Every time semiconductor stocks surge, the same question resurfaces: why aren't Korean companies valued fairly? On July 10, 2026, SK hynix listed on NASDAQ, raising $26.5 billion (approximately ¥4.3 trillion) and setting the record for the largest-ever U.S. IPO by a foreign company. The stock opened 14% above its offering price, pushing the company's market capitalization to $1.27 trillion—but a stock price rally and a justified valuation are two different things. Just 11 days before the listing, SK hynix had announced its participation in a South Korean government-led $576 billion investment plan, and the company is also facing pressure from the U.S. to build new factories domestically. Behind this record-setting IPO lies a company caught between competing demands.
$26.5 Billion Raised, Overtaking Alibaba's 12-Year-Old Record
SK hynix listed on the NASDAQ market on July 10, 2026, under the temporary ticker SKHYV. The company issued 177.9 million ADS (American Depositary Shares, equivalent to 17.79 million ordinary shares at a 10:1 ratio) at $149 per share, raising a total of $26.5 billion (KRW 40.023 trillion). The stock opened at $170 in its trading debut, 14% above the offering price. Demand reportedly exceeded the offered share count by more than sevenfold.
This amount surpassed the $25 billion record set by Alibaba in 2014 by roughly $1.5 billion, making it the largest-ever U.S. IPO by a foreign company. Alibaba's record had stood unbroken for 12 years. The fact that a single semiconductor manufacturer rewrote the record for "largest U.S. IPO by a foreign company" at this scale is unprecedented.
Bloomberg reported the company's total proceeds at approximately ¥4.3 trillion. Notably, this $26.5 billion figure is roughly on par with SK hynix's net profit of KRW 40.34 trillion (approximately $26.6 billion; source: ajupress, medium confidence) for the first quarter of 2026. Revenue nearly tripled year-over-year, while operating profit more than quintupled, driven by surging demand for AI-oriented high-bandwidth memory (HBM).
Looking at the broader landscape of global IPO fundraising, SpaceX's June 2026 offering ($75 billion to $85.7 billion) and Saudi Aramco's 2019 IPO ($29.4 billion) both exceeded SK hynix's $26.5 billion, placing SK hynix around third overall when ranked globally by amount raised. Even so, when narrowed to the specific category of "IPOs by foreign companies on U.S. markets," SK hynix's record remains unbroken for now. The fact that a semiconductor manufacturer alone achieved a record of this magnitude underscores just how much capital is flowing into AI-related stocks.
The Weight of SK hynix's 56.4% HBM Share in NVIDIA's Supply Chain
High-bandwidth memory (HBM) is a type of semiconductor that stacks multiple DRAM chips vertically and places them close to the GPU, achieving data transfer speeds far exceeding those of conventional memory. For NVIDIA's AI-oriented GPUs, the bottleneck in processing speed often lies not in raw computing power but in the exchange of data with memory—meaning HBM supply capacity directly shapes overall GPU production plans. SK hynix has built its position as one of NVIDIA's key suppliers around this component.
According to research from IDC, SK hynix held a 56.4% share of the global HBM market by revenue as of the first quarter of 2026, ranking first worldwide (source: ajupress, medium confidence). This figure means that more than half of all HBM market revenue is concentrated in a single company: SK hynix. Even as individual GPU performance improves, if the primary supplier—SK hynix—cannot keep pace with HBM supply, AI server shipment schedules are directly affected.
The advanced HBM market is currently an oligopoly dominated by three companies: SK hynix, Samsung, and Micron. While China's CXMT has begun mass production of older-generation HBM2, no manufacturer outside these three currently has mass-production capability for the most advanced HBM used in NVIDIA's GPUs—meaning AI GPU production plans are directly tied to the capital expenditure decisions of these three companies. The capital raised through this IPO will directly translate into SK hynix's own capacity for next-generation HBM development and capital investment.
CEO Kwak Noh-Jung has explained that the NASDAQ listing is a strategic move to leverage the company's international profile and raise capital from investors worldwide. Capital raised through the stock market will serve as a resource to accelerate investment in next-generation HBM products. The dynamic where memory manufacturers are being rewarded by investors amid the AI boom differs fundamentally from the logic of traditional semiconductor cycles.
Even at a $1.27 Trillion Market Cap, the Sub-6x PER Wall Persists
Following the listing, SK hynix's market capitalization reached roughly $1.27 trillion, with the stock opening 14% above its offering price. On the surface, these numbers suggest the market has embraced SK hynix's valuation generously. But looking at the projected PER (price-to-earnings ratio, on a 12-month forward basis) reported by CNBC at the $149 IPO price, SK hynix stands at just 4.8x—well below competitor Micron's 6.6x and far below the semiconductor industry's median projected PER of 29.84x (source: CNBC, medium confidence). Even accounting for the stock's post-listing gains, this multiple rises only modestly and remains nowhere near enough to close the gap with the industry median.
The scale of this gap becomes clear through a simple calculation. Based on shares outstanding and the $149 IPO price, SK hynix's market capitalization comes to roughly $1.088 trillion. If we replace SK hynix's projected PER of 4.8x with the industry median of 29.84x, the multiple gap works out to roughly 6.2x. In other words, if investors were to value SK hynix at the same multiple as the industry median, its IPO-price-based market cap would balloon to nearly $6.77 trillion in simple terms—a figure so unrealistic that it only underscores the magnitude of the valuation gap.
CNBC has framed this NASDAQ listing as a litmus test for whether the "Korea discount" can finally be bridged. While the listing itself pushed the stock price higher, it cannot yet be said that investors have come to value SK hynix's profit growth at industry-standard multiples. The rise in the opening price demonstrated strong demand, but by the measure of PER, the valuation wall that has long dogged Korean companies remains firmly in place even after the listing.
