Episode Details
Back to Episodes
No, AI Isn't a Bubble: Hyperion's Case, Built on Margins
Description
Rob Pizzichetta (Mont Wealth) interviews Hyperion Asset Management investment director Jolon Knight about Hyperion’s white paper on the likely long-term economics of AI, focusing on returns above cost of capital, fast payback periods, durable returns, and rising demand as intelligence costs fall. Knight cites strong reporting-season acceleration across key holdings, highlights expanding demand for agentic AI and inference driving a cloud-compute backlog estimated above $2 trillion, and outlines hyperscaler CapEx forecasts rising from about $800 billion to $1.2–$1.4 trillion. They discuss payback cycles on silicon (often under three years), bottlenecks (chips, land, power, materials), market valuation headwinds, and Hyperion’s forecast ~24–25% EPS growth versus the broader market’s ~8–9%. Portfolio updates include SpaceX, TSMC, and Dutch Bros.
00:00 Hyperion Growth Outlook
00:38 Show Intro And AI Paper
01:39 Reporting Season AI Surge
03:07 Agentic AI Demand Boom
04:42 Compute Backlog And Capex
07:15 Payback Cycles Explained
09:43 Bottlenecks And Regulation
11:41 Why Markets Misprice AI
13:24 Hyperion Returns Framework
15:23 Cashflow And Debt Concerns
17:06 Eight Trillion TAM Thesis
18:18 SpaceX Investment Case
20:13 Portfolio Changes Recently
21:27 Calls To Slow AI Down
22:41 Wrap Up And Key Takeaways
#income #bonds #wealth #investing #money #howto #financialfreedom #advice #AI