This was the week the rotation became undeniable. Monday the Russell 2000 surged 2.12% and closed above 3,000 for the first time in history, even as the S&P slipped 0.37% and the Nasdaq fell 1.32% under mega-cap weakness; Alphabet dropped 5% on AI-talent-departure concerns while Amazon, Meta, and Microsoft all fell. The small-cap index has now climbed more than 21% in 2026, lapping both the S&P and the Nasdaq. The mega-cap drag deepened all week. SpaceX, which had rallied to nearly 50% above its $135 IPO price in its first three sessions, gave it all back, falling 16% Monday alone and continuing lower. Midweek a New York Times report that OpenAI was weighing delaying its IPO to 2027, after watching SpaceX trade poorly, hit chip and AI-infrastructure names directly; the concern was the sustainability of AI capital spending if the IPO funding pipeline stalls. The Nasdaq fell 4.6% for the week and logged five straight losing sessions into Friday. Friday PCE inflation printed 4.1% year over year, the hottest in nearly three years, though the monthly increase was smaller than expected. Crude fell roughly 5% as Strait of Hormuz tanker traffic flowed largely unimpeded. The S&P ended the week down nearly 2%, the Dow outperformed at plus 0.6%, and gold rose toward $4,090.
The Russell crossing 3,000 is the headline this desk has waited months for, because it is breadth made literal. For most of this cycle the gains lived in a handful of mega-cap names; small caps leading while the Nasdaq sheds 4.6% in the same week is the rotation thesis confirmed in the tape, not the commentary. The driver is mechanical as much as sentiment: when the AI trade wobbles, the marginal dollar looks for somewhere with cheaper multiples and less issuance overhang, and that is exactly the down-cap universe. The OpenAI IPO blink matters for the same supply reason this desk flagged when SpaceX filed: the mega-IPO pipeline that was draining the marginal equity dollar may be slowing, and a slower drain is a tailwind for everything that trades below it.
The caution rides alongside the opportunity. A 4.1% PCE print keeps the hawkish Fed engaged, and small-cap leadership built on rotation can reverse as fast as it arrived; the Russell gave back its record by Friday and closed the week near 2,990. The durable read for micro-cap allocators is not to chase the index but to own the names that benefit from rotation without depending on it: real businesses, real cash, no refinancing clock. When breadth broadens, quality down-cap names get re-rated; when it narrows again, the same names hold because the business was never the question.
Coverage ledger: BTBT sits on the AI-infrastructure thread that the OpenAI IPO delay put under scrutiny, the name to watch as capital-spending sustainability gets repriced. CLPS trades at a discount the rotation can close. GLMD and ACON hold the cash floors that a 4.1% PCE and a hawkish Fed keep rewarding. Full dossiers in the research library.