Artificial intelligence stocks are experiencing dangerous volatility as crowded positioning and leverage create a treacherous landscape reminiscent of the dot-com bubble's final months. Goldman Sachs partner Bobby Molavi warns that co-correlated strategies feel great on the way up but pose extreme risks during unwinds, with every investor type effectively getting longer AI exposure daily.

Leveraged exchange-traded funds totaling $200 billion in assets amplify moves through short-gamma effects, triggering roughly $9 billion of directional rebalancing for every 1% index shift. Technology and momentum strategies dominate 85% of these funds, creating powerful feedback loops that Nomura's Charlie McElligott calls a perpetual motion machine. Rising prices and volatility across semiconductors, memory, and networking equipment feed this cycle.

Realized volatility spikes threaten volatility-control funds becoming heavy sellers, with McElligott estimating $21 billion in selling flows from two weeks of 1% S&P 500 moves and $41 billion from 1.5% moves. UBS strategists note hedge funds are trimming crowded positions as confidence shifts, questioning remaining upside versus risk.

The market still believes in AI's long-term potential, but doubts about hyperscaler returns and expensive chipmaker expectations are mounting. With capex upgrades potentially hitting speed limits, crowded positions may face underperformance as investors rotate toward the next trade.