Opinion: The concentration of AI stocks has reached a level reminiscent of the eve of the "Tech Bubble" burst, bearing striking resemblance to several classic historical cases.

April 29 data from Barchart shows current AI stock concentration has hit 41%—matching historical bubble peaks like the Nifty Fifty, Japan’s stock market, and the dot-com bubble’s pre-collapse high. A chart combining Bank of America Global Investment Strategy and Bloomberg data reveals the S&P 500’s weight in the top 10 AI companies has surged to 41%, mirroring classic market bubble cases in history. The most recent parallel: the 1995–2000 dot-com bubble, where tech stocks dominated the NASDAQ. The index jumped from ~750 in 1995 to a March 2000 peak of 5048 (a ~600% gain). Tech/internet-related weights in the S&P 500 rose sharply, with the market concentrated in a handful of high-growth tech giants. Many firms hit extreme P/E ratios—at one point, the NASDAQ’s overall P/E topped 200x. Post-March 2000, triggers like rising interest rates, missed earnings, and heavy insider selling sparked a collapse. The NASDAQ plunged 78–80% from its peak to ~1139 in October 2002, erasing all bubble-era gains. Dozens of well-known dot-coms (e.g., eToys, Kozmo) went bankrupt or nearly collapsed to zero. The tech sector then suffered deep losses: the NASDAQ didn’t reclaim its 2000 peak until 2015. Market losses are estimated at over $5 trillion. A handful of quality firms (Amazon, Cisco, for example) survived long-term and rebounded, ultimately validating the long-term value of some “narratives”—though valuations took years to digest.