SK Hynix’s Worst 7-Day Loss Since 2008: Bubble Or Bargain? - SOFTBANK GRP UNSP/ADR by Softbank Group Corp
SK Hynix (SKHY) and Samsung (SSNLF) shares fell sharply, with SK Hynix experiencing its worst 7-day loss since 2008. The selloff was driven by increased supply fears and higher inflation data. SoftBank (SFTBY) and other tech stocks also declined. Analysts note that historical data suggests such losses often precede further declines, but current valuations are low.
How this was made
The 30-second read
Why it matters
Higher-than-expected inflation raises odds of additional Fed hikes, which pressures valuation multiples and increases the cost of funding AI build-outs. The selloff is also tied to fears of increased DRAM supply after a new chip design, plus the idea that buyback support has ended.
Market read
Traders can use the inflation-driven rate and credit repricing narrative, plus DRAM supply fears, to manage semiconductor and AI-financing exposure into the next sessions.
What to watch
The piece emphasizes supply fears and rates, but does not quantify whether demand for AI memory is accelerating or whether supply additions are delayed, which could change the near-term trajectory.
Background
The article frames the move as SK Hynix’s worst seven-day loss since 2008, occurring alongside a broader semiconductor drawdown and a US inflation surprise.
Ticker impact
SK Hynix shares are described as plunging 11.06% in seven sessions, with trading halted after the selloff and supply fears.
Likely further downside pressure as the market digests higher-for-longer rate odds and potential DRAM supply overhang.
The article ties the move to inflation-driven rate expectations, higher credit insurance costs, and a new DRAM design raising supply fears, all of which typically pressure memory valuations during drawdowns.
Micron is cited as also on a seven-day losing streak, implying sector-wide repricing beyond just SK Hynix.
Likely negative bias as traders extend the memory selloff across names while awaiting evidence that earnings hold up.
The article does not provide a company-specific catalyst for Micron, only that it is also losing over the same window, so the linkage is indirect.
CoreWeave is mentioned via rising default-insurance costs, linking AI infrastructure financing stress to the chip selloff narrative.
Potential downside pressure on AI-adjacent risk sentiment, which can weigh on memory suppliers through demand expectations.
CoreWeave is referenced as part of credit-market stress, but the article does not report a CoreWeave-specific event or price move.
Oracle is referenced as an example where default-insurance costs jumped, reinforcing the broader credit-tightening backdrop.
No clear single-name directional call from this article alone.
Oracle is not the subject of a discrete news item here; it is used to illustrate credit-market moves.
Market effects
Memory-cycle and DRAM supply fears are reinforced by a new DRAM design and by valuation compression across the semiconductor complex.
South Korea and Japan are both described as in sharp decline, with the Korea Exchange halting trading, signaling broad regional risk-off.
US inflation surprise is linked to higher rate expectations and higher credit stress, which can transmit to global semis and AI infrastructure financing.
Counterpoint
The article argues memory stocks often look cheapest just before profits peak, so the selloff could be an opportunity if earnings hold and supply fears fade.
Key entities
- companySK Hynix Inc.
Seoul-listed memory supplier whose shares are described as plunging 11.06% and triggering a trading halt.
- companyMicron Technology Inc.
US-listed memory peer cited as also on a seven-day losing streak.
- companySoftBank Group Corp.
AI investment giant cited as dropping 9.3% during the same selloff.
- companyCoreWeave Inc.
AI infrastructure provider referenced via rising default-insurance costs, used to illustrate credit stress.
- companyOracle Corp.
Referenced as another example where default-insurance costs jumped, reinforcing the credit-tightening backdrop.
