On February 28, the US-Israel coalition launched the "Epic Fury" operation against Iran, carrying out airstrikes on over 2,000 targets. Supreme Leader Khamenei was killed in the attack.
This is the largest geopolitical event to occur in the Middle East in decades.
Global capital waited through an entire weekend to verify one question: Were the narratives they had bet on for two years actually true? Is gold a hard asset in troubled times? Is Bitcoin digital gold? Is the debasement trade a real proposition or a narrative bubble?
The results are in.
Gold first rose to $5,400, then fell over 4% along with stocks. Silver plummeted 8% in a single day. Bitcoin fell initially, then fluctuated, ultimately ending up back where it started. The US Dollar Index rose 1.1%.
Narrative Stress Test
Over the past two years, a nearly flawless narrative has been popular in crypto and macro circles: out-of-control US debt, long-term dollar devaluation, with gold and BTC being hard assets that hedge against currency dilution, collectively known as the "debasement trade." In 2025, the data supporting this narrative seemed incredibly solid—gold rose over 50% for the year, BTC reached a high of $126,000, the US Dollar Index fell nearly 11%, marking its worst first-half performance in 50 years. Ken Griffin of Citadel repeatedly mentioned this term in public, and BlackRock's BTC ETF assets under management approached $100 billion.
The core implicit assumption of this narrative is: When a real crisis hits, people will flee the dollar and rush to hard assets.
Last weekend, this assumption faced its first real-intensity stress test.
At Monday's open, gold did rise initially—spot gold in London hit a high of $5,418 intraday, nearly matching the all-time high from late January. But then, as oil prices continued to soar, inflation expectations reignited, and the market began repricing the Fed's rate cut path, everything reversed. Gold closed lower on the day and fell further by over 4% on Tuesday, returning to levels last seen on February 20.
Silver fared worse. It broke above $96 on Monday, then saw intraday losses approaching 8% on Tuesday.
The reason is straightforward. Soaring oil prices mean rising inflation expectations, which mean reduced room for the Fed to cut rates, which means a stronger dollar. A stronger dollar is the direct counterpart to gold and silver. Add in forced liquidations triggered by a synchronous sell-off across all risk assets, and gold and silver, at the "time they should have risen the most," became the most liquid and easiest positions to sell.
This isn't the first time. When the Russia-Ukraine war broke out in 2022, BTC did not act as a safe haven; it fell along with the Nasdaq. Gold rose initially but was later sold off by profit-takers—the same script, just greater intensity.
BTC's situation is more awkward. That weekend the war started, the crypto market was the only market still trading. BTC fell from around $66,000 to below $64,000, then quickly rebounded, largely filling the gap before traditional markets opened on Monday. Superficially, this looks like "resilience." But the truth is: institutional money hadn't entered yet; it was retail investors and arbitrageurs finding direction amidst the volatility. On Tuesday, as the broader sell-off continued, BTC came under pressure again, hovering around $68,000. It neither showed the rise expected of "digital gold" nor the significant drop expected of a "risk asset"—it just... lingered there.
Hormuz, Chips, and a Korean Holiday
The stampede in Seoul was particularly severe for one structural reason: Monday was Samiljeol, a national holiday in South Korea, and the exchanges were closed.
Panic built up over the weekend with no outlet. At 9 AM on Tuesday, all the sell orders that couldn't be executed during the three-day holiday hit the market at the same second. The KOSPI opened directly triggering a circuit breaker warning, ultimately closing down 7.24%, wiping out approximately 377 trillion won in market value in a single day, equivalent to about $257 billion USD.
This was the largest single-day drop since the Yen carry trade crisis in August 2024. That time, the KOSPI plummeted 8.77% in a day, triggered by a crashed US non-farm payrolls report and a surprise rate hike by the Bank of Japan—a systematic unwinding of financial leverage. This time, the direct trigger was geopolitics—but behind it was同样 a tightly stretched string, just made of different material.
Over the past year and a half, Korean retail investors experienced an epic FOMO cycle. The KOSPI rose from 2400 points at the end of 2024 to break through 6000 points by the end of February this year, a gain of nearly 150% in 14 months. Some brokerages raised their target prices to 7000, even 8000. The number of trading accounts exceeded 100 million in January this year—a country of 50 million people has 100 million stock accounts. The Korean government even included "KOSPI 5000" in its policy agenda as a national policy goal.
At the same time, margin debt also expanded同步. As of before the incident, the outstanding balance of on-exchange margin loans in South Korea exceeded 32 trillion won, approximately $22.4 billion USD, the highest level since 2021. The outstanding balance of stock pledge loans was another 26 trillion won, bringing the total to nearly $37 billion USD. The market fear gauge, VKOSPI, had already soared to 54 by the end of February, more than double the "normal" level—the entire market, while hitting new highs, had entered an extreme fear zone.
