A Morgan Stanley report has pulled back the curtain on the financing of the artificial-intelligence boom, revealing that Big Tech hyperscalers and Nvidia have accumulated over $640 billion in purchase obligations and committed to another $675 billion in off-balance-sheet lease payments.
"The lack of disclosure and contractual complexity of these arrangements makes it difficult for investors to interpret true economic leverage versus that reported on balance sheet,” a group of Morgan Stanley analysts led by Todd Castagno said in the report. “The circularity of the AI ecosystem further complicates adequate analysis.”
The scale of these commitments is immense, having more than doubled in the past year and increased sixfold over the past five years. The obligations are rising much faster than the leverage reported on balance sheets. For example, Meta’s commitments stand at approximately 1.7 times its forward operating cash flow. As of the latest disclosures, hyperscalers carry $257 billion in lease liabilities on their balance sheets, but have committed to an additional $675 billion for leases that have not yet commenced.
For investors, this introduces a new layer of risk ahead of Big Tech earnings reports. The massive, opaque spending structure complicates analysis for companies like Alphabet, Meta, and Microsoft, which were once prized for their fortress-like balance sheets and high free cash flow. While the spending fuels AI adoption, it also creates a fragile, intertwined financial ecosystem.
This "circular finance" model works as long as the AI boom continues unabated. Hyperscalers like Google commit to renting space in data centers, which allows the data center suppliers to secure loans for construction, backed by the creditworthiness of their Big Tech clients. While this practice is not illegal, the lack of transparency could make investors nervous, especially as these off-balance-sheet bills are highly likely to come due.
The spending surge comes as companies are increasingly detailing the tangible benefits of AI. A separate Morgan Stanley analysis found that one-quarter of S&P 500 companies mentioned quantifiable AI impacts in the first three months of the year, up from 13 percent in the prior year. Tech and finance lead the pack, with firms like Bank of America noting that AI saves them the equivalent of 2,000 coders.
However, the reality remains that 75 percent of firms have yet to show such benefits, and a Goldman Sachs report found only 10 percent of firms noted AI impact in specific use cases. This highlights the gap between the massive capital outlay on AI infrastructure and the current return on that investment. With tech companies now taking on increasing, and often opaque, debt to fund this build-out, investor scrutiny of capital discipline is set to intensify.
This article is for informational purposes only and does not constitute investment advice.
