Tech Stocks Are Bleeding While Dow Hits a Record High. Here's What Is Really Going On

Tech stocks are bleeding right now. The Nasdaq100 index is down nearly 2% today alone, roughly 3% over the last two days, and electronic tech, communication, and tech services stocks are off up to 4% this week. This is not just a Wall Street story. South Korea's KOSPI has fallen 5% in five days and triggered a circuit breaker this morning. Taiwan's TAIEX has corrected roughly 3%. Japan's Nikkei is down too.

And while all of this is happening, the Dow Jones Industrial Average just hit an all-time high of 51,561, led by healthcare, financials, and industrials, sectors that spent the last two years being ignored while AI stocks took all the money and all the attention.

Let's break down why tech stocks are suddenly the worst place to be, what triggered this sell-off across continents, and what it tells you about where markets are actually heading.

Global Market Sell-Off: Nasdaq, KOSPI and TAIEX Decline

Before getting into the why, here is what actually happened across markets in the last week.

Sources: Google Finance | Data as of June 5, 2026

Why Are Tech Stocks Falling Today?

On June 3, after US markets closed, Broadcom, a company that makes custom AI chips for Google, Meta, Anthropic, and OpenAI, reported its Q2 2026 earnings. The numbers were strong by any normal measure. Revenue of $22.19 billion, up 48% from a year ago. AI chip revenue of $10.8 billion, growing 143% year over year. Earnings per share (EPS) beat Wall Street estimates, per LSEG data.

But None of that mattered.

Broadcom's AI chip revenue guidance for the next quarter came in at $16 billion. Analysts had been expecting $16.36 billion, per Reuters. That gap of roughly $360 million on a $22 billion company was enough to spook the market. CEO Hock Tan also did not raise Broadcom's full-year AI revenue target, and mentioned that Google may look to diversify its chip supply chain.

By the next morning, Broadcom stock was down approximately 15%, erasing close to $280 billion in market value in a single session, one of the largest single-day market cap drops in Wall Street history.

The Broadcom crash dragged AMD down 3.5%, Micron down 7.7%, Arm down 4.5%, and Marvell Technology more than 7%, per Yahoo Finance. One company's earnings report pulled an entire sector with it.

AI Capex Boom: The $725 Billion Spending Question

This is where the real story begins. Hyperscalers like Amazon, Microsoft, Google, Meta and Oracle are expected to spend around $725 billion on AI infrastructure in 2026. That is about 64% higher than last year. Goldman Sachs expects AI-related capex to account for more than 3% of US GDP in 2026, up from 1.6% in 2025.

The pressure is clear. Bank of America estimates that these hyperscalers may spend nearly 90% of their operating cash flow on capex in 2026, compared with 65% in 2025. Alphabet alone has guided for $180–190 billion in 2026 capex, far above its trailing twelve-month free cash flow of $73.3 billion, according to Convergences. That gap has to be funded. And increasingly, it is being funded through debt. Allianz Research says hyperscalers raised more debt in the first three months of 2026 than they did in all of 2025.

Now comes the uncomfortable question: what is all this spending actually buying?

The companies building AI products on top of this infrastructure generated around $35 billion in combined revenue in 2026. OpenAI, the biggest name among them, is at roughly $20 billion in annualised revenue. That is impressive, but it looks small next to the $725 billion being spent to build the infrastructure around AI.

For the last 18 months, investors were willing to believe this spending cycle would keep accelerating. Broadcom’s guidance changed that mood. The issue is not that Broadcom failed. The issue is that Broadcom is one of the clearest real-time indicators of AI infrastructure demand. And this time, the signal was not “we are beating expectations.” It was closer to “we are on track.”

That was enough to trigger a market reset.

Why AI Stocks Are Correcting Despite Strong Demand

Here is what the "Broadcom missed guidance" framing misses.

The Philadelphia Semiconductor Index, which tracks the world's top listed chip companies, had gained approximately 90% in 2026 before this week, according to Bloomberg. Broadcom itself had closed at a record high the night before its earnings. That stock price was not a reflection of what the company does today. It was a bet on what the company would do quarters and years from now.

