AI infrastructure faces mid-cycle slowdown
The guest argued that the AI infrastructure and semiconductor sectors are entering a three-to-six-month mid-cycle slowdown due to past inventory hoarding and unsustainably high market expectations.
The argument
He suggested that the initial demand shock has passed, and the market will transition to a phase of mixed earnings results, where disappointments will likely match positive surprises.
The thesis, stress-tested
✓ What validates it
- ✓Flat or downward revised guidance from major semiconductor or cooling infrastructure companies in upcoming quarterly reports
▸ Risks discussed
- ▸Hyperscaler capital expenditure accelerating beyond current elevated forecasts
Hear it yourself
"I mean, if it was Bill Gurley or someone, he'd be up there for over an hour. What the guy said, okay. The amount of copper we need for the next ten years is equivalent, and I think he said ten years, but it's ten or fifteen, doesn't really matter for the point, is equivalent to all the copper mined over the last ten thousand years. So when people look at this problem, I'm looking for the parts that they're gonna need to actually accomplish this. Because it might take him twenty years to get all it Elon's been known to say it's gonna happen sooner."
AFFILIATE LINK · ZORTIX MAY EARN A COMMISSION · NEVER A RECOMMENDATION TO TRADE