Major announcements this week from SK Hynix and Micron Technology have solidified memory chips as the hottest segment in the AI landscape.
On June 24th, after the U.S. market closed, Micron Technology reported quarterly earnings that significantly exceeded analyst forecasts, with both sales and profit guidance shattering expectations, sending its share price soaring nearly 16% in after-hours trading.
Micron Technology CEO Sanjay Mehrotra stated during the earnings call that the memory industry is undergoing a structural shift due to the proliferation of AI.
He further cautioned that supply shortages are expected to persist beyond 2027, with emerging applications like robotics anticipated to drive additional demand for memory and flash storage. This outlook has effectively alleviated market concerns about a potential slowdown in AI-related capital expenditure.
Simultaneously, SK Hynix disclosed its plan for a U.S. listing, aiming to raise approximately 454.5 trillion won (about $29.4 billion), positioning it to be among the largest initial public offerings in history.
Buoyed by this dual positive news, Asian markets reacted swiftly on Thursday, with shares of SK Hynix and Samsung Electronics both posting significant gains.
Soaring Share Prices Propel Memory Stocks to Lead Semiconductor Indices
The AI infrastructure build-out is propelling memory chip companies to the forefront of capital markets.
Micron Technology shares have surged approximately 267% year-to-date, making it one of the top performers in the Philadelphia Semiconductor Index this year and on track for its strongest annual gain since 2009.
SK Hynix shares have nearly quadrupled since the start of the year, Samsung Electronics has nearly doubled, and Kioxia has surged close to 800%, becoming Japan's highest-valued listed company.
In contrast, AI chip leader Nvidia has gained only about 6.7% this year. This disparity highlights a market rotation up the supply chain toward upstream memory components, where the beneficiaries of supply constraints are even outperforming the core chip design giants.
A Dramatic Industry Reversal: From Losses to AI Darling
Not long ago, the situation for these companies was starkly different.
In 2023, hit by declining demand and inventory overhang, Micron Technology reported a loss of nearly $6 billion for that fiscal year, with SK Hynix and Kioxia also mired in deep trouble.
SK Hynix has a more tumultuous history, born from a government-led rescue merger of LG Semicon and Hyundai Electronics, struggling for survival for years thereafter. Kioxia was spun off from scandal-plagued Toshiba, with its IPO plans repeatedly delayed due to weak investor demand.
The AI infrastructure wave, led by Nvidia, has completely rewritten this narrative. SK Hynix took an early lead in high-bandwidth memory (HBM), gaining an advantage due to its high compatibility with Nvidia's AI accelerators, with Micron Technology and Samsung racing to catch up.
Supply Bottlenecks Persist, Long-Term Agreements Enhance Visibility
The core of the current tight supply situation lies in constrained capacity expansion.
The previous industry downturn compressed long-term capital expenditure by manufacturers, while the mass production of HBM has further consumed capacity traditionally used for DRAM and NAND chips, putting pressure on computer, smartphone, and other device makers with tight supply and rising procurement costs.
Manish Bhatia, Executive Vice President of Global Operations at Micron Technology, remarked that in his 25 years in the memory industry, he has never seen a better time.
Facing persistent shortages, Micron Technology is leveraging the opportunity to sign more long-term agreements with customers, some extending up to five years, providing greater sales visibility for its capital expenditure planning. Mehrotra candidly stated that they currently see no sign of memory supply catching up to rising demand.
Sustainability of the Boom Cycle Remains the Central Question
The memory industry is notoriously known for its severe cyclical swings, with the historical cycle of capacity expansion and oversupply remaining a persistent concern for investors.
However, multiple factors support the longevity of the current cycle: the continuous expansion of AI computing demand, the accelerated deployment of new applications like robotics, and the more disciplined capital expenditure approach adopted by major players following the last loss-making cycle.
Mehrotra has explicitly stated that shortages are expected to extend beyond 2027.
The over $29 billion SK Hynix aims to raise through its U.S. listing will also provide firepower for its continued capacity expansion and technology R&D.
For investors, whether the memory chip sector can maintain this structural premium will remain a core theme throughout the broader AI investment narrative.
Comments