When you price a stock for a specific future, even a small deviation from that future reads as a failure.

This is the AI Perfection Trap. Stocks priced for infinite acceleration, where maintaining a forecast is treated the same as cutting one. And it does not just affect the stock of the company that reported. It resets the expectation for every AI-adjacent company trading at similar valuations.

Why South Korea, Taiwan Stocks Fell After Broadcom Earnings

The sell-off did not stop at Wall Street, and reached Seoul and Taipei before most Indian investors had even woken up.

By the Asian trading session on June 5, South Korea's KOSPI had lost over 5.5%, with the Korea Exchange triggering its circuit breaker after KOSPI 200 futures fell past the threshold. Samsung Electronics dropped over 7%. SK Hynix fell close to 10%. SoftBank in Japan fell over 11%, per CNBC. Taiwan's TAIEX dropped 1.3%, dragging down Foxconn, Wistron, and other AI server assemblers.

The reason these markets are so vulnerable is structural, not emotional.

Samsung Electronics and SK Hynix together made up 42.2% of the entire KOSPI in May 2026. TSMC alone accounts for over 40% of Taiwan's TAIEX benchmark. Nearly half of each of these major national stock indices sits in one sector. One earnings report from one US chip company can trigger a circuit breaker in Seoul.

Sources: AP via Lancaster Online, CNBC, IndexBox, Manulife Investment Management, UOB

South Korea's KOSPI had rallied about 90% in 2026, a pace no other market matched. Taiwan was up about 53%. Both surges were driven by AI chip demand. That concentration is why the correction, when it came, was sharp. There was no diversification to absorb the hit.

Why the Dow Jones Hit a Record High as Tech Stocks Fell

While this bloodbath in tech stocks is spooking investors, Dow Jones hit a record high of 51,660.40 today. This rise to this peak was supported by gains in UnitedHealth. Goldman Sachs, JPMorgan, and Johnson and Johnson. This is not a coincidence. It is a rotation.

Industrial, consumer defensive, and energy stocks are outperforming the broader market by a wide margin in 2026, according to Morningstar, as investors look beyond the AI trade for returns. Companies like Walmart, Caterpillar, and Eli Lilly are gaining because they are profitable, steady, and their valuations were not built on assumptions about infinite AI revenue growth.

When AI euphoria fades, even briefly, money moves toward companies that make money in predictable ways. That money has to go somewhere. Right now, it is going into healthcare, industrials, and financials.

What the AI Stock Sell-Off Means for Investors

TD Cowen reiterated a buy rating on Broadcom with a $500 price target after the crash. BNP Paribas Exane raised its price target to $640. Both suggest the sell-off may have been an overreaction to a minor guidance shortfall on a company with otherwise strong fundamentals.

TSMC's CEO C.C. Wei told investors it will be years before global chip supply catches up with demand. That is not a sign of a dying cycle. It is a sign of a cycle that has gotten too far ahead of itself.

Goldman Sachs strategist Tim Moe said the AI hardware theme is still "clearly what is propelling" Asian markets. The structural story has not changed. The expectations had.

Should Investors Buy the Dip in Tech Stocks?

The answer depends on what you own and why you own it.

For anyone investing in US tech, this week is a reminder that the market is heavily concentrated. The Nasdaq 100’s top ten holdings make up nearly 40-45% of the index. However, the long-term AI story has not disappeared. 

A $725 billion infrastructure buildout does not get written off in one week. Data centre demand is still strong. Cloud companies are still spending. AI adoption is still growing. But there is a difference between a strong business story and a stock that has already priced in perfection. 

Many AI-linked stocks were valued as if they would keep beating expectations every quarter. When that streak breaks, even slightly, the reset can be sharp. For long-term investors, this sell-off may create better entry points than last week. But it should not be treated as an automatic buying opportunity. 

The real question is whether you understand your exposure. So the better question is not “should I buy the dip?” It is: “Do I understand what I own?” If you do, this week may simply be volatility. If you do not, it is a warning.