Big Tech AI Capex Projected to Top $1T in 2027
Get the Finance newsletter
Daily finance — markets, central banks, M&A, the prints that move money. Free.
- Evercore and Bank of America estimate combined Big Tech AI capital expenditures could top $1 trillion in 2027, with 2026 spending now tracking toward $800–$900 billion following Wednesday's earnings.
- Microsoft raised its 2026 capex guidance by 24% to $190 billion — the largest increase among the four hyperscalers and nearly double its prior pace.
- Alphabet reported 63% year-over-year cloud revenue growth and a roughly 10% stock jump, with Google Cloud's backlog hitting $462 billion (a 400% annual increase) that analysts say "supports the capex super-cycle."
- Meta shares fell about 8% as its free cash flow collapsed from $26 billion to just $1.2 billion year-over-year in Q1, with Jefferies saying the company "remains in the penalty box pending clearer capex ROI."
- Meta is doubling 2026 infrastructure spending from 2025's $72 billion to between $125 billion and $145 billion, with CEO Mark Zuckerberg citing higher component costs, "particularly memory pricing."
- Intel, Nvidia, AMD, and custom-AI-chip programs (TPU, Trainium, Maia, MTIA) are positioned as the buildout's biggest beneficiaries, with RBC maintaining positive ratings across the chip ecosystem and Evercore predicting "a CPU renaissance."
Why it matters: The $1 trillion trajectory locks in a multi-year demand pull for chipmakers and infrastructure providers, but the divergent market reactions — Alphabet's 10% surge versus Meta's 8% drop and 95% free cash flow collapse — show investors are now actively pricing in which hyperscaler can actually convert capex into returns, not just who spends most.
