Major technology hyperscalers including Alphabet, Microsoft, Amazon, Meta, and Oracle are on track to spend nearly $1.1 trillion on AI data center infrastructure through 2027, according to research by Wharton finance professor Jessica Wachter. The study estimates that total capital investments could exceed $5 trillion over four years, presenting one of the largest sector-specific buildouts in economic history.
However, a significant revenue gap remains. Annual AI revenues currently stand between $150 billion and $200 billion, compared to expected hyperscaler spending of $750 billion this year alone. To justify the capital outlay and deliver a 15% return, researchers estimate these companies must increase their internal productivity by 2.7x by 2030, raising concerns over potential capital misallocation if demand or revenue growth falters.
Why it matters
Hyperscaler Capex plans create severe downside risks for public tech equities and infrastructure providers if enterprise revenue monetization fails to accelerate.
Data center builders and hardware suppliers must monitor potential spending pullbacks if hyperscalers adjust return-on-investment timelines past 2030.
Source: technologyreview.com
