BIS Warns AI Capital Expenditure Collapse Would Transmit to Credit Markets and Output
BIS identifies AI investment reversal as systemic risk channel to credit and growth. Report links concentrated capex, bond issuance, and lending standards without assuming continued expansion. Analysis shows states face trade-off between productivity claims and financial stability buffers.
The BIS June 2026 assessment identifies three linked threats: an AI bust, persistent inflation, and fiscal strain. It notes that AI-related capital expenditure has reached levels comparable to prior technology cycles, with concentrated exposure in a small number of US firms whose debt and equity financing now influence broader lending standards.
Primary records show the warning appears in the BIS Annual Economic Report chapter on financial vulnerabilities. Data cited include the rise in AI-linked corporate bond issuance since 2023 and the correlation between technology equity drawdowns and subsequent tightening in bank lending surveys across G10 economies.
State incentives explain the pattern. Export-oriented economies have tolerated elevated valuations to sustain productivity narratives, while central banks have subordinated macroprudential tightening to growth targets. A contraction would force simultaneous re-pricing of both equity and credit risk, an outcome the BIS records as historically rare outside synchronized downturns.
Next steps hinge on whether supervisors incorporate AI-exposure metrics into stress tests scheduled for 2027. Absent that adjustment, credit allocation will continue reflecting current valuations until a price signal arrives.
BIS: G10 bank lending standards tighten by at least one net percentage point in the 2027 stress-test round if AI-related capex contracts 25 percent year-over-year.
Sources (2)
- [1]Primary Source(https://www.bis.org/publ/arpdf/ar2026e.htm)
- [2]Supporting Source(https://www.bis.org/publ/qtrpdf/r_qt2606a.htm)