On July 10, BCA Research’s investment strategist said that the current AI investment cycle has entered “extra time,” with the AI bubble showing up more on the earnings side than on the valuation side. Profit growth, pressure arising from changes in demand structure, and pressures driven by supply expansion may become concentrated around 2027; at that time, the growth rate of AI capital expenditures may slow down significantly.

Investors may focus on four bubble warning indicators: GPU rental prices, AI storage chip prices, AI application adoption rates and enterprise spending, as well as Token prices and downloads of AI programming agents. Among them, a decline in Token prices and a slowdown in AI programming tool download volumes are the most worth watching early signals, reflecting that companies are gradually shifting from chasing the most advanced models to cost control.

If the AI bubble ultimately bursts, its impact on the U.S. economy and capital markets could exceed the 2000 internet bubble. Based on estimates, the U.S. stock market may see a 30% to 50% pullback at that time, and it would drag down U.S. consumption and economic growth through the wealth effect. #预测世界杯西班牙VS比利时