Bank of America warns the AI spending boom is starting to mirror the classic pattern of past capital busts.

Strategist Sebastian Raedler described a sequence of high margins, heavy investment, excess capacity and falling prices.

Expected hyperscaler capex for the next 12 months has tripled from $330 billion to about $1 trillion.

That is roughly 3% of US GDP, above the 1870s railroad boom’s 2.2% and the 1990s tech boom’s 1.1%.

Average AI token prices have reportedly fallen about 50% since summer, partly due to Chinese open-weight models.

Ramp data shows US business AI spending easing since August even as token volumes rise sharply.

Tech giants are increasingly using debt and equity markets to fund AI, adding risk if returns disappoint.

Raedler says a technology can transform industries without delivering attractive returns to every investor.

Read the full article for more details on BofA’s AI spending boom warning.