More striking still is what the report describes as a web of “circular” financing arrangements concentrated within a small number of AI firms. Since 2023, 14 publicly disclosed deals channeled approximately $300 billion of equity, led by hyperscalers, into financially weaker AI firms, alongside $650 billion of debt. In exchange, those investors secured $1.4 trillion of “future revenue commitments” through services or chip orders, according to the report.
Jason Wu, assistant director of the IMF’s monetary and capital markets department, explained the risk plainly: if AI commercialization falls short of expectations, the losses could arrive simultaneously – equity investments depleted, debt hard to service, and revenue commitments failing to materialize.
“If there are questions about the monetization in AI,” Wu said at the October 13 press briefing, “the equity investments will deplete in value, the debt may be difficult to service, and the revenue commitments in the future may not materialize.” He called it a “triple whammy.” Concentration in a small number of nodes means that stress in any one entity could reverberate quickly.
Hedge funds are the amplifier
Part of the IMF report, focusing specifically on hedge funds, connects directly to the yield risk. Hedge funds’ share of outstanding US Treasury securities rose from approximately 4 percent in 2022 to 9 percent in 2025, according to the report – a shift largely driven by the cash-futures basis trade, a leveraged arbitrage strategy that exploits small price differentials between Treasury bonds and futures contracts.
That footprint, the IMF warns, leaves the Treasury market “susceptible to a disorderly unwind.” The fund’s researchers – Andrea Deghi, Taneli Mäkinen, Mahvash S. Qureshi, and Felix Suntheim of the IMF’s Monetary and Capital Markets Department – find that the sector’s rising leverage and increasingly overlapping positions mean that multiple funds hold the same trades simultaneously.