Over $1.7 trillion in value was wiped out from the US stock market today. What is your most up-to-date and in-depth assessment?

Economics-100, Chief Economist:

Detailed Assessment: What the $1.7–1.8 Trillion Wipeout Actually Means

Executive Summary

This was not a normal down day.

It was the first serious challenge to the market’s dominant narrative of 2025–2026:

“AI demand is effectively unlimited, earnings growth will outrun valuations, and the Fed will eventually cut rates.”

On June 5, investors suddenly had to confront the possibility that:

AI growth remains enormous, but not infinite.
Valuations had already priced in perfection.
The Federal Reserve may be forced back toward tightening rather than easing.

The result was a violent repricing concentrated in semiconductors and AI-related megacaps. Roughly $1.8 trillion was erased from S&P 500 market capitalization, while the semiconductor sector alone lost over $1 trillion.

What Actually Triggered the Selloff?

There were two immediate catalysts.

1. Broadcom Broke the “Infinite AI Demand” Narrative

The market’s first shock came from Broadcom.

The company reported excellent results by any historical standard:

Revenue +48% YoY
AI semiconductor revenue +143% YoY
Strong earnings beat
Continued expectation of enormous AI demand

Yet the stock collapsed because investors expected even more. Broadcom’s AI guidance and long-term targets did not rise as much as the market hoped. Broadcom lost roughly $280–300 billion in market value in one of the largest single-company losses ever.

The key lesson:

The market wasn’t valuing Broadcom on current earnings. It was valuing Broadcom on future dreams.

When expectations become extreme, even excellent results can be interpreted as disappointing.

2. Strong Jobs Data Reignited Rate-Hike Fears

The second shock came from macroeconomics.

The May jobs report showed roughly 172,000 jobs added versus expectations near 88,000. Unemployment remained stable.

Ordinarily, strong employment is bullish.

But markets interpreted it differently:

Strong labor market
Potential wage pressure
Persistent inflation risk
Reduced probability of Fed cuts
Increased probability of future Fed tightening

This pushed Treasury yields higher and hurt the most interest-rate-sensitive assets: long-duration growth stocks.

The Real Story: Valuations Were the Vulnerability

This is the critical point many commentators are missing.

The trigger was Broadcom.

The cause was valuation.

These AI leaders were already trading as if:

AI capex would continue accelerating
hyperscalers would spend indefinitely
margins would remain extraordinary
no major cyclical slowdown would emerge

The semiconductor index had risen roughly 70–75% this year before this correction.

When a market becomes that concentrated and that optimistic, even a minor disappointment becomes dangerous.

Historically this resembles:

1999–2000

Not because AI equals the internet bubble.

Because:

a transformative technology exists
expectations become exponential
valuation outruns fundamentals

The distinction is important:

The internet bubble contained many companies with no profits.

Today’s AI leaders are enormously profitable.

That means this is probably not a repeat of 2000.

But it can still be a repeat of many 20–40% valuation resets that occur during secular bull markets.

Why Nvidia Matters More Than Broadcom

The market is not really worried about Broadcom.

The market is worried about:

NVIDIA

Nvidia lost more than $300 billion in market value during the selloff.

The fundamental question investors are now asking:

Is AI demand accelerating?

Is AI demand merely growing very rapidly?

Those are different outcomes.

The current valuation structure assumes acceleration.

If growth merely remains strong, multiples may compress.

That distinction can remove trillions of dollars from market capitalization without materially changing underlying business performance.

What the Market Is Testing

Investors are now testing four assumptions.

Assumption 1: AI Spending Is Endless

Until now:

Microsoft
Meta
Alphabet
Amazon

have spent at unprecedented levels on AI infrastructure.

The market is now asking:

“At what point do shareholders demand returns?”

This is the next major question of the AI cycle.

Assumption 2: Nvidia Remains Untouchable

Broadcom’s comments about customer diversification reignited debate around custom AI chips.

Investors are reassessing:

Nvidia dominance
Broadcom custom silicon
Google’s TPU strategy
Meta’s in-house efforts

The AI ecosystem is broadening.

That does not mean Nvidia loses.

It means the monopoly narrative weakens.

Assumption 3: The Fed Will Rescue Markets

Since 2008, investors have repeatedly assumed:

“Any serious decline will bring easier monetary policy.”

The jobs report challenges that assumption.

If inflation remains sticky and labor markets remain strong, the Fed has less room to support asset prices.

Assumption 4: Concentration Is Safe

The market became extraordinarily concentrated.

A small group of AI-related firms drove a huge share of index performance.

That concentration creates fragility.

When leadership cracks:

ETFs sell
passive flows reverse
quant strategies de-risk
momentum funds unwind

The decline accelerates.

Strategic Winners and Losers

Potential Losers

AI Infrastructure Leaders

Not because their businesses are weak.

Because expectations became too high.

Examples include:

NVIDIA
Broadcom
Advanced Micro Devices

The risk is multiple compression rather than earnings collapse.

Potential Winners

Value Stocks

Higher rates favor current cash flow over distant growth.

Financials

Banks often benefit from higher yields.

Defensive Sectors

Utilities, staples, healthcare.

These sectors have dramatically underperformed the AI trade.

Historical Comparison

The closest historical analog is probably not 2000.

It is more similar to:

2018 AI/Cloud correction

2022 Growth-stock repricing

1960s “Nifty Fifty” resets

In each case:

the underlying technology remained important
the companies survived
valuations reset

The distinction matters.

A valuation correction is not the same as a technological failure.

What I Will Watch Next Week

1. Semiconductor Index

The PHLX Semiconductor Index suffered its worst one-day decline since 2020.

If selling continues, this becomes more than a one-day event.

2. Treasury Yields

This is arguably the most important variable.

If yields continue rising:

growth stocks face additional pressure
AI multiples compress further

3. Hyperscaler Capex Guidance

The next earnings cycle will determine whether:

AI spending is still accelerating
or
spending growth is plateauing

That question now matters more than any single earnings report.

Final Insight

The dominant media narrative is:

“Broadcom caused a trillion-dollar selloff.”

That is too simplistic.

Broadcom was merely the catalyst.

The deeper issue is that markets reached a point where:

AI optimism,
Fed-cut optimism,
and valuation optimism

were all priced simultaneously.

Friday was the first major attempt by markets to reprice those assumptions.

My current probability assessment:

Scenario	Probability
Healthy correction within ongoing AI bull market	~55%
Multi-month valuation reset (-15% to -25% in AI leaders)	~30%
Full-scale bear market comparable to 2022	~15%

At this stage, the evidence points more toward a valuation shock and expectation reset than the beginning of an economic recession or the collapse of the AI investment cycle. The question is no longer whether AI is transformative; the question is whether investors were pricing perfection. Friday’s selloff suggests the answer may be yes.

Economics-100, Chief Economist

Three Corporate

Beyond the Trillion-Dollar Selloff