From 4% Crash to Monday Bounce: Inside the US Market's Wild 72 Hours

On Friday, June 5, the Nasdaq had its worst session in over a year. Over $1 trillion in market value disappeared. Semiconductor stocks fell off a cliff. Bond yields spiked on a jobs report that was twice as strong as anyone expected. By Monday morning, Asian markets were in freefall, South Korea's stock exchange halted trading altogether after the index fell far enough to trigger a circuit breaker. If you opened a financial app anywhere in the world on Monday, June 8, the color was red. Then the US markets opened and the Nasdaq went up.

Let's break down what actually happened across these 72 hours, why the bounce showed up, and what the volatility reveals about the AI trade that investors have been riding for the past two years.

Why the US Market Crashed on Friday

The trigger was Broadcom. The chip and software giant reported Q2 FY2026 earnings on June 3 that cleared the bar Wall Street had set: revenue of $22.19 billion against a $22.13 billion consensus, non-GAAP EPS of $2.44 versus $2.39 expected. By traditional standards, a beat is a beat. But the market was not grading Broadcom on what it delivered. It was grading the company on what it refused to promise.

Broadcom guided Q3 AI chip revenue at $16 billion, against analyst estimates of $17.2 billion. More critically, the company did not raise its full-year AI semiconductor sales forecast. That single omission lit the fuse. Broadcom fell 14% the following session. Then, on Friday, the broader semiconductor sector collapsed.

Sources: Yahoo Finance, CNBC, Trading Economics

The jobs report made it worse. The US Labor Department reported that the economy added 172,000 jobs in May, roughly double the 80,000 economists had forecast, with part of the beat attributed to hiring around the FIFA World Cup set to begin June 11. Strong employment is normally good news. For a market that had been pricing in Federal Reserve rate cuts, though, it meant the opposite: a hot economy gives the Fed far less reason to ease. The 10-year Treasury yield jumped above 4.5% and the 30-year crossed 5%. Higher yields compress valuations, and AI chip stocks, trading at rich multiples after two years of a bull run, bore the brunt of it. Investors rotated into Coca-Cola (+3%), Johnson & Johnson (+2%), and Colgate-Palmolive (+4%). The defensive trade was on.

How Asian Markets Reacted to the US Selloff

Asian markets had no buffer. They opened Monday carrying Friday's Nasdaq rout, plus a geopolitical shock that landed over the weekend: Iran and Israel exchanged missile strikes, the conflict entered its 100th day, and Israeli forces struck cities inside Iran despite direct requests for restraint from US President Trump. Brent crude surged past $96 per barrel.

Sources: Google Finance

South Korea's KOSPI took the sharpest hit. Samsung Electronics and SK Hynix sit at the center of global AI memory chip supply. When Nvidia and Marvell sell off in New York, Seoul feels it acutely. The circuit breaker, an automatic trading halt triggered when an index falls past a set threshold, was the starkest sign of how badly sentiment had deteriorated overnight.

Why Are US Markets Rising Today?

Three separate events gave markets a reason to buy back in before the opening bell.

• Marvell Technology was announced as a new S&P 500 constituent, effective June 22. The AI chipmaker had already tripled in 2026, buoyed partly by Nvidia CEO Jensen Huang endorsing it as a potential "trillion-dollar company" at Computex Taipei. S&P 500 index inclusion means every passive fund tracking the benchmark is now required to buy Marvell shares before June 22. That forced buying is a known, near-term catalyst. Marvell surged over 9% at the open, according to CNBC.
• Amazon announced a multibillion-dollar agreement with Corning to supply optical fiber for its expanding US data centers. Corning popped more than 9%. This was not a routine procurement deal. You do not sign fiber supply contracts at scale if you are quietly pulling back on AI infrastructure. The deal was a concrete signal that hyperscaler spending had not paused.
• And Iran's military told CNBC on Monday that it had ceased strikes against Israel. Oil's early spike partially unwound. Risk appetite returned. Nvidia added 2.3% after announcing a memory partnership with SK Hynix. Micron, which had fallen 13% on Friday, bounced over 7%.

