The most consequential move for SK Hynix this week did not happen on the Seoul exchange. It happened in the corridors of corporate strategy, where SK Group chairman Chey Tae-won floated the prospect of a joint semiconductor venture with Japan’s Kioxia — spanning production, research, and supply-chain integration. No contracts have been signed, but the public signal from the group’s top executive marks a notable shift in how the memory giant intends to navigate an increasingly crowded and capital-intensive landscape.
Investors, meanwhile, had to digest a more immediate set of pressures. US Commerce Secretary Lutnick warned midweek that targeted semiconductor tariffs could follow unless SK Hynix builds out meaningful US manufacturing capacity. That threat lands squarely on a company already committing billions to American soil, including a planned HBM4E packaging facility in Indiana that is not slated to reach mass production until the third quarter of 2029.
The stock closed Friday at 1,647,000 won, up 2.1 percent on the day — or 3.2 percent by the secondary article’s measure of the recent recovery. The weekly picture remains slightly negative at minus 0.4 percent, and the 30-day return sits at minus 1.3 percent. More tellingly, the shares trade roughly 11 percent below their 50-day moving average of 1,843,245 won, and remain about 45 percent off their 52-week high. Yet the year-to-date gain still stands at a formidable 154 percent.
Washington’s Leverage and the Indiana Timeline
Lutnick’s tariff warning cuts to the heart of SK Hynix’s expansion calculus. Alongside Samsung, the company has set in motion Korean megaprojects valued at 800 trillion won, and any new duties on finished products could unsettle those plans. Rival Micron has positioned itself as comparatively insulated, pointing to a $250 billion US investment and pulling forward the production start of its Idaho fab to mid-2027.
The Indiana facility itself tells a more complicated story about execution risk. The groundbreaking took place in late August, but the cleanroom is not expected to open until October 2028, with series production of the next HBM generation following in the second half of 2029 — roughly a year behind the original schedule. The delay is a reminder that even the dominant player in high-bandwidth memory cannot compress construction timelines or regulatory approvals to match the pace of AI-driven demand.
There is also the question of supply-chain dependence. SK Hynix currently relies on TSMC for 12-nanometer base-chip fabrication for HBM4E and is exploring Intel as a second foundry partner. Samsung, by contrast, handles those components in-house at 4-nanometer. Reducing reliance on a single supplier has become a strategic priority, though no definitive move has been announced.
Should investors sell immediately? Or is it worth buying SK Hynix?
CXMT’s Ascent and the DRAM Chessboard
Geopolitical friction is only half the story. Chinese rival CXMT has been steadily eating into SK Hynix’s DRAM territory, lifting its global share to 10 percent in the second quarter of 2026 from just 4 percent a year earlier. Over the same stretch, SK Hynix slipped to 25 percent. CXMT has also managed to bring LPDDR6 chips to mass production ahead of both SK Hynix and Samsung, and is targeting an entry into HBM3E by 2027.
The numbers behind that push are substantial: CXMT reported first-half 2026 revenue of 150.31 billion yuan and net profit of 77.60 billion yuan. The company has also filed a lawsuit challenging its placement on the US sanctions list — a legal front that could have broader implications for how Washington’s export controls are enforced.
The NAND Question and Solidigm’s Unresolved Future
Chey’s Kioxia overture speaks to a quieter but persistent problem: SK Hynix’s NAND business, run through subsidiary Solidigm, has struggled for strategic clarity while the market’s attention — and margins — flowed to HBM. A partnership with Kioxia, one of the few remaining independent NAND makers, would recast the competitive dynamic from rivalry toward consolidation in a segment that has drawn far less investor enthusiasm.
Solidigm itself remains in limbo. Reports of a 5 trillion won pre-IPO fundraising round have not been confirmed; the company says only that it is reviewing measures to strengthen competitiveness. That ambiguity suggests a decision may be approaching, but its form is far from settled. Investors hoping for rapid resolution are likely to be disappointed.
Analyst Conviction and the AI Catalyst
Despite the layered uncertainties, sell-side sentiment has held up. Nomura reaffirmed a buy rating with a price target of 4.7 million won, while Kiwoom sits at a more conservative 2.1 million won. The near-term catalyst came from OpenAI’s unveiling of GPT-6 Astra on September 3: SK Hynix’s US-listed depositary receipts jumped 8.14 percent within two trading sessions, outpacing the Philadelphia Semiconductor Index’s 3.37 percent gain.
Whether that AI-driven momentum can offset the combined weight of tariff risk, CXMT’s advances, and the unresolved NAND strategy is the question that will define the coming weeks. US inflation data due September 10 and 11 could move the needle on interest-rate expectations, with knock-on effects for growth-oriented names like SK Hynix. For now, the equity remains a bet on execution across multiple fronts — not merely a play on demand.
Ad
SK Hynix Stock: Buy or Sell?! New SK Hynix Analysis from September 6 delivers the answer:
The latest SK Hynix figures speak for themselves: Urgent action needed for SK Hynix investors. Is it worth buying or should you sell? Find out what to do now in the current free analysis from September 6.
SK Hynix: Buy or sell? Read more here...








