How One US Company Crashed Asian Stock Markets (The Broadcom Story)
On June 4, 2026, Broadcom reported AI chip revenue of $10.8 billion — up 143% year over year.
The stock fell 12.59% the same day.
And the next morning, stock markets across Asia crashed.
South Korea’s KOSPI dropped 5.91% — triggering a circuit breaker. Samsung Electronics fell nearly 7%. SK Hynix dropped over 8%. Japan’s Nikkei fell 1.53%. Hong Kong’s Hang Seng dropped 0.81%. Broadcom wiped $300 billion in market value in a single session.
One US chip company most people have never heard of moved entire national economies. Here’s exactly what happened — and what it tells us about the AI era we’re living in.
What Broadcom Actually Reported
To understand why the reaction was so severe, you need to understand both what Broadcom reported and what the market expected.
The actual results were extraordinary by any normal standard. Total revenue of $22.2 billion, up 48% year over year. AI chip revenue of $10.8 billion, up 143%. Free cash flow of $10.3 billion. The company guided third quarter revenue to $29.4 billion — an 84% increase year over year.
But markets aren’t judged against normal standards when stocks are trading at extreme valuations after extraordinary rallies. They’re judged against expectations. And Broadcom’s expectations had become extreme.
The specific disappointments were two: software revenue of $7.18 billion missed the analyst consensus of $7.32 billion. And AI semiconductor guidance for the next quarter — $16 billion — came in below the most aggressive buy-side models, which had penciled in approximately $17.2 billion.
A $1.2 billion miss on AI guidance. In a company guiding for $29.4 billion in total revenue. And it crashed stock markets across a continent.
Why Korea Got Hit Hardest
South Korea’s KOSPI took the most severe hit of any major market — a 5.91% decline that triggered a circuit breaker for the tenth time this year.
The reason is straightforward: Korea’s stock market is more concentrated in semiconductors than almost any other major market in the world. Samsung Electronics and SK Hynix together represent an enormous share of the KOSPI’s total market capitalization. When global semiconductor sentiment turns negative, Korea feels it more than anywhere else.
Samsung Electronics is the world’s largest memory chip maker. SK Hynix is the world’s second largest. Both companies supply the HBM (High Bandwidth Memory) chips that go into NVIDIA’s most advanced AI accelerators. Their fortunes are directly tied to AI infrastructure spending.
When Broadcom’s results raised even a hint of doubt about the pace of AI infrastructure investment, investors sold Korea’s chip stocks first and asked questions later.
The irony: Broadcom’s CEO Hock Tan, on the same earnings call, said the company expects AI semiconductor revenue to exceed $100 billion in fiscal year 2027. That’s an extraordinary long-term outlook. The market focused on the near-term miss and ignored the long-term guidance.
The Ripple Effect Across Asia
The selloff wasn’t limited to Korea.
Japan’s Nikkei fell 1.53%, led by semiconductor equipment companies Tokyo Electron (down 6%) and Advantest (down 5%). These companies make the machines that manufacture chips — when chip demand expectations fall, so does demand for chip-making equipment.
Taiwan’s market was more resilient — TSMC actually edged 0.4% higher, suggesting investors distinguished between the foundry business (which manufactures chips for everyone, including both NVIDIA and Broadcom’s custom chip customers) and the semiconductor design companies that were more directly impacted by the guidance miss.
Hong Kong’s Hang Seng fell 0.81%. The broader MSCI emerging markets index fell 1.7% — its worst day since mid-May — with the technology subindex down 2.8%.
What This Tells Us About AI and the Global Economy
The Broadcom-triggered selloff illustrates something important about where we are in the AI investment cycle.
AI infrastructure has become the nervous system of the global economy. The chips that power AI data centers are designed in California, manufactured in Taiwan, and assembled into servers by companies in Texas and Tennessee. The memory chips that go into those servers are made in Korea. The networking equipment comes from California. The data centers themselves are built globally.
This supply chain is so tightly integrated — and the companies involved are so large relative to their home markets — that a disappointment at any major node sends shockwaves everywhere. Broadcom disappointed in California. Korea felt it most acutely the next morning.
Expectations have become extreme. When $10.8 billion in AI revenue (up 143%) causes a stock to fall 12%, the market is pricing in perfection. Not just good results — perfect results that exceed already-elevated expectations.
This is what happens late in a momentum cycle. Stocks rise until they price in extraordinary growth. Then even extraordinary results aren’t enough. The correction, when it comes, looks violent relative to the underlying business reality — because the underlying business is fine. The valuation was just stretched.
Korea is the most sensitive barometer of global AI investment sentiment. If you want to know how the market feels about AI infrastructure spending on any given day, watch the KOSPI. Its semiconductor concentration makes it the world’s most direct proxy for AI capex confidence.
Context: What Happened Two Days Before
The selloff is even more striking when you consider what happened just two trading days earlier.
On June 2, 2026, NVIDIA CEO Jensen Huang visited South Korea. Anticipation of his visit — and expectations about partnerships and investment announcements — sent the KOSPI to an all-time high of 8,788. Samsung Electronics surged more than 10% in a single session. The KOSPI’s market capitalization crossed 7,000 trillion Korean won for the first time in history.
Two days later, the same market fell 5.91%, triggered by a guidance miss from a company most Korean retail investors couldn’t have named.
The gap between those two days — the euphoria and the crash — captures the volatility inherent in a market so concentrated in a single theme at a moment when that theme has attracted extraordinary capital and extraordinary expectations.
Is This a Sign the AI Rally Is Over?
I don’t think so. Here’s why.
Broadcom’s CEO explicitly guided for $100 billion in AI semiconductor revenue in fiscal 2027. The company has beaten earnings estimates for seven consecutive quarters. The underlying demand for AI infrastructure — from hyperscalers spending hundreds of billions annually on AI capex — hasn’t changed because of one quarter’s software revenue miss.
What has changed is the market’s tolerance for missing elevated expectations. Stocks that have risen dramatically on AI momentum are being held to extraordinary standards. When they miss even slightly, the correction can be sharp.
But sharp corrections in the middle of genuine technology transitions are normal. They happened repeatedly during the internet era, the smartphone era, and every other major technology transition. The companies and technologies that were genuinely transformative kept growing through the corrections. The ones that were mostly hype didn’t.
NVIDIA, Broadcom, AMD, and the broader AI infrastructure complex are generating real revenue at extraordinary scale. One bad day — even one bad week — doesn’t change that underlying reality.
My Personal Take
I find the Broadcom-Korea connection genuinely fascinating as an illustration of how interconnected the global economy has become around AI infrastructure.
The fact that a guidance miss at a chip company in California can trigger a circuit breaker in Seoul — affecting the retirement savings of millions of Korean investors who have never heard of Broadcom — is a remarkable illustration of the AI era’s supply chain dependencies.
It’s also a reminder that investing in concentrated themes — even genuinely transformative ones — comes with volatility that can be difficult to stomach. The AI transformation is real. The path from here to there will include days like today.
As a regular dad investing primarily through index funds, days like today are uncomfortable to watch but don’t change the long-term thesis. The S&P 500 gained 0.41% on the same day that Korea fell 5.91% — a reminder that diversification across sectors and geographies dampens the impact of any single company’s bad day.
Broadcom disappointed relative to extreme expectations. The AI buildout continues. One doesn’t cancel the other.
Related: What is Broadcom? covers the company’s business in detail. And Is AI a Bubble? explains why I think the technology transformation is real even when individual stocks disappoint.