Author: Xiaobing

On June 5, US stocks experienced their worst day since the tariff crisis in April 2025.

The Nasdaq Composite Index plunged 4.18% to close at 25,709 points, wiping out more than 1,121 points in a single day. The S&P 500 fell 2.64% to close at 7,383 points, marking its biggest one-day drop since October. The Dow Jones Industrial Average fell 695 points (-1.35%), just the day before it had reached a record high. The VIX volatility index surged 34% in a single day, breaking through the 20 mark, and the CNN Fear & Greed Index plummeted from "Greed" to "Fear".

Just 72 hours earlier, on June 2nd, the S&P 500 closed above 7600 for the first time. All three major indices were at record highs. The market had risen for nine consecutive weeks, a scene of euphoria, all reversed in 48 hours.

To understand this crash, we need to see how three triggers were ignited simultaneously.

Article 1: Broadcom's financial report tore open the first crack in the AI ​​narrative.

The story begins after the market closed on June 3.

Broadcom released its financial results for the second quarter of fiscal year 2026. On the surface, it was an impressive report card: revenue of $22.2 billion, exceeding Wall Street expectations; adjusted earnings per share of $2.44, also exceeding expectations; and AI chip revenue surged 143% year-over-year to $10.8 billion, far exceeding the company's own forecast.

The problem lies in the outlook for the next quarter.

Broadcom projects third-quarter AI chip revenue of $16 billion. Analyst consensus estimates are for $17.2 billion. This $1.2 billion shortfall would normally trigger only a mild correction in a normal year, but 2026 is not a normal year.

Over the past year, the valuation of the entire semiconductor sector has been based on a core assumption: that capital expenditure on AI infrastructure is unlimited and that hyperscale cloud computing companies (Google, Microsoft, Amazon, Meta) will purchase computing power at any cost.

Broadcom's financial report did not deny the high growth of AI; the 143% year-on-year growth rate is sufficient to demonstrate strong demand. It merely suggests a possibility: the growth rate may not be as steep as even the most optimistic expectations.

More critical details emerged during the earnings call. CEO Hock Tan acknowledged that Google might bring in more chip suppliers, meaning Broadcom would no longer be the sole darling. He also pointed out that the rapid growth of the AI ​​chip business was diluting the company's overall gross margin.

With a stock that has risen 88% in the past year and its valuation already "priced for perfection," these signals are enough to trigger a stampede.

Broadcom plunged 12.6% on Thursday. By Friday, panic had spread throughout the semiconductor supply chain: Micron Technology plummeted 13.2%, Marvell plunged 16.7%, Intel fell 11.3%, AMD fell approximately 11%, ARM fell 12.8%, and Qualcomm fell 11%. The Philadelphia Semiconductor Index plunged 10.26% in a single day, with all 30 component stocks suffering losses.

US-listed chip companies lost a total of approximately $1.3 trillion in market capitalization on this day.

One crucial detail: none of these companies that experienced sharp declines released any bad news. Intel, AMD, and Micron's declines were simply due to investors "extrapolating" signals from Broadcom. If Broadcom's AI growth slows, will the entire AI supply chain need to be revalued?

This is the opposite of "narrative alpha." When a story is powerful enough, all related assets will be drawn in the same direction, regardless of their individual fundamentals.

Article 2: Overly strong employment data has become poison for the market.

At 8:30 a.m. on Friday, the U.S. Department of Labor released its May non-farm payrolls report: 172,000 jobs were added, and the unemployment rate remained at 4.3%.

This number might seem modest at first glance. But compared to expectations, it's a bombshell: the Dow Jones consensus forecast is only 80,000, and the Reuters poll median is 88,000. 172,000 is a full double what Wall Street expected.

What's even more concerning is that the data for the first two months were significantly revised upwards: March's figure was revised from 185,000 to 214,000, and April's from 115,000 to 179,000, totaling an additional 93,000 jobs. The average monthly increase over the past three months was approximately 188,000, far exceeding the Federal Reserve's internal estimate of 150,000 "break-even point." As long as employment remains above this level, there is no reason to cut interest rates.

In normal economic logic, strong employment data is good news, meaning that the economy is resilient, businesses are expanding, and consumers have money to spend.

However, the United States in June 2026 will not operate according to "normal economic logic".

