KEY POINTS
- Broadcom guided Q3 AI semiconductor revenue to $16 billion, below the $17.2 billion Street estimate, and held its full-year AI chip forecast at $56 billion — triggering a 12.6% stock decline and dragging the entire semiconductor sector lower.
- The iShares Semiconductor ETF (SOXX) dropped 10% on Friday for its worst session since March 2020, with Marvell down 16%, Intel and AMD each off 11%, and Micron falling 13% after briefly trading above $1,000 earlier in the week.
- The selloff raises a fundamental question: whether the AI capex cycle is decelerating or Broadcom simply set the bar too high, with $800 billion in planned Big Tech AI spending in 2026 suggesting the latter.
Broadcom reported fiscal Q2 results after the close on June 3 that were, by most conventional measures, outstanding. Revenue hit $14.9 billion. AI semiconductor revenue of $10.8 billion grew 143% year-over-year. The company guided Q3 total revenue to $29.4 billion, ahead of the $28.5 billion consensus. None of it mattered.
The Number That Broke the Trade
What mattered was what Broadcom did not do. Management guided Q3 AI chip revenue to $16 billion, below the $17.2 billion the Street was modeling. More critically, CEO Hock Tan declined to raise the company's full-year fiscal 2026 AI semiconductor forecast from $56 billion, even as analysts had been expecting an upward revision. The stock dropped 12.6% on Thursday in heavy volume.
The message the market heard was simple: AI chip demand growth may be peaking. Whether that interpretation is correct is debatable — $56 billion in annual AI chip sales growing at 180% year-over-year is not exactly a slowdown — but in a market that had priced semiconductor stocks for perfection, "in line" felt like a miss.
Contagion Spread Fast
The damage on Friday was indiscriminate. The iShares Semiconductor ETF fell 10%, its worst day since the COVID crash. Marvell Technology, which had surged 32% on Tuesday after Jensen Huang called it the "next trillion-dollar company," gave back 16%. Micron, which had broken above $1,000 per share for the first time on June 1 on the back of 196% revenue growth and its HBM capacity sold out through 2026, dropped 13% to $989. Intel and AMD each fell roughly 11%.
The selling was exacerbated by a deepening memory chip supply crisis. IDC projected that the global smartphone market will decline 13.9% in 2026 — the steepest annual contraction ever — as memory chip manufacturers pivot capacity to higher-margin AI data center customers. The average smartphone selling price has climbed 14% to $523 as DRAM and NAND shortages squeeze consumer electronics. SK Hynix, Samsung, and Micron's memory chip divisions have each crossed $1 trillion in market capitalization, creating what analysts call a $3 trillion oligopoly that is rationing supply.
Is the AI Trade Over?
The bears' case writes itself: the Nasdaq dropped 4.18% in a single session, the biggest AI beneficiaries were the biggest losers, and the hot jobs report sent yields spiking. But the bull case remains formidable. Goldman Sachs estimates that U.S. tech giants — Alphabet, Microsoft, Amazon, and Meta — will spend roughly $800 billion on AI-related capital expenditure in 2026. Alphabet alone just announced an $84.75 billion equity raise specifically to fund AI infrastructure, with $10 billion coming from Berkshire Hathaway. Broadcom reiterated a fiscal 2027 AI chip revenue target of over $100 billion.
The question is not whether AI spending continues — it clearly will — but whether the market had priced in too much too fast. With the SOXX now down more than 10% from its June 2 high, valuations are resetting. For traders, the key level is Broadcom at $155, where its forward P/E compresses to roughly 22x fiscal 2027 earnings. Below that, and the AI premium has been fully extracted. Above it, and Friday's selloff was a buying opportunity in disguise.
Watch Nvidia's response. Jensen Huang spent the week boosting Marvell and the AI infrastructure narrative. If Nvidia's management echoes Broadcom's caution in coming weeks, the correction has further to run. If Huang doubles down, the dip buyers will have their catalyst.
