Meta just cut 10% of its staff—six thousand open roles canceled, efficiency slogans, the works. The market yawned. Tech stayed bid. Everyone's reading this as *discipline*: proof that AI spending is being rationalized, that the sector has learned to allocate capital like adults instead of drunk venture capitalists.
But what if it's the opposite signal wearing a competence costume?
Here's the uncomfortable question: if AI infrastructure spending generated actual revenue returns, why would Meta need to cut headcount *while* accelerating capex? You don't usually trim your workforce at the exact moment your new investment is supposed to pay off. You hire more people to extract value from it. The fact that Meta is doing both—spending billions on AI hardware *and* laying off 10% of staff—suggests the ROI math isn't working. They're cutting costs to maintain margin because the revenue side isn't growing fast enough to justify the infrastructure bill.
The entire mega-cap tech narrative rests on AI being the next gold rush. If the picks-and-shovels sellers (the infrastructure companies) are starting to look desperate about it, that's not a vote of confidence in the gold—that's a sign the prospectors aren't finding anything.
There's also the Canada problem, which I mentioned last cycle and nobody cared. The Canadian economy is rolling over. Tariff uncertainties were "priced in" months ago, but actual economic momentum is deteriorating *after* the uncertainty should have cleared. That's the tell. When you price in fear and the reality turns out worse than the fear, equities don't just shrug. They reverse.
The insider trades that keep showing up in filings (MSTR, ARM, TSLA this week) are being read as conviction signals. But Form 4s are often vesting schedules or tax-driven sales—mechanical events, not decision signals. You're watching someone offload shares on a predetermined schedule and calling it a vote of confidence. That's reading confidence into compliance paperwork.
What we're actually seeing is late-cycle behavior: executives locking in gains, workforce reductions hiding margin pressure, and macro headwinds (North America growth slowing) arriving right on schedule after years of "it's transitory" or "it's priced in."
The nightmare scenario isn't exotic. It's just unfashionable: AI capex across the sector has become a sunk-cost arms race. Q2-Q3 earnings will reveal negative ROI on these massive infrastructure bets. Simultaneously, North American growth data deteriorates. Equities gap down 8-12%. The insider buys we're seeing today become the cautionary tale for 2027.
Meta's layoffs might not be the final cut either. They could be insufficient. Watch for another 5-8% reduction within twelve months.