Since the outbreak of the war with Iran in late February, the de facto blockade of the Strait of Hormuz has driven up global oil prices. WTI crude oil remained above $92 per barrel on June 5th, while Brent crude exceeded $94. High oil prices have pushed up everything: from transportation costs to food prices, inflationary pressures have seeped into the capillaries of the economy from the supply side.

Against this backdrop, a better-than-expected jobs report sent a different signal: the economy is too hot, so hot that the Federal Reserve may not only not cut interest rates, but may even be forced to raise them.

The bond market reacted faster and more honestly than the stock market. The yield on the 10-year U.S. Treasury note jumped from 4.47% to 4.54%, hitting its highest level since late May. Data from the CME FedWatch tool was even more alarming: just a day earlier, the market was pricing in a roughly 50% probability of a rate hike before the end of the year; after the report was released, that figure jumped to 73%, and broke 80% by the close. Expectations for a rate cut virtually vanished.

This has a double impact on tech stocks.

The first layer is valuation compression. The valuations of tech stocks, especially high-growth AI-related stocks, are highly dependent on the discounted value of future cash flows. When the risk-free interest rate rises, the value of every dollar of future profit decreases today. For every percentage point increase in interest rates, the theoretical valuation of a growth stock with a forward P/E ratio of 40 could shrink by more than 10%.

The second layer is capital rotation. When bond yields rise above 4.5%, you can obtain good returns without taking any risk. For investors who have already made a fortune on AI stocks, selling overvalued tech stocks and locking in profits in government bonds becomes a simple math problem.

An interesting counter-example is that the Russell 2000 Small Cap Index bucked the trend, rising 1.45% that day. Funds flowed out of overvalued large-cap tech stocks, with some flowing into more reasonably valued, less interest rate-sensitive small and mid-cap stocks. This divergence itself illustrates that the market wasn't panicking enough to indiscriminately sell everything; it was simply repricing the extreme parts of the AI ​​narrative.

Beneath the surface of this large figure of 172,000, the quality of employment is also sending unsettling signals. Supporting this number are hotel workers (leisure hospitality +70,000), government employees (local government +55,000), and nurses (medical +35,000); however, industries that truly reflect the state of the economy are shrinking: the financial sector has lost 22,000 jobs, and employment in the information technology sector has declined by 11% since its peak in November 2022.

Wage data also doesn't stand up to close scrutiny. Average hourly wages rose 3.4% year-on-year in May, which sounds good, but the CPI had already reached 3.8% in April. Do some simple subtraction: real wage growth is negative. Nominally, wages are rising, but purchasing power in pockets is shrinking. This isn't economic prosperity; it's "the harder you work, the poorer you get."

Article 3: The lingering shadow of inflation from the Iran war

The third clue is more like an undercurrent; it won't trigger a crash on its own, but it amplifies the destructive power of the first two triggers many times over.

On February 28, 2026, the United States and Israel launched a military operation against Iran. Iran subsequently blocked the Strait of Hormuz, cutting off approximately 20% of the world's oil supply routes. The International Energy Agency characterized it as "the largest supply disruption in the history of the global oil market."

Three months have passed, and the war is far from over. Although the US and Iran reached a framework agreement for a temporary ceasefire last week, new developments in the situation in Lebanon have stalled the final agreement. Oil prices have fallen from their March high of $110, but WTI remains above $90, well above pre-war levels.

This persistently high oil price presents the Federal Reserve with a dilemma. On the one hand, supply-side inflation caused by war is not a problem that monetary policy can solve, and raising interest rates will not reopen the Strait of Hormuz. On the other hand, if inflation expectations become unanchored due to high oil prices, the Federal Reserve will have to react.

The June FOMC meeting is approaching. The Fed's latest Summary of Economic Projections (SEP) still suggests the next step is a rate cut, maintaining an accommodative stance. But the market is no longer buying it. Federal funds futures are pricing in rate hikes, not rate cuts. If the Fed is forced to shift to a hawkish stance at the June meeting, it will be the official end of the "soft landing" narrative of the past two years.

On June 5, Citigroup analysts warned that the global stock market bubble had reached its highest level since 2008.

