US equities are running on two irreconcilable stories, and this week's published research makes the split unusually stark.
The bull case, argued across multiple recent research notes, rests on a striking earnings backdrop: what could be a fifteenth straight quarter of double-digit S&P 500 earnings growth, with roughly $3 trillion in projected AI data center spending flowing into semis, infrastructure, and financials. Several analyst camps frame AI-driven productivity as the core engine and see rotation into small caps, financials, healthcare, and industrials as money leaves the most stretched momentum names.
A contrarian thread of recent research pushes back hard. It flags semiconductor stocks up more than 230% in fourteen months — a pace matched only by the final leg of the dot-com bubble — momentum names at a five-sigma extreme, record household equity concentration, and hyperscaler capex dependent on AI returns that haven't been proven. One published view adds that a Fed leaning toward fighting structural inflation rather than easing raises the downside for US risk assets.
The genuine disagreement: on a 1–2 year view, is AI capex productive investment or speculative excess?