Samsung and SK hynix’s -9% to -11% memory selloff is less “AI demand” and more “who pays for the capex”—Micron becomes the shock absorber

In a capital-intensive industry, financing fears act like a multiplier on uncertainty.

Even if end-demand remains strong, buyers (hyperscalers and integrators) may delay or re-schedule capex tranches when credit conditions or internal ROI gates get tighter. That can hit memory companies through two linked mechanisms: (1) price expectations soften because suppliers accelerate output while the buyer’s timing becomes less certain; and (2) the market discounts the value of future capacity because payback timelines stretch.

Micron Technology is not “HBM-first” the way SK hynix is, but it sits in the same AI-memory economics: pricing for DRAM/HBM-adjacent memory and the speed at which suppliers can flex output.

In the most recent full-year data available through the data tools, Micron shows a cycle-like cash conversion pattern: FY 2025 had $17.53B operating cash flow and $1.67B free cash flow after heavy capex ($15.86B property/plant/equipment investment in the same FY). That profile makes Micron’s equity re-rate sensitive to what happens next to memory pricing and capex pacing—exactly the transmission channel implied by the Reuters framing.

Reuters also connects the selloff to intensifying competition from China. That matters because China entrants can sustain price pressure even when global demand is fine—so the market can’t assume “strong demand will lift all boats.”

One concrete market datapoint from the same supply ecosystem is CXMT’s Shanghai debut: Reuters reported the IPO raising 57.92 billion yuan (about $8.6B)—evidence of continued funding behind memory ambitions. (Even though this doesn’t quantify “HBM vs DRAM share” in the Reuters snippet we accessed, it does show capital is still flowing into the China memory buildout.)

Korea’s equity plumbing creates mechanical correlation. When two dominant memory names (Samsung and SK hynix) drop sharply on a shared narrative, funds often hedge or de-risk across the same index exposure. Reuters’ framing of the selloff as both “financing risk” and “China competition” increases the likelihood the selloff is modeled as macro-repricing rather than earnings-only.

That’s why the US readthrough matters: when the market believes the next quarter’s memory pricing is less dependable, it tends to re-price the global AI-memory complex—not just the Korean equities.