Net Cash Has Become Debt

Calculations by the Flossbach von Storch Research Institute, based on regulatory filings with the SEC, describe a spectacular reversal. The contractual commitments of the five major US AI groups, excluding leases (software, infrastructure, energy), rose from $129 billion to $1.6 trillion. In total, $2.7 trillion in firm commitments do not yet appear on balance sheets.

The aggregate financial position has flipped: $209 billion in net cash has given way to $106 billion in adjusted net debt. Oracle alone carries $88 billion of it.

The Mechanism: Spending Faster Than You Collect

The causal chain is simple. When infrastructure spending outpaces internal cash flows, companies sign leases and multi-year contracts, then turn to credit. Liquidity reserves shrink and external financing becomes a necessity.

The $63 billion package announced by Bank of America, Blackstone, Citi and Morgan Stanley for computing capacity, in partnership with Broadcom, is one illustration. The risk does not disappear: it shifts to bank balance sheets and investors.

A Market That Demands Proof

Market reactions show selectivity. Alphabet fell after disappointing sales in most of its core businesses, as investors question the hundreds of billions it is devoting to AI.

By contrast, SK Hynix gained more than 13% in New York. The stock traded at four times expected 2027 earnings. Its revenue beat expectations by about 6%, with a gross margin of 85.7%. Shipment growth should reach the middle of the 20% range this quarter, up from 10% in the third, and production is said to be sold out through 2029. Micron (nearly 6%), Arm (4%) and AMD (3%) followed.

Equinix even sees an advantage: if projects are cancelled, existing capacity becomes scarcer, which could give it room to raise prices.

The Limits of Interpretation

Caution, however. A firm commitment is not a due debt, and these figures say nothing about the revenue that will finance them. The gains and losses cited are single-session moves, not an established trend. As for the day's indices, their displayed variations are inconsistent and could not be used.

A Proportionate Conclusion

AI financing increasingly depends on credit and on commitments that are barely visible. In this context, markets value proof of profitability rather than the scale of spending. That is not enough to conclude there is systemic risk, but financial discipline is becoming a central criterion for investors.

This article was written with the assistance of artificial intelligence and verified by our editorial team.