
In an interview on Excess Returns, Kai Wu from Sparkline Capital discusses his research on the AI CapEx boom, drawing parallels to historical infrastructure booms like railroads and the dot-com era. He highlights the significant acceleration of capital expenditure by major tech companies (Mag 7) into AI infrastructure, estimating trillions of dollars in investment over the next five years. Wu raises concerns about whether the potential returns will justify this massive spend, the concentration of market gains in a few AI-linked stocks, and the historical underperformance of companies with high asset growth. The conversation explores the shift of Mag 7 from asset-light to asset-heavy models, the implications of circular deals, and the prisoner's dilemma dynamics in the AI arms race. Wu suggests investors consider early adopters of AI across various sectors and geographies, rather than focusing solely on infrastructure players, and advocates for a value-based investment approach that accounts for intangible assets to navigate the hype and potential valuation risks.
Sign in to continue reading, translating and more.
Continue