Moineauworld

Like this:

SK hynix Rises After Scrapping Solidigm | Analysis by Brian Moineau
Discover why sk hynix solidigm ipo cancellation boosted shares and what funding plans mean for investors—read the full analysis now.
Brian Moineau
October 24, 2026
Business Technology

TL;DR

  • SK hynix shares rose about 4% on October 23, 2026 in Seoul after it scrapped plans for a U.S. IPO of Solidigm, its NAND/SSD unit—signaling no dilution while AI‑memory cash flows strengthen. [1][3][4][8]
  • Management shifted to fund Solidigm’s potential U.S. fab from internal cash, debt, and subsidies rather than selling 10%+ of the unit into a choppy 2026 semiconductor tape. [2][3][6]
  • The core story is which governments finance a first‑in‑years U.S. NAND fab and how Korea’s July 2026 “split listing” guardrails change capital raising across chaebol structures. [2][11][12][13]

What the source said

Barron’s reported SK hynix scrapped a planned U.S. IPO for Solidigm and that shares gained roughly 4% in Korea trading as investors welcomed reduced dilution. It tied the move to weak 2026 market conditions and said the company would seek alternative funding for capacity expansion while riding secular AI demand in memory and storage. It noted the October 23, 2026 pop in Seoul and a positive move in ADRs as relief that management would not carve out the fast‑growing unit at a discount. The piece framed internal and policy routes—including CHIPS Act incentives—as more likely paths than a U.S. listing in 2026. [1][3][4][8]

Why it matters

Shareholders of SK hynix keep 100% of Solidigm, a unit that press and sell‑side chatter valued at up to $150 billion if listed, and they still get a 40 trillion won buyback with full cancellation plus a target to return “over 50% of FCF” in 2025–2027. That combination preserves upside in enterprise SSDs while concentrating AI‑cycle cash at the parent. The October 2026 decision also avoids selling a crown jewel into a discount window. [2][4][14]

Policy makers in Seoul and Washington now hold the keys. Korea’s Financial Services Commission and KRX hardened “split/duplicate listing” rules in July 2026 to protect minority holders—even for overseas IPOs—raising execution risk for carve‑outs like Solidigm. The U.S. wants more domestic memory output and can swing the outcome via CHIPS Act grants and state‑level packages if a U.S. NAND fab proceeds. [11][12][6]

Original analysis

SK hynix stock jumps after scrapping U.S. IPO: what’s really priced in

Consensus view in October 2026: shelving Solidigm’s U.S. IPO handicaps growth and risks a U.S. policy miss. Contrarian read: it’s rational de‑risking under Korea’s July 2026 governance regime that preserves control of a high‑margin asset while signaling confidence that HBM‑driven cash can shoulder capex without selling equity at a 2026 discount. [12]

Evidence for the contrarian case:

  • Valuation math favored staying private. Media chatter put Solidigm’s potential raise at $10–$15 billion against talk of up to $150 billion valuation; selling ~10% into a volatile semiconductor tape risked a headline discount and new governance overhead. SK hynix finished up about 4.4% on October 23, 2026 in Seoul as dilution fears cleared. [3][4][8]
  • Regulators raised the bar. Korea’s FSC/KRX tightened “split listing” rules in July 2026, requiring explicit minority‑protection measures and, in cases, shareholder consent for subsidiary listings abroad, increasing timing and political risk. Why fight that while running a record buyback and harvesting AI‑cycle cash? [11][12][14]
  • HBM leadership throws off cash. SK hynix’s HBM capacity is tied to NVIDIA roadmaps through multiyear partnerships announced in 2026, supporting superior DRAM economics and funding optionality without ceding Solidigm economics to new minorities. [9][10]

Back‑of‑the‑envelope financing math:

