News & analysis · 7 June 2026
AI chip selloff wipes $1.3 trillion: when a strong jobs report became bad news for semiconductors
Wall Street entered the weekend nursing its worst tech bruise in more than a year. U.S. chipmakers lost roughly $1.3 trillion in market value on Friday as the PHLX Semiconductor Index plunged 10.3% — its steepest one-day drop since the March 2020 pandemic panic. Nvidia fell about 6%, cleaving more than $300 billion from the world’s most valuable chip company. Intel dropped 11%, Micron 13%, Marvell 17%, and AMD nearly 11%. The proximate spark was not a single earnings miss but a collision: a May nonfarm payrolls print of 172,000 jobs — more than double the consensus near 80,000 — that forced traders to reprice Federal Reserve policy toward hikes, not cuts, at the same moment Broadcom’s quarterly outlook reminded investors that even AI-levered silicon can disappoint sky-high expectations.
Good news for workers, bad news for duration
The May employment report landed like a policy grenade. Employers added 172,000 jobs while the unemployment rate held at 4.3%. March and April were revised higher by a combined 93,000, extending the strongest hiring stretch in two years. For households, that is resilience. For markets priced on the assumption that a softening labor market would let the Fed ease in 2026, it was the opposite.
Within minutes, the CME FedWatch tool showed traders assigning a majority probability to at least one rate hike before year-end — roughly 57% by some readings, overtaking the combined odds of a hold or a cut. The 10-year Treasury yield pushed past 4.5%, a 12-month high; the policy-sensitive 2-year climbed toward 4.16%. Bond markets were not celebrating growth — they were discounting a more restrictive path for money.
Growth stocks, and AI names in particular, are long-duration assets: much of their present value sits years in the future. When discount rates rise, those distant cash flows shrink in today’s terms. That mechanical link is why a jobs beat that would have cheered equity bulls in 2021 instead triggered a broad risk-off move on Friday. The S&P 500 fell 2.6%, snapping a nine-week winning streak. The Nasdaq 100 dropped roughly 5% — its worst session since April 2025. Bitcoin, already under pressure from ETF outflows and IPO-window rotation, slid toward the $62,000 area alongside the equity flush.
For readers tracking the macro framework, our interest rates and markets guide walks through how yield moves transmit into equity multiples — the same transmission channel that amplified this week’s cross-asset selloff.
From Broadcom whispers to sector-wide capitulation
Macro repricing alone might have produced a routine growth-stock pullback. What turned Friday into a semiconductor earthquake was timing. The PHLX chip index had hit a record high on Wednesday, riding a year-to-date gain still north of 70% even after the carnage. Valuations had compressed years of AI revenue optimism into weeks of price action — including a Marvell surge of more than 30% in a single session earlier in the week after a few bullish sentences from Nvidia CEO Jensen Huang.
Broadcom’s report broke the spell. Custom AI accelerator demand — the segment Wall Street had treated as an endless backlog — came in short of elevated expectations. The disappointment rippled because Broadcom sits in the middle of the hyperscaler supply chain: if the biggest cloud buyers are not accelerating orders as fast as models assumed, every vendor selling into the same capex wave looks vulnerable. Nvidia, AMD, Marvell, Arm, and memory suppliers like Micron were sold as a basket, not as individual stories.
Reuters tallied the damage: roughly $1.3 trillion erased across U.S.-traded chip names in one session. Even after Friday, the sector’s year-to-date advance remains enormous — which matters for how traders frame the move. This was not a fundamental repudiation of AI demand so much as a valuation and positioning reset at the moment rates repriced against it. The same week brought reminders that AI infrastructure competes for physical resources everywhere — from data-center DRAM to consumer GPUs, a tension we covered in our gaming memory shortage analysis.
Who fell hardest — and what the moves actually mean
Nvidia (−6%): Still the gravity well of the trade. Losing $300 billion in a day sounds catastrophic, yet Nvidia entered the week with record revenue guidance and dominant share in AI training chips. Friday’s drop is macro and sentiment as much as fundamentals — a reminder that market cap can leave earnings reality far behind, then snap back when rates move.
