Goldman Sachs: Leveraged Funds Go Long Semiconductors, Short Software — AI Moat Bets Show "Negative Reversal" Pattern

Nashnova编辑部
Published todayAbout 4 min read

Goldman's chief trading strategist Tony Pasquariello says the long-semis / short-software divergence among leveraged funds has hit a decade-high extreme, meaning any shift in AI profits toward software could trigger a violent unwind.

01

What exactly is this "photo-negative" trade?

Leveraged traders are long semiconductors (SMH) and short software (IGV) at the same time — the two positions mirror each other like a photo and its negative.
This means → their core bet is that AI's moat belongs to the companies that make chips, not the companies that write software.
In plain terms = they believe the shovel-sellers will outperform the gold-diggers — so they buy shovel companies and short the miners.
02

How extreme is the divergence?

Goldman's prime brokerage data — spanning a full decade — shows the long/short split has reached a historic extreme.
This means → in ten years of tracking, leveraged funds have never held such opposite views on semis versus software.
Pasquariello calls the positioning a "photo negative" — where the photo is bright, the negative is dark, and vice versa.
03

Is the extreme positioning itself a risk signal?

The more crowded the trade, the harder the reversal — everyone is on the same side, and the exit is narrow.
Two triggers could flip it: ① AI application-layer profits start flowing to software companies; ② semiconductor supply-chain expectations soften.
This means → if the market suddenly decides "software makers can also cash in on AI," the short-software positions face forced covering, producing sharp volatility.

Content is for reference only, not financial advice.