The problem? Expectations were even higher. The LSEG SmartEstimate consensus called for ₩64 trillion in operating profit, meaning SK Hynix missed by roughly ₩3.5 trillion . Revenue of ₩79.3 trillion also fell short of the ~₩84 trillion estimate
.
The market interpreted the miss as a signal that the AI memory cycle may have peaked — that even extraordinary growth could disappoint when inflated expectations meet decelerating momentum .
The KOSPI had already been under pressure. After hitting a record close of 9,114.55 on June 22, it had fallen more than 20% entering bear territory by early July . The SK Hynix earnings report turned a correction into a crisis.
The two-day crash. On July 28, the KOSPI plunged nearly 11%, triggering multiple trading halts as Samsung Electronics and SK Hynix each fell over 13% . The next day, the index dived as much as 12.6%, marking the worst two-day rout in history
. By the end of July 29, the KOSPI had dropped more than 40% from its June peak — a collapse that erased roughly $2 trillion in market value
.
SK Hynix stock. The company's U.S.-listed shares (SKHY) fell 8.98% to $130.17 in regular trading on July 28 and slipped further in after-hours trading . On the Korea Exchange, SK Hynix shares had already corrected ~33% from their June record high after an earlier profit warning on July 13
.
Samsung Electronics. Samsung shares fell over 9% on July 2 and again over 13% on July 28 as part of the contagion . The two chipmakers together account for more than half of the KOSPI's weighting, making the index especially vulnerable to their declines
.
The selloff was not contained to South Korea. As AI-demand fears spread, major semiconductor stocks declined worldwide:
Single-stock leveraged ETFs were introduced in South Korea on May 27, 2026, allowing retail investors to place leveraged bets on individual stocks like SK Hynix and Samsung .
The flows. Korean retail investors poured a net ₩14 trillion into these products, compared with roughly ₩2 trillion from foreign investors . These funds became a source of extreme volatility, accounting for 35.9% of total ETF trading on July 7 alone
.
The trigger. When SK Hynix shares corrected, these leveraged funds suffered catastrophic losses. Forced liquidations cascaded through the market — one single-stock futures contract flash-crashed 18-20%, triggering approximately $57 million in liquidations . The selling pressure amplified the broader rout.
The apology. On July 29, Finance Minister Koo Yun-cheol apologized before the National Assembly for rushing the introduction of single-stock leveraged ETFs "without sufficient deliberation," acknowledging the product had intensified the rout and torched retail investors . The Financial Services Commission had already admitted on June 22 that it acted too quickly in approving these funds
.
Regulatory response. South Korea temporarily banned new listings of single-stock leveraged ETFs tied to major technology firms and raised minimum deposit requirements for retail investors to 30 million won ($20,300) .
Capex increase. SK Hynix guided 2026 capital expenditures to the "high-40 trillion won" range, a substantial increase reflecting aggressive investment in HBM and advanced DRAM capacity, including accelerating the M15X facility . This signaled that even with record profits, the company must keep spending heavily to maintain its technological lead.
No shareholder return policy. Despite sitting on ₩88 trillion in net cash, SK Hynix did not announce a new shareholder return policy . Investors were disappointed that the company offered no clear plan for dividends or buybacks, which contradicted the broader "Value Up" program the Korean government has been pushing to improve corporate governance and investor returns
.
During the earnings call, SK Hynix management stated it has concluded long-term supply agreements covering high-bandwidth memory (HBM) for AI applications with around 10 key customers, providing some revenue visibility into 2027 and beyond . However, this reassurance failed to calm markets. The key question was not whether demand exists, but whether the pace of growth can justify current valuations
.
The rout unfolded against several unresolved macro and sector risks:
The crash was a stark reminder that in markets driven by AI euphoria, even record profits can feel like failure when expectations are set impossibly high.