In the short run, the market’s AI trade is still the most popular. Since April’s post-selloff rebound, just ten stocks—largely the Mag 7 plus AVGO, ORCL, and PLTR—have driven roughly 80% of S&P 500 gains. That concentration, reinforced by policy tailwinds, has pushed valuations to extremes that rely on uninterrupted capital flows into a handful of names. Hartnett warns the signal for a genuine AI bubble break will be widening tech credit spreads, echoing the 1999 pattern when cash burn fears ended the dot-com melt-up.
‍

Until that stress shows up, momentum could keep running, especially with $2.9 trillion in AI-related capital expenditures projected by 2028. When media headlines shift from celebrating the AI boom to warning that valuations are “getting out of hand,” it’s often the last stage before a sentiment break. For now, both the productivity dream and the bubble risk are in play—a combination that can stretch markets higher in the short term but leaves little room for disappointment during earnings results.