1. Home
  2. Analysis
  3. Macro
Macro

Fed Uncovers $3T in Off-Balance-Sheet Commitments Behind Big Tech's AI Boom

The Fed's Financial Stability Report reveals $3T in off-balance-sheet commitments—including $1T+ in uncommenced data center leases—across Big Tech, raising fresh risks for AI valuations.

BMInsider desk · · 3 min read
A wide low-angle photograph of an expansive hyperscale data center construction site in rural Virginia on a cold November morning, showing steel framing and concrete foundations flanked by massive industrial cooling units and high-voltage transmission lines under an overcast autumn sky.
Foto: Hay Kranen / Wikimedia Commons, CC BY 4.0

The Federal Reserve's November 2026 Financial Stability Report identified more than $3 trillion in off-balance-sheet commitments financing the artificial intelligence boom across major technology firms, according to an in-depth review reported by Fortune on Tuesday. For equity investors, the revelation exposes a massive layer of hidden corporate liabilities that could severely compress tech multiples if returns on generative AI investments fail to accelerate.

What happened

In a dedicated focus section (Box 2.2) titled 'Leverage in the Data Center Buildout,' the Fed warned that major technology leaders—including hyperscalers Microsoft, Amazon, Alphabet, Meta, and Oracle, alongside chipmakers Broadcom and Nvidia—have accumulated over $3 trillion in nominal off-balance-sheet obligations. More than $1 trillion of this sum consists of uncommenced data center leases, contracts where construction is underway but rent has not yet begun.

Another $2 trillion spans take-or-pay energy contracts, long-term compute capacity purchase pacts, and residual value guarantees. While traditional on-balance-sheet debt for these seven tech leaders is projected to top $800 billion by year-end, their off-balance-sheet obligations have expanded at more than triple that pace.

The report rattled tech equities on Tuesday, November 17, 2026. After a 2.5% tech sell-off on Monday, mega-cap cloud and hardware providers faced renewed pressure as trading desks absorbed the extent of non-bank credit financing the buildout.

Why it matters

By structuring data center buildouts through uncommenced leases and special-purpose developer entities, hyperscalers have kept direct liabilities off their corporate balance sheets. Third-party developers use these long-term commitments as collateral to obtain construction loans from commercial banks and private credit funds, shifting development risks while preserving tech credit ratings.

However, standard tools for reading the balance sheet understate the true financial leverage supporting the sector. U.S. bank holding companies now hold nearly $1 trillion in direct and indirect exposure to AI developers and specialized tech entities.

If enterprise AI adoption slows or compute capacity experiences oversupply, hyperscalers face rigid long-term cash drains from take-or-pay terms. As central bankers note, circular vendor financing and massive fixed obligations leave little margin for error at current equity valuations.

What to watch next

  • Nvidia Q3 earnings (Tuesday after the close): Management guidance on customer infrastructure capex and hyperscaler financing commitments.
  • Enterprise AI monetization metrics: Q4 software spend conversion rates and hyperscaler cloud margin trajectories.
  • Regulatory disclosures: Potential SEC and FASB moves toward stricter reporting rules for forward data center commitments.
  • Credit spread monitoring: Borrowing costs for private data center developers and specialized infrastructure funds.

Stocks in this analysis

AI stock analysis

Free on aktienanalyse.ai: score profile, ratios versus the industry and business model. Plus, if you want it, the full analysis with fair value and bull, base and bear case. Prices and ratios as of Nov 16, 2026.

Sources (2)
  1. Fortune: The Fed's first close look at the AI buildout finds $3 trillion of AI commitments that aren't on big tech's balance sheet as debt
  2. Federal Reserve: Financial Stability Report - November 2026

This analysis is for information only and is not investment advice. Prices and figures as of publication.

AI stock analysis

Any stock analysed in minutes

Business model, balance sheet, valuation and scenarios: aktienanalyse.ai builds a thorough analysis of any stock in minutes, with a fair value and bull, base and bear cases.

Analyse a stock → Stock profiles