Meta says it will lay off 10% of its workforce

This is a margin-protection story, not a simple cost-cutting story. The market’s main mistake is treating the workforce actions as a sign of distress rather than as a financing mechanism for an AI capex cycle that could keep resetting consensus estimates lower on near-term earnings power while improving long-duration optionality. For META, the immediate beneficiary is free cash flow optics: payroll reduction helps fund infrastructure spend without forcing an even larger step-up in external financing, but it also signals that management is willing to sacrifice operating leverage to preserve the AI race.

The second-order effect is competitive, not just company-specific. If META and MSFT keep absorbing hundreds of billions into AI buildout, the real losers are mid-cap software and internet names that rely on abundant platform demand and cheap cloud pricing; hyperscaler capex tends to crowd out adjacent spend and keep pricing pressure elevated on downstream vendors. For AMZN, the relative immunity makes sense: its diversified cash engine and infrastructure leverage mean investors are less likely to punish a restructuring narrative unless AWS growth slows materially.

Near term, the risk is multiple compression rather than earnings collapse. The market can tolerate layoffs if they are viewed as disciplined capital reallocation, but it will not tolerate evidence that AI spend is rising faster than monetization — that would hit META and MSFT over the next 1-3 quarters through lower terminal margin assumptions. A reversal likely requires either clearer AI revenue conversion or a pause in capex growth; absent that, every incremental restructuring announcement may be read as confirmation that AI returns are still back-end loaded.

The contrarian view is that the selloff in META and MSFT may already reflect too much skepticism if investors are underestimating the operating leverage from fewer hires and slower backfill. If management uses this window to keep total headcount flat while infrastructure scales, 2027-2028 earnings power could surprise to the upside once utilization improves. The key is that the bad headline is also the discipline signal: in a market worried about waste, showing willingness to cut labor to fund AI may eventually support a rerating if growth holds.