This structure, when hit with an acute shock, leads to a textbook liquidity squeeze.
Falling stock prices trigger margin calls, brokers initiate forced liquidations, liquidations further depress stock prices, lower prices trigger more margin calls—a self-reinforcing feedback loop. Foreign investors net sold over 5.17 trillion won on this day, about $3.5 billion USD, the largest single-day net selloff this year. Retail investors did the opposite: buying the dip during the stampede, continuing to add leveraged ETFs, betting on a rebound.
Other Asian markets were under pressure同步. The Nikkei 225 fell 3% that day, with Toyota down 5.5% and Sony down 4.3%. The Hong Kong Hang Seng fell 2.1%, leading the decline in the Asia-Pacific region. The Stock Exchange of Thailand announced a suspension of short selling for most securities. The broader MSCI Asia Pacific index fell about 2%.
But the core structural fragility was in Seoul.
What fell wasn't just stocks. It was the KOSPI market, which had been repeatedly narrated as the "best proxy for the AI super cycle." Samsung Electronics fell nearly 10%, breaking below the key psychological support level of 200,000 won. SK Hynix fell 11.5%. These two companies together account for about 40% of the KOSPI's market capitalization and are also the two most critical nodes in the global AI chip supply chain—Samsung is the world's largest DRAM and NAND flash memory manufacturer, SK Hynix is a core supplier of HBM (High Bandwidth Memory), the memory used in most of Nvidia's AI GPUs comes from here.
These factories are all in South Korea. South Korea imports 2.76 million barrels of oil per day, mostly from the Persian Gulf, passing through the Strait of Hormuz.
Iran announced the closure of the Strait of Hormuz, then retracted the statement. But insurers have already voted unilaterally: war risk insurance suspended, shipping companies stopped scheduling. An unofficially announced blockade is effectively in operation.
Simultaneously with the market stampede, another piece of news quietly surfaced: Samsung's US wafer fab in Taylor, Texas, a core asset in the US "chip reshoring" strategy, had its mass production timeline delayed again, pushed from 2026 to 2027. This factory, built in the desert of Texas—still needs oil that comes from the Persian Gulf.
The Dollar Won, Everyone Else Lost
Now for the most uncomfortable conclusion.
Throughout this cycle, "de-dollarization" has been one of the most mainstream macro narratives. The US dollar's share of global reserves has fallen below 47%, central banks are buying gold at a record pace, BRICS is building the mBridge cross-border settlement platform, and the scale of on-chain stablecoins has expanded from $205 billion to over $300 billion. Almost everyone in the circle was betting in some way: the dollar's golden age is over, the next era belongs to multipolarity, hard assets, decentralization.
Then the war started, and the US Dollar Index rose 1.1%, marking its largest single-day gain since last May.
On this day, there was同步 selling across the board: stocks fell, bonds fell, gold fell, silver fell, commodities fell. The flow had only one direction: the US dollar.
This is the nature of a liquidity squeeze. It's not that the dollar got better; it's that the dollar is the asset with the deepest liquidity in the global financial system. In an acute crisis, when everyone needs to exit at once, the dollar is the widest door. Leverage unwinding requires dollars, margin replenishment requires dollars, the first step in fleeing cross-border assets is also dollars. No other option has this scale.
Iran provided the clearest footnote for this logic.
The Iranian rial has depreciated over 30% year-to-date. Nobitex is Iran's largest crypto exchange, handling over 87% of Iran's on-chain crypto activity. Within minutes of the airstrike beginning, withdrawals from Nobitex surged 700% (Elliptic data). Chainalysis tracked the fund flows: a significant portion of the outflowing funds ultimately landed at overseas exchanges historically receiving Iranian inflows, to be converted into USDT and USDC.
In a real currency crisis, what people do using crypto networks is flee into the US dollar.
This is not to say the debasement trade is a false proposition. The structural pressures of US debt, inflation, and the long-term erosion of the dollar's purchasing power are real. The logic of gold outperforming in the long run was not overturned by this stress test.
But this tells us one thing: the debasement trade is a slow-variable narrative; it requires the刻度 of time to hold true. Geopolitical shocks are fast variables; they have only one刻度—today.
When two time frames are placed at the same moment, the slow variable yields.
The dollar won. In the next crisis, the dollar will most likely win again. Until one day, it doesn't—but that day will most likely not begin with this kind of dramatic opening.
