Tuesday, June 23, saw global semiconductor stocks caught off guard by events in South Korea. A Wall Street strategist labeled the sharp selloff a "chip-wreck."
The initial collapse occurred in what has been this year's top-performing major market: South Korea. The KOSPI index plunged 10% in a single session, triggering multiple circuit breakers, with shares of SK Hynix and Samsung Electronics each falling more than 10%.
A few market rumors ignited the storm. Korean media reported that NVIDIA was expected to cut production of its upcoming Rubin platform, and that SK Hynix was slowing expansion of its high-bandwidth memory (HBM4) production to shift toward cheaper conventional DRAM. Separately, Yonhap News Agency reported that multi-party lawmakers in South Korea were discussing taxing unrealized gains on assets like stocks and real estate, meaning taxes would be levied on paper profits even before assets are sold.
This "chip earthquake" quickly transmitted to U.S. markets.
The Philadelphia Semiconductor Index (SOX) fell 7.9% overnight, with all 30 of its component stocks declining. Micron Technology shares dropped 13%—prior to Tuesday, the stock had already surged over 300% this year, making it the SOX index's strongest performer. Micron, NVIDIA, and AMD together accounted for roughly 50% of the S&P 500's decline. The Nasdaq closed down 3.3%, the Dow fell just 0.1%, and the S&P 500 dropped 1.4%.
Jonathan Krinsky, Chief Market Technician at BTIG LLC, stated, "Regardless of any short-term bounce, we still see intermediate-term downside risk for tech/AI." He believes the semiconductor sector could fall another 10% to 15% and described Tuesday's action as a "Chip-Wreck."
However, in a note on June 24, Peter Callahan, a TMT sector expert at Goldman Sachs Global Banking & Markets, wrote a crucial observation: "Most conversations with investors today revolved around the theme of 'what are you seeing' rather than signs of a broader narrative shift." This framed the selloff's scope: the price action was ugly, but at least for that day, there was no widespread abandonment of the AI trade.
Thus, the issue is not as simple as a Korean rumor single-handedly derailing a global AI bull market. It more closely resembles a sector that was already overextended, crowded, and leveraged collectively de-risking upon encountering a catalyst. In the short term, it indeed has characteristics of a "technical correction"; from a medium-term perspective, the vulnerabilities within the AI trade have not disappeared.
This Was a Transmission, Not an Accident
South Korea's crash appeared sudden, but the underlying logic is not complex.
The news of SK Hynix slowing HBM4 expansion hammered its stock. This stock holds a weight in the Korean market similar to Apple's in the Nasdaq—its sheer size means trouble for the entire index. More critically, Korean retail investors heavily use leveraged ETFs to participate in AI/semiconductor trades. These products are forced to sell automatically during market declines to maintain leverage ratios, creating mechanical selling pressure.
The news was the spark, but the leveraged structure was the explosive. Concurrently, some market observers are asking: "Could Korean leveraged retail investors be the ones to end the U.S. tech bull market?"
This question is somewhat exaggerated, but it points to a real vulnerability: AI/semiconductor trades are highly concentrated, with global investor positioning strikingly similar. A selloff at any node can transmit along this chain.
According to Goldman Sachs' post-market data, both long and short sellers were net sellers on the day: long-only funds (LO) had a sell skew of -18%, while hedge funds (HF) also sold consistently throughout the day, with shorts accounting for 60% of sales volume (compared to a recent average of ~50%). Both types of institutions sold over $1 billion in notional exposure.
The worst-hit U.S. stocks were the "crowded longs," this year's most profitable names: Goldman's memory stock basket (GSTMTMEM) fell 10%, its AI semiconductor basket (GSCBSMHX) dropped 620 basis points, its AI stock basket (GSTMTAIP) fell 440 basis points, and its 12-month momentum basket (GSXHUHMOM) declined 420 basis points.
Market Perspective: A Shift in Narrative or Not?
Given such a steep decline, how is the market interpreting it? Based solely on the drop, Tuesday looked like a repricing of the AI trade. But judging by trading volume and capital flows, the conclusion is less definitive.
In his post-market note, Goldman's TMT trading desk expert Peter Callahan described the day's feel with the word "orderly"—despite the significant decline, overall Nasdaq volume was roughly in line with its 20-day average, and cash and volatility desks operated normally.
More importantly, he described the day's investor conversations: "Most conversations with investors today revolved around the theme of 'what are you seeing,' with no observed signs of a broader narrative shift, nor incremental inquiries into 'new ideas' or 'laggards.'"
In other words, no one was rotating into new positions or seeking new investment directions. They were simply checking in with each other.
Another Goldman Sachs market strategist, Chris Hussey, provided supporting data: of the 12 tech stocks that fell over 8% on the day, all but one were still up double-digits for the year, with most having already doubled or more. His assessment was:
"Today's selloff looks more like a 'froth-skimming' of a frenzied price advance than a fundamental reassessment of the AI infrastructure trade. Investors aren't dumping the index wholesale, but are re-evaluating: what price should be paid for stocks that have doubled in six months?"
Jack Janasiewicz, Portfolio Manager at Natixis Advisors, offered a similar view:
"This looks more like a technical selloff than anything else. Market breadth was decent after the open, despite all the big red numbers—a sign of a narrow selloff." He also cautioned, "When we see such massive crowding in beta and momentum, it can easily lead to an ugly deleveraging event."
Beyond the Technical Correction: Structural Vulnerabilities That Cannot Be Ignored
The term "technical correction" sounds comforting, but while it can explain a lot, it may also mask genuine risks. The day's price action did have technical features: the decline was concentrated in winners, trading volume wasn't completely out of control, and investor chatter didn't indicate an immediate end to the AI narrative. However, there is no clear wall between a technical correction and structural risk—the former, if severe enough, can fully evolve into the latter.
Several contextual numbers are worth considering together.
First, the rally was too rapid. The Nasdaq is up over 30% since late March. In June alone, the Philadelphia Semiconductor Index saw 8 out of 16 trading days with intraday swings exceeding ±5%—meaning half the month's sessions featured violent chip stock gyrations. Even after Tuesday's drop, the SOX index is still up about 5% for June, outperforming the Nasdaq and S&P 500 by roughly 8 percentage points. A pullback from these levels is justified by technical correction and also reflects fragility at elevated heights.
Second, positioning is too crowded, and a key "support" force is temporarily absent. Julian Emmanuel, Chief Equity and Quantitative Strategist at Evercore ISI, told Bloomberg Television: "People are looking for reasons to hedge while wanting to stay invested." This accurately captures the market's current contradictory mindset. Simultaneously, about 65% of public companies are currently in their buyback blackout periods. In past selloffs, corporate buybacks were a crucial support, but that card cannot be played this time.
Third, the macroeconomic backdrop is shifting. Expectations for Federal Reserve rate hikes are rapidly heating up—Bank of America anticipates three more hikes this year, and market pricing for a July hike has risen from near zero to about 50%. The valuation logic for high-growth tech stocks is built on low-interest-rate discounting. Once rates rise, the present value of future earnings naturally contracts, and the first to feel the impact are those stocks with high valuations propped up by expectations.
Michael O'Rourke, Chief Market Strategist at JonesTrading Institutional Services, wrote: "The hyperscale cloud companies are the new software stocks. This group is dragging down the 'Magnificent Seven' while being unable to escape their own malaise."
Torsten Slok, Chief Economist at Apollo, listed three core questions facing the market: What happens if AI companies start cutting compute budgets due to insufficient ROI? What are the implications for equities and credit markets if the Fed hikes in September and December? There are no simple answers, but the market is shifting from 'willing to ignore these risks' to starting to take them seriously.
The reason this technical correction deserves serious attention is not the magnitude of the decline itself, but that it is occurring when valuations, positioning, interest rates, and sentiment are all at extreme levels.
Looking Ahead: Historical Precedents and the Next Test
Historical data shows that sharp declines in the Korean market are often violent but brief. This is a "silver lining" bulls might cite.
But the context this time differs from past purely domestic Korean events: it touched a core nerve of the global AI trade—is memory chip demand truly as robust as expected? Has the frenzy for data center construction already borrowed from the future?
Some answers may come after Micron Technology reports earnings on Wednesday. Micron has been the strongest component of the Philadelphia Semiconductor Index this year, up over 300% before Tuesday. Its earnings report will be a true stress test.
Perhaps BTIG's Krinsky put it most directly: "Regardless of any short-term bounce, intermediate-term downside risk for semiconductors remains."
Comments