AI capex scrutiny hits Amazon, Google as 2026 spending nears $375B
Alphabet and Amazon are facing investor scrutiny over AI capex plans as the broader tech sector sells off. Alphabet expects 2026 capital expenditures of about $175–$185B, while Amazon guides roughly $200B, with nearly 80% of Amazon’s outlay estimated to be AI-related.
Market reaction has been sharp. On June 22, Alphabet shares fell about 6% and Amazon dropped around 4%. The selloff reflects the second “capex punishment” wave this year, after a February selloff that wiped out over $1T in combined Big Tech market value.
The core issue is the ROI timeline. Cloud revenue and enterprise AI monetization are growing, but not fast enough to justify hundreds of billions in AI capex. As spending accelerates faster than topline growth, free cash flow is pressured, reducing room for buybacks/dividends and weakening financial flexibility.
Operational bottlenecks also matter. Advanced-chip supply constraints, power buildouts, and data-center cooling needs can delay returns and push payoff further into the future.
What traders should watch: cloud segment margins, revenue per AI workload metrics, any AI infrastructure return on invested capital disclosure, and—most importantly—free cash flow trends. With major hyperscalers racing to build similar capacity, there is also an overcapacity risk if AI demand growth plateaus or arrives later than expected.
If free cash flow deteriorates further, the negative sentiment could spill into the broader Nasdaq. If cash flow holds despite AI capex, markets may eventually reward the long-term positioning.
Bearish
This news is bearish because it targets a key valuation support: free cash flow. Big tech’s AI capex ramp (Alphabet $175–$185B; Amazon ~$200B, mostly AI-related) is outpacing topline growth, raising the market’s fear that cash will be consumed faster than it can be monetized. That pattern has historically driven drawdowns in the tech sector when investors concluded “spending > returns,” such as during prior capex/AI infrastructure repricing episodes in early 2020s.
Short term, the immediate trigger is earnings-model risk: if cloud margins soften and free cash flow deteriorates quarter over quarter, traders typically cut exposure and weigh Nasdaq beta contagion (index-level pressure).
Long term, the story is not purely negative—enterprise AI demand is growing—but timing uncertainty (chip/power/cooling constraints) can keep return horizons farther out. Unless disclosures show strong ROI metrics (e.g., revenue per AI workload and AI-specific ROIC) alongside stable free cash flow, the market may continue to treat AI capex as a drag rather than a catalyst. Net: risk of extended volatility and downward pressure remains higher than upside confirmation until cash-flow durability is proven.