TS Lombard and Gartner warn AI bust could wipe out $20T; $3.5T infra at impairment risk
SumitGup · x · 2026-10-25
Two major research firms are warning about the fallout if the AI spending boom collapses.
TS Lombard (Dario Perkins) estimates a sharp AI spending bust could trigger:
- S&P 500 down 30%, tech stocks down 40%
- $125–190B in credit losses
- 800,000 jobs lost
- Over $20 trillion in stock market value wiped out
Gartner goes further on AI infrastructure: companies are expected to spend nearly $1T on infrastructure in 2026 alone, plus another $6T between 2027 and 2030. Data center owners would need to generate $8.5T in revenue through 2030 to cover spending, financing costs and returns. Per Gartner's official release (image), more than $3.5T of AI infrastructure spending could be exposed to impairment by end of 2030 if capex continues at the current pace.
More from Venture
- AI labs are flipping GPU contracts: H100/B200 rental prices up 2-3x since early lock-ins — Yuchenj_UW · 2026-10-25
- Gary Marcus: bear case for AI strengthens as S&P 493 shows zero productivity gains — GaryMarcus · 2026-10-25
- Musk: Agentic AI demand is growing faster than anything he's ever seen — elonmusk · 2026-10-25
- Owning scarce physical assets plus AI beats building models, argues investor — karinanguyen · 2026-10-25
- Top 10% of firms account for 99.5% of AI inference spend, Apollo data shows — SumitGup · 2026-10-25
- Anthropic revenue hits $16.3B in H1 2026, a 14.8x jump in a year — PeterDiamandis · 2026-10-25