U.S. Hyperscalers Increase AI Spending Amid Bubble Warnings

U.S. hyperscalers have entered an AI capital expenditure arms race, with spending expectations for 2026 and 2027 continuing to rise. Data from Goldman Sachs Group, Inc. indicates that these companies are increasingly relying on debt to fund AI infrastructure as their cash-to-capex spending ratio declines.

This financial trend mirrors the late 1990s dotcom boom, where a widening gap between available cash and capital spending preceded a market bust. Current U.S. stock market concentration levels are now comparable to those seen at the peak of previous bubbles, including the Japanese bubble of the late 1980s.

While analysts remain optimistic about operating margins for the S&P 493, the combination of high P/E multiples and extreme market concentration suggests a poor 10-year outlook for equity returns.