高盛称半导体多头与软件空头的AI动量交易仍处历史高位

The leveraged trading community has engineered a historic inverse trade that Goldman Sachs's Tony Pasquariello on Tuesday characterized as a "photographic negative."
Prime brokerage data spanning a full decade reveals semiconductor (SMH) long exposure and software (IGV) short exposure tracking each other in near-perfect opposition—when one rises, the other falls, Pasquariello wrote in a note. This structural divergence intensified dramatically in the second half of last year as investors pressed hard on the AI (AIQ) enablers-versus-disrupted narrative, aggressively building long positions in hardware makers while shorting software names perceived as vulnerable.
The fever broke in recent months through a combination of July's deleveraging event, which flushed out crowded positions, and a fundamental reassessment of software that has evolved into "a more 2-sided debate," Pasquariello wrote,
Yet the structural wedge between these sectors remains historically wide. The semiconductor-software split went so far as defining the extreme edges of the broader momentum trade itself.
As actual momentum factor performance trended upward over the past decade, most pronounced in recent years, clients "steadily dialed up their leverage to it," the note said. Net factor exposure to medium-term momentum has climbed to levels rarely observed in the ten-year dataset.
Looking ahead, Pasquariello projects that elevated positioning will persist with a significant caveat: "As long as the AI buildout remains on track, client exposure to this factor will run on the structurally higher end of history, but with that comes a more volatile path."