When the foundation of the narrative begins to crumble

Looking at these three triggers separately, you'll find that they each attacked a different dimension of market confidence:

Broadcom's financial report attacks the narrative of "limitless AI growth." It doesn't say AI is bad; it simply states that growth may not remain exponential forever. But when the entire sector's valuation is based on the assumption of "exponential growth," even the slightest hint of slowdown is enough to trigger a collective revaluation.

The non-farm payroll data attacks the expectation that the Federal Reserve will soon cut interest rates. Another pillar of the stock market rally over the past year has been liquidity expectations. If the Fed not only doesn't cut rates but may even raise them, then both pillars supporting high valuations (the growth narrative and liquidity expectations) will simultaneously falter.

The Iran war attacks the consensus that "inflation has been tamed." With oil prices remaining above $90 and the Strait of Hormuz not yet fully open to navigation, the specter of inflation constantly looms over the market, making every decision by the Federal Reserve more difficult.

The combination of these three factors creates a dangerous feedback loop: slowing AI growth, pressure on tech stock valuations, rising interest rate expectations, increased funding costs, further pressure on overvalued stocks, and widespread sell-offs.

The plunge in US stocks quickly spread globally.

South Korea's KOSPI index plunged 5.54% on Friday, with Samsung Electronics falling 6.4% and SK Hynix plummeting 9.9%. Tokyo stocks also declined sharply. In Europe, ASML in the Netherlands fell 3.8%, and Infineon in Germany plunged more than 6%.

The cryptocurrency market was not spared either. Bitcoin fell by about 4% to around $60,000, Coinbase shares dropped 7.1%, and Strategy (formerly MicroStrategy) fell 6.9%. As risk assets retreated across the board, the "digital gold" narrative of the crypto market was once again put to the test by reality.

Gold futures edged down 0.35% to $4,489 an ounce, failing to fulfill its traditional safe-haven role. In an environment of rising interest rate expectations, the attractiveness of non-interest-bearing assets is also declining.

Is this the beginning of the AI ​​bubble bursting?

This is the question everyone cares about most, but the answer is not as simple as it seems.

The bearish arguments are clear: the Philadelphia Semiconductor Index plunged 10% in a single day, and this level of sell-off usually signifies a fundamental questioning of the market's growth assumptions for the entire sector. Marvell plummeted over 16% in two days, and Micron plunged 17% in two days; this indicates a wavering of confidence.

However, the bullish arguments also carry weight. Broadcom's AI chip revenue grew by 143% year-over-year, and its full-year AI semiconductor revenue guidance still exceeds $56 billion. These are figures that an industry experiencing a burst bubble should be reporting. The problem lies in the slope of growth: AI demand remains real and enormous, but can the growth rate match Wall Street's wildest imaginations?

A more accurate characterization might be: this is a "valuation repricing" rather than a "narrative collapse." The market is waking up from the euphoria of "AI can make everything skyrocket" and beginning to examine more calmly which companies can truly profit from AI and which are merely riding the wave.

The S&P 500 remains near record highs after the plunge. It has retreated about 5% from this week's high, which is historically considered a normal technical correction. The real test is: will this pullback stop at 5%, or will it slide to 10% or even deeper?

Over the next two weeks, three key moments will determine the direction of the market.

First, the June FOMC meeting. Will the Fed maintain its stance that the next step is to cut interest rates, or will it formally shift to a hawkish stance? If the Fed acknowledges the possibility of a rate hike, the market may face another round of valuation compression.

Second, more earnings reports and guidance from AI companies. Broadcom has opened Pandora's box, and the market needs other AI winners (especially Nvidia) to prove that the AI ​​growth story is not over. The next earnings season will be a key window for verification.

Third, the evolving situation in Iran. If a ceasefire agreement is ultimately implemented, oil prices will fall below $80, inflationary pressures will ease, the Federal Reserve's policy space will be significantly expanded, and the market is expected to rebound rapidly. If the war continues to drag on, everything will become more complicated.

The plunge on June 5th was a warning, not a verdict. The underlying logic of the AI ​​revolution remains unchanged, and the demand for chips still exists. What has changed is the market's expectation of growth and the price investors are willing to pay for that expectation.

Only when the tide starts to recede can you see who's swimming naked.

On June 5th, the tide was still rising, just a beat slower, but that one beat was enough to bring tears to the eyes of those fully invested, like poor editor.