Intel (−11%): Painfully ironic timing. Foxconn announced an AI infrastructure partnership with Intel on Thursday; the stock sold off anyway, swept into the sector tide. Intel’s recovery narrative needs years of execution; traders did not wait to price near-term multiple compression.
Memory and networking (Micron −13%, Marvell −17%): Higher-beta expressions of the same capex cycle. Memory pricing had benefited from AI server builds; networking silicon had ridden custom accelerator enthusiasm. When Broadcom hinted that the slope of demand might bend, these names absorbed the convex downside.
The Globe and Mail noted that two-thirds of NYSE-listed stocks were in the red Friday — this was not purely an AI story. Energy and defensives held up better as investors rotated toward cash flow today rather than promises tomorrow. That breadth suggests the jobs report reset the entire risk appetite dial, with chips as the most crowded expression of the prior regime.
The Fed meeting that now matters more than any earnings call
Markets head into June 16–17 with a new chair at the helm. Kevin Warsh’s first Federal Open Market Committee meeting arrives under white-hot political pressure for cuts — and fresh data arguing for patience. Traders do not widely expect a hike on the 17th itself; the fight is over the path for the second half of 2026. Goldman Sachs pushed its first-cut forecast to December; swap markets began pricing meaningful odds of a quarter-point move as early as October.
For AI equities, the policy calendar now rivals the product calendar. Hyperscalers are funding data-center buildouts with a mix of cash flow and debt; higher rates raise the carrying cost of that bet. Alphabet’s massive equity offering earlier in the week had already reminded investors that AI leadership requires continuous capital — not a one-time GPU purchase. When bonds sell off and stocks follow, the AI trade faces a double bind: funding costs up, discount rates up.
Meanwhile, the liquidity picture off-exchange remains crowded. SpaceX, OpenAI, and Anthropic IPO windows are absorbing attention and cash from risk assets, a rotation we analyzed alongside Bitcoin ETF outflows in our AI rotation and IPO window piece. Friday’s chip rout did not happen in isolation — it stacked on a week when crypto already posted its worst weekly decline since the FTX collapse and when corporate treasuries were being questioned for the first time in years.
What to watch from here
Inflation prints before the 17th: Strong jobs with sticky CPI would cement hike pricing. Soft inflation could let equities breathe even if the Fed holds.
Hyperscaler capex commentary: The next Microsoft, Amazon, Google, or Meta update on AI spend will either validate Broadcom’s caution or rekindle the backlog narrative. Silicon stocks will follow the tone more than the line items.
Positioning and volatility: A 10% single-day sector drop after a record high often produces reflex bounces — and second-leg selling if macro data do not cooperate. Readers managing exposure should think in scenarios, not headlines; our risk management and position sizing guide covers how to size trades when correlation spikes across asset classes.
Cross-asset spillover: Bitcoin and high-beta crypto often trade as liquidity gauges when growth stocks wobble. ETF flow data next week will show whether Friday was capitulation or the start of a longer de-risking — especially with mega-IPOs on the calendar for mid-June.
Bottom line
Friday’s AI chip selloff was two stories braided into one tape. Macro: a labor market too strong for the rate-cut consensus that carried stocks for months. Micro: an AI supply chain so hyped that Broadcom merely meeting reality felt like a miss. Together they erased $1.3 trillion in chip market cap and reminded investors that the AI buildout is both real and cyclical — financed with borrowed money, sensitive to discount rates, and competing with every other risk asset for the same pool of capital.
The semiconductor index is still up dramatically in 2026. Nvidia is still Nvidia. What changed over one session is the price of time: when the Fed’s next move might be up, not down, and when even the companies selling shovels in the AI gold rush can miss the whisper number. Chair Warsh’s debut meeting arrives with markets no longer assuming he gets to ease his way out of the politics. That is the undercurrent beneath the PHLX chart — and it will outlast any single jobs report revision.
Sources: Reuters via The Express Tribune — chip slump erases $1.3T; The Globe and Mail — why AI stocks plunged; Crypto Briefing — jobs data and rate hike fears; TOPONE Markets — 172K jobs and Fed hike odds; FX Leaders — Nasdaq 100 sinks 5%. Related on Solana Garden: Bitcoin ETF outflows and AI rotation, Gaming memory shortage, Interest rates and markets, Risk management and position sizing.