Contrary to the AI growth narrative, immense CapEx is transforming 'cap-light' tech giants into capital-intensive businesses. This spending pressures margins, reduces returns on capital, and mirrors historical capital cycles where infrastructure builders rarely reaped the primary rewards.
Inspired by baseball's 'Wins Above Replacement' (WAR) metric, M&A should be evaluated not against doing nothing, but against a 'replacement-level' use of capital, such as a share buyback. A buyback is a readily available, low-risk alternative that most acquisitions fail to clear as a comparable benchmark.
The apparent spike in median home prices is a statistical artifact. Owners with ultra-low mortgage rates are not selling, so transactions are skewed toward higher-priced homes, artificially raising the median. This obscures significant pent-up demand that could be unleashed if rates fall.
A recession could perversely benefit the housing market. An economic crisis would likely force the Fed to lower rates and restart QE, making mortgages affordable again. This would unlock huge pent-up demand from sidelined buyers, making well-positioned construction companies a unique recession hedge.
The massive, redundant CapEx in AI infrastructure is analogous to the late-90s fiber-optic boom. While that fiber enabled future giants like Netflix, the initial investors went bankrupt. This suggests the ultimate beneficiaries of AI may be society and end-users, not the companies spending trillions on the build-out.
To illustrate Buffett's success: an investor who bought the S&P 500 at its absolute bottom in 1932 would have been outperformed by someone who simply held cash for 32.5 years and then bought Berkshire Hathaway in 1965. His compounding ability dwarfed even perfect market timing.
Current S&P 500 valuations, with near-record profit margins and a 26x multiple, make historical 10.5% annual returns mathematically improbable. Achieving this would require absurd P/E expansion to 43x or margin expansion to over 20%, suggesting a best-case scenario of only 5% annual returns.
The idea of an infinite holding period is a myth, even for great companies. After Buffett bought Coca-Cola, it eventually traded at 58x earnings in 1998. By not selling, Berkshire endured a meager 4.5% annual return for the next 27 years, proving that even great businesses become sells at exorbitant prices.