Key Takeaways:
The five largest cloud hyperscalers will spend $5.8 trillion on artificial intelligence infrastructure over six years, a sum that exceeds the capacity of any single debt market to finance.
Goldman Sachs equity research published a report Monday projecting that Microsoft, Amazon, Meta, Google and Oracle will collectively invest $5.8 trillion in AI capital expenditures between fiscal years 2025 and 2030. The scale of the buildout — equivalent to roughly 20 percent of current US gross domestic product — will require funding from public bonds, bank loans, private capital and project financing, the bank said, because the dollar investment-grade bond market alone cannot absorb that volume of issuance without disruption.
"The sources of funding are a little bit less. The uses of funds are a lot more. And the only way you get this to equal without a recession or anything awful is higher yields," Mohamed El-Erian, Allianz chief economic adviser, said in a CNBC interview Monday. He pointed to Amazon's weak bond issuance last week as an early warning sign: investors had to sell other holdings to buy the deal, and even then demand was lackluster. The 10-year Treasury yield stood at 4.54 percent on July 9, in the 95th percentile of its 12-month range, while the 20-year yield climbed to 5.08 percent.
The $5.8 trillion forecast covers the current spending trajectory across the five hyperscalers, whose combined capital expenditure is already running at an unprecedented pace. Amazon has guided to roughly $200 billion in 2026 capex for AI infrastructure, custom silicon, robotics and its Leo satellite constellation, with long-term debt rising to $119.1 billion from $65.6 billion. Microsoft spent $38 billion on capex in a single recent quarter, up 63 percent from a year earlier, while Alphabet's first-quarter free cash flow fell 47 percent year over year to $10.12 billion as its buildout accelerated. Meta raised its 2026 capex guidance to $125 billion to $145 billion, issued $55 billion in debt over six months and halted share buybacks. Oracle has already crossed into negative free cash flow for fiscal 2026, with its CFO telling analysts the company expects to raise approximately $40 billion in debt and equity in fiscal 2027.
The bond market is already signaling strain. The 10-year to 2-year Treasury spread has compressed from 0.74 percent in February to 0.35 percent, a flattening consistent with a market absorbing heavy supply at the long end while short rates hold firm. El-Erian noted that traditional Middle Eastern funding sources will be less forthcoming as those economies redirect capital toward domestic reconstruction. The result is a structural mismatch: too many borrowers — governments, technology companies and other corporations — competing for a pool of capital that is not growing fast enough.
The financing challenge has direct implications for which companies can sustain their buildout timelines. Evercore and Bank of America already project 2027 combined hyperscaler capital expenditure will clear $1 trillion, extending the pressure on debt markets. Companies with weaker balance sheets or less predictable cash flows may face higher borrowing costs or be forced to slow their spending plans. Nvidia, whose data center revenue reached $75.25 billion in its most recent quarter and which holds $119 billion in total supply-related commitments, sits at the center of the spending cycle — but its customers are the ones who must finance the infrastructure to run its chips.
For investors, the Goldman Sachs report crystallizes a tension that has been building through 2026: the AI infrastructure cycle is the largest concentrated technology buildout in history, but its financing depends on debt markets that are showing signs of capacity constraints. Microsoft shares are down 20 percent year to date through July 10, while Amazon has gained 6.3 percent and Nvidia 13.3 percent — a dispersion that reflects differing views on which companies can execute their buildout without destroying shareholder value. The companies that can access diverse funding sources at favorable rates will have a structural advantage over those that must rely on a single, increasingly crowded bond market.
This article is for informational purposes only and does not constitute investment advice.
