ec1}} Concentration, passive flow mechanics, and the behavior of new-issue supply are reshaping how liquidity forms and how price discovery works — not just how much capital is raised.

Liquidnet’s Q2 2026 U.S. Liquidity Landscape report echoes this: trading has shifted back toward lit venues, off-exchange behavior is evolving, and macro uncertainty is making execution conditions challenging for institutional investors even as the market continues to climb.{{cite:ee5e6bf23ec1}} The backdrop for new issuance is constructive on the surface — index gains, narrow credit spreads, steady monetary policy — but the plumbing underneath is in flux.

This is the tension that makes the H2 outlook genuinely uncertain. The H1 record was produced in a market where demand for AI-infrastructure exposure overwhelmed everything else. The structural conditions that supported that demand — concentrated passive inflows, a four-year backlog of postponed listings finally clearing, and a multiple-friendly rate environment — are either intact or shifting. Which direction they shift determines whether 2026 sets a full-year record or sees H2 cool sharply.

The July calendar: what is actually pricing

The Renaissance Capital IPO calendar for the week of July 6, 2026, shows three deals pricing alongside SK hynix:{{cite:da8d8e44eb0a}}

After this week, the pipeline includes Csquare (CSQR, $1.25B, Morgan Stanley/TD Securities), Standard Nuclear (STDN, $356M, BofA/Goldman), and MetaOptics (MOT, $18M, Roth Capital).{{cite:da8d8e44eb0a}} The Csquare deal — a $1.25 billion offering led by Morgan Stanley and TD Securities — is the first non-semiconductor mega-deal of the quarter and will be a useful breadth indicator: if a deal outside the AI-infrastructure narrative clears $1 billion with a healthy book, the market’s appetite is broadening. If it struggles, the concentration thesis hardens.

H1 2026 in context

Sources: Reuters/Yahoo Finance, June 27, 2026; Renaissance Capital 2Q 2026 U.S. IPO Market Review.{{cite:fcde628e3f6a}}

The composition data tells a more nuanced story than the aggregate. Eleven separate IPOs cleared the $1 billion threshold in H1.{{cite:fcde628e3f6a}} The 2026 debutants are up a weighted-average 16% from their offer prices, and the average IPO valuation is roughly three times last year’s average and ten times the 2022 average.{{cite:fcde628e3f6a}} That is not a “SpaceX plus everyone else” tape — but the valuation stretch is the part of the picture that should temper any straight-line extrapolation.

Three risks to hold in view

Concentration. Two deals drove ~68% of H1 proceeds.{{cite:fcde628e3f6a}} Strip them out and the market is still record-paced, but the breadth looks thinner. The H2 print will show whether the pipeline can sustain volume without a generational anchor.

Valuation stretch. Average IPO valuations are ~10x their 2022 level.{{cite:fcde628e3f6a}} A single sentiment reset — an AI capex disappointment, a credit event, or a renewed inflation surprise — can re-rate the cohort fast. The SK hynix pricing at a $1.2 trillion market cap{{cite:a6626b900819}} is the kind of number that looks either reasonable or extreme depending on whether HBM demand accelerates or plateaus.

Calendar risk. Morgan Stanley flagged that Q3 deals are likely to be front-loaded ahead of midterm-election volatility, meaning supply may compress into July and August.{{cite:fcde628e3f6a}} The H2 print will tell whether 2026 sets a full-year record or sees H2 cool sharply.

What to watch next

• SK hynix pricing and first-session performance (week of July 7). The bookbuild oversubscription ratio and first-day move are the cleanest signal of whether AI-infrastructure demand extends beyond SpaceX. A healthy pop with sustained volume validates the thesis; a flat or negative debut raises the concentration concern.
• SpaceX lockup expiration dates. When insider selling restrictions lift, the freely tradable float expands and the supply-demand balance that supported the debut run resets.{{cite:8fc9c9c2a887}} The volatility pattern around those dates will be a template for every 2026 debutant facing the same mechanics.
• Csquare (CSQR) pricing. The first non-AI mega-deal of the quarter. If it clears $1 billion with a strong book, the market is broadening. If it reprices or withdraws, the concentration thesis hardens.
• Renaissance IPO ETF (NYSE: IPO). The cleanest market proxy for newly listed share performance. When broken IPOs spike, sentiment turns and the new-issue window closes.{{cite:fcde628e3f6a}}
• PE-backed S-1 filing pace. JPMorgan identifies buyout-backed exits as the dominant H2 supply source.{{cite:fcde628e3f6a}} The pace of new filings from PE sponsors will pre-signal whether the calendar can sustain Q2’s volume.

The base case — 60/40 — is that the window stays open through Q3. The $251 billion H1 print, the SK hynix cornerstone anchor, and the structural demand for AI-infrastructure exposure all point to a market that can absorb $28 billion of new paper without breaking. The 40% case is that SpaceX’s lockup volatility, the valuation stretch, and a front-loaded Q3 calendar compress the window before the midterm election cycle fully engages. SK hynix prices this week. The answer starts there.