AI hardware build-out mirrors Cisco vendor-financing bubble
The bearish thesis for the current AI hardware trade argues that chip demand is artificially manufactured through vendor financing, mirroring the late-1990s Cisco dot-com bubble.
The argument
Michael Green argued that major AI players are financing their own customers' acquisitions through debt guarantees or direct equity investments. Once these customers fail to generate underlying profits from their high-cost infrastructure, this closed-loop demand mechanism will run out, exposing overstated profit margins.
The thesis, stress-tested
✓ What validates it
- ✓Secondary market sales of AI chips by cash-strapped customers swamp primary production
- ✓A decline in capital expenditure guidance from major cloud service providers
▸ Risks discussed
- ▸End-user demand for AI applications scales rapidly to profitability
- ▸Inflow of retail 401(k) capital continues to support high valuations indefinitely
Hear it yourself
"When we talk about the closed loop with AI, what we're really referring to is the financing mechanism that's allowing the acquisition of customers, either through guaranteeing debt contracts or by lending the money directly in the form of equity investments, et cetera."
00:00 / 00:18
AFFILIATE LINK · ZORTIX MAY EARN A COMMISSION · NEVER A RECOMMENDATION TO TRADE