Key Takeaways:
Five hyperscalers and two chipmakers have pledged more than $3.1 trillion in off-balance-sheet support to finance the AI buildout — money that is not showing up on their balance sheets.
Five hyperscale cloud providers and chipmakers Nvidia and Broadcom have committed more than $3.1 trillion in off-balance-sheet financial support to fund AI infrastructure, according to Morgan Stanley — a figure that exceeds the roughly $1 trillion in cash flow the companies expect to generate in 2027.
"Alphabet's commitment is the largest at $890 billion, largely procurement guarantees," Todd Castagno, the Morgan Stanley analyst who led the report, said.
The five hyperscalers — Alphabet, Amazon, Meta, Microsoft and Oracle — are on track to spend more than $1.2 trillion in cash capital expenditure in 2027 against roughly $1 trillion in expected cash flow, the bank estimated. More than 40 percent of hyperscaler sales revenue is being reinvested into AI capex. Alphabet and Amazon both turned free cash flow negative in the second quarter, and Meta is expected to follow this quarter.
The gap between capex and cash flow is being bridged with debt, guarantees and vendor financing rather than retained earnings — a structure that behaves differently in a downturn and concentrates risk across equities, credit and emerging markets tied to the AI supply chain.
The $3.1 trillion figure covers guarantees, leases and other forms of financial support that keep debt off balance sheets. The Bank for International Settlements has warned that some AI financing structures may mask leverage by moving it off balance sheet — "leverage does not disappear by being out of sight," the BIS noted.
Circular financing is the most visible mechanism. Nvidia, the largest beneficiary of the buildout, is also financing its own customers. The company is in talks to provide a financial backstop of up to $250 billion so OpenAI can lease computing capacity from a planned 10-gigawatt data center campus in Ohio being developed by SoftBank's energy subsidiary. Nvidia is separately discussing financing as much as $350 billion of OpenAI's chip purchases, on top of the $30 billion it has already invested in the company. A newly unveiled partnership with SK Group, parent of memory chipmaker SK Hynix, is worth more than $500 billion in business between the two sides.
Google has agreed to backstop lease payments at five data center locations for Anthropic, helping the OpenAI rival obtain what amounts to a $35 billion loan.
The pattern echoes dot-com era vendor financing, when telecom and tech companies extended credit to their own customers to inflate demand figures that looked strong right up until they were not.
The financing is not merely a balance-sheet abstraction — it is showing up in operating results. Alphabet, long viewed as a money-printing machine, turned free cash flow negative in the second quarter of 2026 for the first time, while its long-term debt more than doubled to $98 billion over the first half of the year. Amazon's long-term debt jumped 81 percent to $119 billion in the first quarter alone.
The Magnificent 7's combined free cash flow has fallen to 7.9 percent of quarterly sales, its lowest level since early 2024, as AI-driven capex across the group surged 75 percent year on year to $136.6 billion in a single quarter. Alphabet and Meta have halted share buybacks, while Apple and Microsoft have scaled theirs back, as the group shifts from an asset-light model to a capital-intensive one.
Morgan Stanley now estimates the five largest hyperscalers will spend roughly $805 billion on capex in 2026, up from $261 billion in 2024, with projections of $1.1 trillion for 2027.
The concentration risk extends beyond the seven companies funding the buildout. South Korea and Taiwan together now make up close to 50 percent of the emerging markets index, and both economies are deeply tied to AI semiconductor production. A portfolio holding US equities, US credit and EM equities — diversified on paper — may actually be making one concentrated bet three different ways.
Whether the technology succeeds and whether today's capital gets repaid are two separate questions. History suggests it is entirely possible for the answer to be yes to the first and no to the second, with the eventual winners being adjacent suppliers and disciplined capital allocators, not necessarily the companies and financiers funding the buildout at today's prices.
This article is for informational purposes only and does not constitute investment advice.
