Meta Stock Surged 9% to $612.91 on July 1 After Reports That Mark Zuckerberg Is Building a Cloud Business to Compete With Amazon, Microsoft, and Alphabet

Meta plans to sell excess AI compute capacity via “Meta Compute,” effectively creating a new cloud-like revenue stream versus AWS/Azure/Google Cloud. The article flags capex risk: Meta’s capex rose 84% YoY to $72.2B in 2025, with projected total spend of $125B–$145B this year, and shares are down 26% from its prior all-time high. Offsetting the concern, Q1 advertising revenue increased 33% YoY, suggesting core operations remain strong, while the compute selloff is framed as either a monetization fix or a sign of AI overbuild.

Analysis

This is less a new cloud strategy than a balance-sheet admission that part of the AI buildout must be monetized externally to preserve ROIC. The market should focus on incremental margin, not gross revenue: if Meta is renting capacity at attractive rates with minimal sales/SG&A, the first $1-2B of annualized external revenue could materially soften free-cash-flow concerns. But if this is mostly a way to fill stranded inventory, it is a warning that internal demand is not absorbing prior capex at the implied hurdle rate.

Relative winners are META and, secondarily, NVDA if outside demand keeps GPU utilization tight. The more subtle loser is the cloud incumbents’ pricing power: even a small, aggressive entrant can pressure spot/short-duration compute pricing and force AWS, Azure, and Google Cloud to defend workload share on non-core training jobs. That said, this is unlikely to move full-stack cloud share meaningfully in the next 6-12 months; the real competitive effect is on pricing discipline and on the market’s willingness to assign software-like multiples to AI-heavy platforms.

Catalyst path matters: over days, the stock can trade as a relief rally if investors read this as monetization; over 1-3 months, the key is whether management quantifies contracted external demand and moderates capex guidance; over 6-18 months, sustained third-party revenue could justify a higher multiple only if utilization stays high and depreciation does not outrun cash flow. Contrarian view: the consensus may be too quick to call this bearish. The more important signal is that Meta has a fast monetization lever unavailable to most peers. Falsifier: capex stays elevated into the next print while external compute revenue remains immaterial.

Key Decisions for Investors

• META: do not chase the first headline bounce; wait for next earnings or a 3-5% pullback and only add if management discloses booked external compute demand and capex growth flattens.
• Pair trade: long META / short a basket of AMZN, MSFT, and GOOGL on any evidence that Meta is price-competitive in training/inference workloads; target this only if utilization and contract terms are disclosed, otherwise stay flat.
• NVDA: use weakness to add modestly if this story is read as demand expansion rather than supply relief; the better setup is if Meta’s external sales imply more GPU hours, not cheaper GPU hours.
• If capex remains above ~$125B with no visible external revenue by the next quarterly update, fade META rallies with a short-dated bearish call spread; risk/reward improves if the market starts questioning asset turnover.
• Watch item, not trade yet: monitor AWS/Azure/GOOGL cloud pricing commentary over the next 1-3 months for signs that Meta’s entry is forcing discounting in AI compute.