Peter Howitt 1.7 8 ideas

Howitt notes that while industrial concentration suppresses some innovation, the current AI boom requires massive capital expenditure and "flexible financial systems" to tolerate losses. He highlights that "Googles and Microsofts" are the ones "going at it" alongside new entrants they often back (OpenAI/Anthropic). In a "Superstar Market," the most productive firms with the deepest capital moats capture the majority of the value. AI is a capital-intensive game where incumbents with massive R&D budgets can endure the "fumbling around" phase of technology adoption that bankrupts smaller players. Long the "Superstar" incumbents who are effectively privatizing the infrastructure of the AI economy. Regulatory breakup (Howitt explicitly mentions antitrust needs to reorient toward preventing innovation suppression) or a "Schiller PE" valuation reset.

Howitt highlights that as manufacturing jobs disappear (even in China), labor moves to services. He explicitly identifies "elderly care" as a sector with "tremendous potential" due to aging populations in North America, Europe, and China. This is a "Second-Order" AI trade. AI will handle menial data entry (as seen in the Gates/Rwanda example), making human care workers more efficient. The demand is demographic (inevitable), and the supply constraint (labor) is eased by AI productivity tools, improving margins for care facility operators. Long Senior Living (BKD) and Hospital operators (HCA) as beneficiaries of demographic tailwinds + AI efficiency. Government reimbursement rate changes or labor shortages outpacing AI productivity gains.

Howitt observes that the Shiller PE (CAPE) ratio is currently at levels seen only during the 2000 Dotcom bubble and 1929. High CAPE ratios do not predict an immediate crash, but they are statistically significant predictors of low real returns over the next 10–20 years. The "AI Boom" is real, but the price paid for that growth is historically expensive. Exercise caution with broad passive indexing; returns will likely be compressed compared to the last decade. "Irrational exuberance" can persist longer than solvency (the market could melt up another 50% before correcting).

Howitt uses Tesla as a case study for the transience of monopolies. Tesla had ~90% market share in EVs, but that lead "only lasted a few years" before being overtaken by Chinese competitors. This illustrates the "Creative Destruction" cycle speed. Being a first-mover (like Tesla in EVs or potentially Nvidia in chips) does not guarantee a permanent monopoly. Investors should watch for rapid market share erosion in dominant tech leaders. Watch for signs of commoditization in EV/AI hardware similar to what happened to Tesla's EV dominance. Tesla re-innovating (Robotaxi/Optimus) to reset the cycle.