As a result, the Nasdaq opened up 1.8%, the S&P 500 opened up 0.77%, and the Dow opened up 0.25%, as per Google Finance.

Was This a Market Correction or a Bigger Crack in the AI Trade?

Not every selloff in chip stocks is a structural signal. The challenge is separating genuine deterioration in the AI thesis from a violent shakeout of overextended positioning. Three filters help here.

Filter 1: Are hyperscalers cutting AI capex? The answer right now is no. Every major cloud provider raised its 2026 guidance on its last earnings call. Until that changes, the demand floor for AI infrastructure hardware remains.

Filter 2: Are AI revenues actually growing? AWS grew 28% in its latest quarter. Azure AI crossed a $37 billion annual revenue run rate. Google Cloud's backlog exceeded $460 billion. Demand is not collapsing, what the market questioned on Friday was the rate of future demand growth, which is a different and less alarming concern.

Filter 3: Was this a fundamental break or a positioning problem? The Philadelphia Semiconductor Index had gained over 50% in the 12 months before Friday's correction. Relative strength indicators had been flashing overbought warnings for weeks. Broadcom's cautious guidance was the match. The powder keg was built from weeks of crowded positioning in a single trade. It is also worth noting that Broadcom triggered almost the same fear in December 2024 with cautious guidance after a strong quarterly beat and the sector recovered. This pattern has now run twice.

What Could Break the AI Stocks Recovery Thesis?

The honest version of this analysis includes the scenarios where the recovery story falls apart.

1. A hyperscaler cuts AI capex on its next earnings call. Any sign that Amazon, Microsoft, or Alphabet is pulling back on data center spending removes the primary support under chip demand estimates. Q2 2026 earnings season, arriving in July, is the next test.
2. Treasury yields stay elevated. If the 10-year holds above 4.5% through the summer, which a strong jobs market makes possible, tech valuations face sustained compression, not a one-day spike. Rate-sensitive growth stocks are structurally vulnerable in that environment.
3. Broadcom's guidance was a leading indicator, not a one-off. The cautious AI chip outlook could reflect genuine softness in orders that has not yet shown up in Nvidia's or Marvell's numbers. If subsequent chip earnings also disappoint, the bear case gains real traction.
4. Geopolitical escalation disrupts energy and supply chains. The Iran-Israel ceasefire is described as fragile. Oil above $96 and any disruption to the Strait of Hormuz pushes inflation higher and narrows the Fed's options in a direction that hurts growth stocks.

What Investors Should Learn From the US Market Volatility

The 72 hours from Friday's crash to Monday's bounce did not resolve the AI debate. What they clarified is that the market is now grading AI stocks on execution, not enthusiasm.

Two years ago, announcing a partnership with a hyperscaler or mentioning AI in an earnings call was enough to move a stock. That era is over. On Friday, Broadcom delivered a beat and still fell 14%, because the guidance implied that demand growth might be slowing. On Monday, Marvell and Corning moved sharply on concrete, specific catalysts: index inclusion and a named supply contract.

For investors in India watching this from the outside, the question is whether Monday's bounce means the dip-buying window closed. It may have narrowed. But the AI infrastructure thesis, anchored by $700 billion-plus in committed 2026 hyperscaler capex and actual revenue growth in cloud services, has not changed in the past 72 hours. What has changed is the risk-reward shape: future returns will depend more on whether AI monetization keeps pace with the infrastructure bill, and less on multiple expansion from pure excitement about the technology.

That makes the research question for investors more specific now. Not "is AI real?", that debate is largely settled in the revenue data. The question that matters for returns over the next 12-18 months is whether the companies spending $700 billion to build AI infrastructure will generate enough return on that investment to justify the depreciation costs it will create. Watch the hyperscaler earnings in July for the first real signal.