DongA Ilbo: SK Hynix gained ~$15B on its Kioxia stake earlier in 2026, and plans to return ~$12B to shareholders via a special dividend
SK Hynix Inc.'s shares bucked a broad selloff in AI-related stocks thanks to the prospect of a large dividend payout related …
Context & Ripple Effects
SK Hynix entered the October selloff after a run of record financial results tied to AI HBM demand, including record fourth-quarter revenue and operating profit in early 2026. Management had already made shareholder returns a central response to concerns over the durability of AI spending, with a planned $29 billion treasury-share buyback and cancellation.
The reported Kioxia proceeds add a separate source of distributable capital: an investment gain rather than operating cash flow. That distinction matters as SK Hynix weighs returning cash to investors while AI-linked memory demand and chip valuations face sharper scrutiny.
First-order effects
- The reported roughly $12 billion special dividend would transfer most of SK Hynix's reported gain from reducing its Kioxia holding directly to shareholders.
- SK Hynix would pair a one-off cash distribution with its planned buyback, broadening its shareholder-return toolkit during a period of pressure on AI-related chip stocks.
Second-order effects
- Investors would have a clearer basis to separate SK Hynix's operating performance from the realized value of its Kioxia investment when assessing the company’s capital returns.
- Kioxia’s reduced role in SK Hynix’s investment portfolio makes the latter’s AI-memory earnings and capital-allocation decisions more central to its equity case.
Third-order effects
- If large AI-memory producers continue to convert investment gains and peak-cycle cash generation into distributions, capital returns may become a more prominent counterweight to concerns about the durability of AI infrastructure spending.
- The pattern points to a memory-industry divide between companies able to fund both aggressive capacity needs and shareholder payouts, and those without comparable AI-driven profitability.
The trend: AI-memory leaders are using exceptional cycle profits and portfolio gains to reinforce shareholder returns as investors test the staying power of AI infrastructure spending.