  • Capex proxy: Micron’s Boise memory fab plan is $15 billion over roughly a decade, a reasonable U.S. front‑end benchmark; a Solidigm U.S. NAND fab likely sits in a $10–$20 billion band over several years. [6]
  • Currency math: SK hynix approved a 40 trillion won buyback/cancellation and guided to “over 50%” FCF returns for 2025–2027; using Reuters’ cited rate of ~1,326 KRW per USD, 40,000,000,000,000 ÷ 1,326 ≈ $30.2 billion—enough to cover a Micron‑scale fab before debt or subsidies. [14][3]
  • IPO trade‑off: Raising $10–$15 billion via IPO would fund part of a fab but permanently sell ~10% of Solidigm; if HBM cash and debt markets remain available in 2026–2027, internal funding keeps strategic freedom. [3][6]

Named‑stakeholder breakdown:

  • SK hynix: Retains 100% of enterprise SSD upside as AI training/inference shifts data centers to QLC‑rich, ultra‑high‑capacity roadmaps; avoids new subsidiary governance under July 2026 rules. [2][10][12]
  • Solidigm: Keeps a single‑owner sponsor through site‑selection talks for a U.S. fab, including state incentive negotiations in 2026–2027. [2][6][13]
  • NVIDIA: Benefits from a focused HBM partner funding next‑gen ramps and ecosystem R&D for 2026–2027 AI platforms. [9][10]
  • Micron and Kioxia/Western Digital: Lose a near‑term public valuation “comp” event in U.S. markets; competition shifts back to eSSD share, cost/bit, and onshoring narratives in 2026. [6][10]
  • Regulators (FSC/KRX; U.S. CHIPS): Korea validates its minority‑protection framework; the U.S. keeps a live target for a rare U.S. NAND fab, now hinging on incentives and export‑control trade‑offs rather than equity capital markets. [11][12][13]

Historical analogue: In 2012–2015, Samsung Electronics and SK hynix funded memory super‑cycle capex mostly from internal cash and structured financing, not equity carve‑outs, and the control premium paid off into the 2016 upturn. That playbook, adjusted for 2026 AI supply chains, points toward internal funding first and equity optionality later.

What others are missing

Media coverage fixates on the IPO headline while missing Solidigm’s eSSD unit economics and QLC strategy that turn bits into 2026 cash. TrendForce estimates enterprise SSD vendor revenue reached nearly $37.59 billion in 2Q26, and its trackers show eSSDs approaching 48% of NAND bit shipments as PCIe 5.0 ramps through U.S. hyperscalers. If eSSD penetration climbs into the mid‑50s by 2027, mix‑driven gross margins could rival early‑cycle HBM cash impact at SK hynix. Selling ~10% of Solidigm in 2026 would have clipped the very cash engine needed for a U.S. fab and for the 40 trillion won buyback. [7][10][14]

What to watch next

  1. By Q2 2027 (June 30, 2027), SK hynix will announce a specific U.S. state and site for a Solidigm front‑end NAND fab with a committed capex of at least $10 billion.
  2. By year‑end 2027 (December 31, 2027), enterprise SSDs will reach at least 55% of global NAND bit shipments in major trackers, up from 48% in 2Q26. [7]
  3. By H1 2027 (June 30, 2027), Korea’s FSC/KRX “split listing” regime will publicly scuttle or materially alter at least one planned subsidiary IPO by a listed parent due to minority‑protection requirements. [11][12]

My take

SK hynix chose substance over sizzle in October 2026 by protecting Solidigm just as AI storage demand steepens and Korea’s July 2026 rules make carve‑outs harder. I prefer internal cash, debt, and CHIPS/state incentives to fund a first‑in‑years U.S. NAND fab rather than selling a 10% slice at a discount. If management lands a competitive package and keeps HBM execution tight, the 40 trillion won buyback/cancellation and “over 50% of FCF” target keep compounding shareholder value. The stock’s 4% move says investors read it the same way. [2][6][11][12][14]

Sources

🔖Tags: AI Brian Moineau buyback capex capital raising CHIPS Act data center enterprise SSD free cash flow HBM IPO Korea FSC KRX memory Micron NAND Nvidia onshoring QLC semiconductor share repurchase SK Hynix Solidigm split listing SSD subsidies
Share on Facebook
Share on Twitter

Leave a Reply

Like this: