South Korean stocks were hit by two straight circuit breakers on July 28 and 29 in Seoul, an event the article describes as unprecedented in the country’s market history. The Kospi dropped 10.84% on the first day and another 5.98% on the second. SK Hynix, the index heavyweight, lost about 23% across the two sessions. At the same time, the Nasdaq sold off, semiconductor stocks fell globally, and leveraged exchange-traded funds were caught in the same downdraft.
After those two days, the Kospi’s drawdown from its June high had widened to 40%, and July was on track to become the worst month on record for the index. The piece frames the move not as a stock-specific shock but as a global forced deleveraging episode. In its telling, equities, not crypto, were the assets trading with the kind of violence usually associated with the digital-asset market.
Spot shares and leveraged products both took heavy damage
The article first points to the cash market. SK Hynix reported second-quarter operating profit of KRW 60.54 trillion, a record high. But the figure missed the LSEG estimate of KRW 64.22 trillion, and the stock was sold aggressively, ending July 29 at KRW 1.401 million.
In the author’s reading, good news that fails to lift a stock becomes bad news. The company, which had recently listed on Nasdaq, also fell below its $149 offering price.
Derivatives tied to the name suffered an even steeper collapse. CSOP’s 2x long SK Hynix ETF (07709.HK) fell from HK$193.65 on June 25 to HK$32.7 on July 29, a decline of 83%. At its peak, the product had more than HK$1.3 trillion in assets and was described in the article as the world’s largest single-stock leveraged ETF. More than HK$1 trillion in market value disappeared within a month.
The issuer responded by changing product rules. Starting Aug. 3, 12 single-stock leveraged products under its umbrella will shift from a fixed 2x structure to a flexible leverage model with a minimum of 1.1x, with the exact ratio set daily by the fund manager. Korean regulators are also preparing to limit retail purchases of leveraged ETFs.
Bitcoin stood out for a different reason. While it is usually treated as the high-volatility asset in the room, it rebounded from about $57,800 on July 1 to around $66,300, gaining nearly 15%. The article describes the contrast bluntly: stocks traded like crypto, while Bitcoin sat through the storm.
What triggered the selloff
From the June 22 high, the S&P 500 was down only 2.1%, and the Nasdaq had lost 6.6%. The Philadelphia Semiconductor Index, by comparison, had plunged 28.6%. The piece says that was not broad panic across all assets. It was a concentrated unwind in the market’s most crowded long positions.
It identifies two immediate catalysts. One was SK Hynix’s earnings report, which set a record but still fell short of expectations. The other was what the article calls the China variable: CXMT completed Asia’s largest IPO of 2026, with the proceeds earmarked for DRAM capacity expansion, creating a fresh challenge to the AI memory scarcity narrative.
Japan added pressure in the background. The Bank of Japan raised rates to 0.75% in December 2025, the highest level in 30 years, while the 10-year Japanese government bond yield climbed to around 2.9% in July, its highest since 1997. The market estimates that yen carry positions worth between $300 billion and $500 billion remain a source of stress for global risk assets. According to UBS, only about half of that carry unwind has happened so far.
Technology investor Dan Niles said this was not the collapse of the AI thesis itself, but a short-term bottom created by forced liquidations from retail traders and hedge funds. Prime brokers, he said, are accelerating risk cleanup to avoid a repeat of an Archegos-style blowup. He also argued that this was only a speed bump within an AI supercycle: the top 1% of companies are being cautious with computing power, while the remaining 99% are still spending. In that view, the industry logic is intact and leverage is what broke.
Bitcoin did not really receive the money leaving equities
The article’s answer is direct: no, the capital exiting stocks did not flow into Bitcoin in any meaningful way. Bitcoin’s relative resilience came because it had already been hit earlier.
From May 15 to June 3, U.S. spot Bitcoin ETFs recorded net outflows for 13 straight trading days, totaling about $4.4 billion, the longest such streak on record. Over the same period, Bitcoin fell from around $80,000 to $63,000, a drop of about 21%.
For June as a whole, net outflows reached about $4.5 billion, the worst month since spot Bitcoin ETFs launched. Nearly 80% of those exits came from BlackRock’s IBIT alone. The article argues that much of the weak positioning had already been flushed in June, leaving Bitcoin with less downside by the time technology stocks were being repriced in July.
There was a rebound in ETF demand later in the month. From July 14 to July 22, spot Bitcoin ETFs posted net inflows for seven consecutive trading days, totaling about $981 million, the longest and largest inflow streak of 2026. IBIT again led the move. But compared with the losses in May and June, the article says, that recovery was still modest. Analysts cited in the piece estimate that filling the hole would require several months of steady buying.
Defensive capital chose gold, not Bitcoin
The article says the real safe-haven bid went to gold. By the end of July, gold had risen to $4,086 an ounce, up more than 20% year over year.
Data from CryptoQuant showed Bitcoin’s 30-day correlation with gold falling to -0.88 at one point, the lowest reading since the deepest stage of the 2022 bear market. On that basis, the piece argues that the “digital gold” narrative did not hold up in this stress event. Institutions, it says, are now treating the two assets differently: gold as protection, Bitcoin as a vehicle for upside.
That leaves a clear path for capital flight. Money first exits high-valuation technology stocks into cash and U.S. Treasuries, then moves into gold. Bitcoin sits farther out on the risk curve, so it does not attract the first wave of defensive allocation.
The article also flags a new overhang. In late June, MicroStrategy announced a $1.25 billion Bitcoin monetization authorization, the first formal selling framework in the company’s history. One of the market’s biggest buyers, in the article’s words, has started to leave itself an exit option.
What could bring capital back to Bitcoin
The piece lists three conditions. Global liquidity pressure would need to ease. The Federal Reserve would need to cut rates without the economy falling into recession. And the CLARITY Act would need to clear, removing what the article calls Wall Street’s remaining compliance concerns.
The third point is the most politically delicate. The CLARITY Act passed the U.S. House in July 2025 by a vote of 294 to 134, with 78 Democrats voting in favor. But by July 2026, it was still stuck in the Senate and would not reach a vote before the August summer recess. The article says Democrats believe ethics provisions aimed at limiting Trump-related crypto interests are not strong enough, while bank lobbying groups oppose the bill’s stablecoin yield provisions.
SEC Chair Paul Atkins has said that if Congress does not pass the law, the SEC will write its own rules.
The article also notes a change in market structure. After peaking at $126,000 in October 2025, Bitcoin entered a deep correction on its own, and its tie to the Nasdaq has been loosening. Technology stocks are priced mainly on AI capital spending and corporate earnings. Bitcoin, by contrast, is priced off global liquidity. They can look synchronized in easy conditions, but they split under stress.
That lower correlation is exactly what some institutions want. A BlackRock research report, cited in the article, recommends a 1% to 2% Bitcoin allocation for institutional portfolios. For capital unsettled by concentrated AI bets, the appeal of an asset that does not move in lockstep with the Nasdaq may grow. The article ends on that note: Bitcoin is not a safe harbor yet, but it may already be early in the queue when global capital starts to reallocate again.