It is also true that the memory industry as a whole is prone to supply-and-demand cycles and tends to be valued at structurally low multiples. Micron's own projected PER of 6.6x also falls below the industry median of 29.84x. Yet SK hynix's 4.8x is notably lower even compared to Micron. The fact that a Korean company in the same memory-specialist category is being valued even more conservatively suggests that the Korea discount, long flagged even before the listing, has simply carried over unchanged.
$576 Billion Pledged to Korea, Silence on New U.S. Factories
Just days before SK hynix's celebratory NASDAQ debut, U.S. Commerce Secretary Howard Lutnick was applying pressure in the exact opposite direction. On July 9, Lutnick attended a construction milestone event at Micron's factory in Clay, New York (celebrating the pouring of the first concrete; construction itself had broken ground in January 2026), and used the occasion to call on Samsung and SK hynix to build memory chip factories in the United States as well. Micron has laid out plans to expand its total U.S. investment—spanning New York, Idaho, and Virginia—to more than $250 billion by 2035, with the Clay facility positioned as its core site.
Lutnick stated, "Sanjay Mehrotra, the CEO of Micron, may not want this, but I want to bring competitors Samsung Electronics and SK hynix to the United States and have them build chip plants here," adding, "Micron is blazing the trail, and other rival companies will envy it and eventually follow." U.S. semiconductor policy is intensifying pressure to secure domestic production capacity for AI-oriented memory, and SK hynix's NASDAQ listing itself could become fodder for that pressure campaign.
But SK hynix and Samsung had already, on June 29—ahead of Lutnick's request—announced their participation in a South Korean government-led AI and semiconductor investment plan totaling $576 billion. Together, the two companies plan to invest KRW 800 trillion (roughly $518 billion to $520 billion) in domestic chip facilities within South Korea, to be executed in phases depending on demand trends and board approvals. In effect, even as the U.S. was demanding new factories, the two companies had just committed to a massive investment plan directed at South Korea.
SK hynix has, in fact, already begun construction on an HBM advanced packaging and R&D facility in West Lafayette, Indiana, with an investment of approximately $3.87 billion (also stated in official announcements as "approximately $4 billion"), breaking ground in the first half of 2026 with operations planned to begin in the latter half of 2028. However, this is not a front-end semiconductor fab that manufactures wafers, but rather a back-end packaging and R&D facility—meaning its investment character likely differs from the wafer fabrication plants that Lutnick appears to have had in mind.
Based on a review of reporting from TechCrunch, Korea JoongAng Daily, CNN, and other sources cited in this article (confidence: medium), there is no evidence that SK hynix or Samsung has formally announced a construction timeline or investment amount for a new memory chip factory of the kind Lutnick is demanding. While there have been welcoming remarks, no commitment involving specific figures or schedules has been reported.
According to SEC filings (Form F-1 registration statement), the intended use of the $26.5 billion raised is spelled out in far greater detail. The funds are earmarked for construction of Fab 1 in Yongin, South Korea (KRW 26.6 trillion in additional investment out of a total project cost of KRW 31 trillion, or approximately $17.46 billion); the P&T7 advanced packaging facility in Cheongju (KRW 18.9 trillion in additional investment out of a total project cost of KRW 19 trillion, or approximately $12.41 billion); and the acquisition of EUV lithography equipment (KRW 11.9 trillion, or approximately $7.8 billion). The combined total of these additional investments exceeds the amount raised, with the shortfall reportedly to be covered by operating cash flow and borrowing—but in any case, no allocation to U.S. factories appears in the stated use-of-proceeds list.
The already-underway West Lafayette facility (packaging and R&D) represents a separate investment from the funds raised in this offering and does not directly respond to Lutnick's demands. While the use of capital raised in the U.S. market has been clearly defined and directed toward South Korea, no financial commitment has yet been shown for factories within the United States. This asymmetry is precisely the issue SK hynix will next be pressed to answer.
Samsung and Micron Closing In, and the Shadow of a 2027 Cycle Turn
SK hynix's 56.4% HBM market share may appear commanding, but viewed in industry context, Samsung and Micron continue to close the gap. Morningstar has positioned SK hynix's record-setting fundraising as one of the most closely watched listings amid the current AI-related stock boom, while also flagging risks inherent to investment in this sector (source: Morningstar).
Morningstar points out that the broader expansion of the memory market—fueled by robust demand for AI servers—carries the risk of tipping into a cycle turn between 2027 and 2028, marked by rising supply and eroding pricing power. The capital SK hynix raises through this listing could accelerate a capital expenditure race among all three companies, including Samsung and Micron, and if the supply-demand balance tips, the very companies that benefited now could find themselves facing the backlash. Indeed, as of the period just before SK hynix's listing (per a Morningstar article dated July 8, 2026), shares of Micron, SanDisk, and Western Digital had fallen by double-digit percentages over the preceding weeks, while Samsung Electronics dropped 7% following its July 7 earnings announcement—down 18% from its recent high. Morningstar warns that if the rapid inflows from fast-money and leveraged ETFs were to reverse, it could further sour market sentiment (source: Morningstar).
Whether SK hynix can truly overcome the Korea discount will depend less on stock price movements than on investment decisions grounded in tangible substance. Only when the company formally announces a construction timeline and investment amount for the new factories Lutnick is demanding—and demonstrates to the market that this is compatible with its $576 billion investment pledge in South Korea—will the PER wall finally begin to give way. Until then, the record-breaking fundraising and market capitalization remain figures that merely mask the underlying valuation gap.