Alphabet, Apple, Amazon, Microsoft, and Facebook collectively spent $80B in the last year on physical assets, including specialized manufacturing equipment
Alphabet, Apple, Amazon, Microsoft, and Facebook—are fast becoming industrial giants https://www.bloomberg.com/... via @BW Tamara Dull / @tamaradull : “Companies that deal in data are some of America's biggest buyers of equipment.” ~ via @business #bigdata #emergingtech http://j.mp/2DtokHr
Context & Ripple Effects
Bloomberg's 2018 tally — $80B across Alphabet, Apple, Amazon, Microsoft, and Facebook — reads today as the opening entry in a spending curve that has since steepened relentlessly: combined capex for the same cohort hit $140B by 2021, passed $200B in 2024 (over $200B total), and is now forecast at roughly $650B for 2026 (the ~$650B 2026 forecast). The through-line is that companies whose products are data have become America's heaviest buyers of physical equipment, exactly as Tamara Dull's framing in the piece suggests.
What has changed since 2018 is not just scale but funding structure: the top US data-center spenders have layered roughly $350B of new debt over five years (that debt build-up) onto what was once cash-flow-funded capex, turning an engineering story into a balance-sheet one.
First-order effects
- Suppliers of servers, networking gear, and specialized manufacturing equipment gain their largest and most reliable customer base overnight, with five buyers now setting demand signals for entire hardware categories.
- Each of the five firms commits its balance sheet to owning physical capacity rather than renting it, converting R&D-led cost structures into depreciation-heavy ones.
Second-order effects
- Rivals outside the group face a matching-funds problem: staying competitive in cloud and consumer services requires comparable facility investment, pushing peers like Oracle into the same spending race seen in later coverage.
- Equipment vendors and construction supply chains begin pricing and allocating capacity around hyperscaler orders, squeezing smaller data-center operators' access to components and sites.
Third-order effects
- If the pattern holds, the industry's structure inverts: software businesses become capital-intensive utilities whose returns depend on utilization of owned plants, and their financing migrates from operating cash to debt markets — the dynamic visible in the 2026 debt analysis.
- Sustained concentration of equipment purchasing among a handful of firms gives those firms outsized influence over hardware roadmaps, effectively steering supplier innovation toward their workloads.
The trend: Big Tech is completing a decade-long transition from asset-light platform companies to capital-intensive infrastructure owners, with each year's capex record becoming the next year's baseline.