just now
The week the market stopped paying any price for a perfect story.
For the better part of a year, the S&P 500 has run on a single engine: artificial intelligence, and the belief that the spending behind it has no ceiling. Every quarter the numbers got bigger, every dip got bought, and the index climbed a clean channel from the April low near 6,300 all the way to a record print above 7,600. Then, in the space of 48 hours, the market was handed two of the things it had been quietly counting on — and sold them both.
That is the story worth telling this week. Not that stocks fell, but why they fell, and what it says about the year ahead.
The two cracks
The first crack came from Broadcom. On the surface the quarter was a triumph: second-quarter revenue of $22.19 billion, up 48% from a year earlier, with AI chip sales reaching $10.8 billion, a 143% increase. By any normal standard, that is a blowout. And yet the stock fell roughly 14% in after-hours trading. The reason is the tell. Management guided next-quarter AI chip sales to about $16 billion versus the roughly $17.2 billion some analysts expected, and crucially kept its full-year AI target unchanged rather than raising it. The number was extraordinary. The problem was that the market had already priced more than extraordinary. When the full-year math landed at $56 billion rather than higher, traders read it as a ceiling rather than a floor.
The second crack came from the labour market — but in reverse. On Friday, nonfarm payrolls jumped a seasonally adjusted 172,000 in May, far above the consensus estimate for 80,000, with the unemployment rate holding at 4.3%. A strong economy should be good news. The catch is what it does to the Fed. A labour market this resilient reduces the urgency for near-term rate cuts, suggesting a prolonged period of stable rates — and high-valuation stocks faced downward pressure as a result. Good news for Main Street, bad news for a market that had been leaning on cheaper money to justify rich multiples.
Put them together and you get the real theme: the market is no longer being rewarded for good. It needs perfect, and it needs it cheap. Broadcom proved that even a 143% growth print can disappoint when expectations run ahead of reality. The jobs report proved that the rate-cut cushion stocks were resting on is thinner than hoped. Two different sources, one message — the margin for error has gone.
The longer arc
Step back and this looks less like a single bad week and more like the first real test of the move off the April low. The entire rally from 6,300 has been a story of expanding belief: belief that AI capex compounds forever, belief that the Fed rides to the rescue at the first sign of trouble, belief that any pullback is a dip to be bought. Channels like the one SPX has traced don't break because the story turns false — they break when the story stops getting better. Broadcom's "unchanged" guidance and Friday's "too strong" payrolls are both, in their own way, the story failing to improve.
That doesn't make this a top. The bull case is still intact — AI bookings are contracted, not hoped for, and a resilient economy is a strange thing to be bearish about. But the easy phase, where price simply tracked the upper rail and nobody had to think, looks finished. From here the index has to earn its level rather than assume it. The question for the back half of 2026 is no longer "how high," it's "what price is fair" — and that is a far more two-sided conversation.
The chart: S&P 500, Daily
Price has just slipped out of the lower boundary of the rising channel that has guided the entire move since April, closing the week around 7,383. That break is the technical echo of the fundamental shift above — the trend isn't broken, but it's no longer effortless.
The level to watch is 7,400. It was support on the way up; it now sits just overhead as the first test of whether buyers step back in or sellers defend the breakdown. A clean rejection below 7,400 opens the door toward the anchored VWAP from the April low (the green line), which has tracked beneath this entire advance and currently sits near 7,200. That is the line the rally has never seriously challenged — and the most logical magnet if this turns into a genuine mean-reversion rather than a one-week wobble. Reclaim 7,400 convincingly and the breakdown reads as a shakeout; lose it and the VWAP becomes the conversation.
Chart: S&P 500 Index, Daily timeframe. Rising channel off the April low, anchored VWAP in green.
The Week Ahead — Economic Calendar (8–14 June)
After the Broadcom-and-payrolls double-header, the calendar front-loads the quiet and back-loads the fireworks. Monday and Tuesday are bare. Then everything lands at once.
| DayEventWhy it matters | ||
| Mon 9 Jun | — | No major scheduled releases. |
| Tue 10 Jun | — | No major scheduled releases. |
| Wed 11 Jun | US CPI | The headline US event. After a hot payrolls print, a firm CPI would cement the "no cuts soon" read and pressure rate-sensitive growth names further. |
| Wed 11 Jun | China CPI | Read on global demand and the commodity complex; matters for risk appetite broadly. |
| Wed 11 Jun | Bank of Canada decision | A second G7 central-bank data point in the same week. |
| Thu 12 Jun | ECB rate decision + staff projections | Markets price ~91% odds of a 25bp hike to 2.25%. The projections and Lagarde's tone are the real catalyst, not the near-certain hike. |
| Thu 12 Jun | US PPI | Pipeline inflation read; confirms or complicates the CPI signal. |
| Fri 13 Jun | German final CPI | Confirms the euro-area inflation picture into the ECB. |
| Fri 13 Jun | UoM Consumer Sentiment (prelim) | Closes the week on the US consumer's mood and inflation expectations. |
The thread that ties it together: Wednesday's CPI is now the single most important release for US equities. Friday's jobs report told the Fed it can wait; a hot CPI would tell it it must wait. For an S&P 500 that just lost its channel on exactly that fear, the inflation print is the data that either rescues the dip or confirms the breakdown.
