Investors Reassess AI Exposure as Selloff Hits Wall Street's Favorite Trade
Key Takeaways
- •Portfolio managers are auditing their holdings to distinguish obvious AI positions from seemingly unrelated investments that depend on the sector's spending and growth.
- •At least $25.3 trillion is invested in the largest global companies with direct exposure to artificial intelligence.
- •The MSCI AI-related physical infrastructure index fell 22% between September 30 and November 6.
- •Hedge funds are deleveraging AI risk at the fastest pace in a decade, typically by selling assets to pay down borrowed money tied to their positions.
- •Nomura Asset Management cautioned that when trillions of dollars are concentrated in the same companies, simultaneous trims of just 1-2% by many investors could amount to a rout.

For a stretch, artificial intelligence was the trade that seemed unable to lose. Now, with a selloff underway in AI-related stocks, portfolio managers are combing through their holdings to work out how much of the sector they actually own — and how much of it is hiding in positions that only look unrelated.
The reassessment spans both direct and indirect exposure to AI. The stakes are substantial: at least $25.3 trillion sits in the largest global companies with direct AI exposure.
From Conviction Trade to Forensic Exercise
What was once a conviction trade has become a forensic exercise. Investors are attempting to separate the obvious AI holdings from positions that only appear unrelated.
Direct exposure covers the companies building or selling AI products. Indirect exposure covers everything that depends on those companies continuing to spend, grow, or attract capital.
The numbers underline the urgency. The MSCI AI-related physical infrastructure index fell 22% between September 30 and November 6 — a decline of that scale in roughly five weeks.
Why Small Trims Could Cause Big Moves
Andy Goldberg of Nomura Asset Management has warned that a portfolio trim of just 1–2% could resemble a "rout."
The issue is one of scale and crowding. When trillions of dollars are concentrated in the same group of companies, a small percentage cut made by many investors at the same time adds up to a very large volume of selling.
Nomura Asset Management had flagged the risks of AI concentration months before the current selloff.
Hedge funds appear to be acting on the same concern. The pace of hedge-fund deleveraging of AI risk is running at its fastest level in a decade.
Deleveraging means funds are paying down borrowed money tied to their positions, usually by selling assets. When it happens quickly, it tends to amplify price swings rather than calm them.
What This Means for Investors and Markets
Investors are discovering that sector labels do a poor job of capturing AI risk, since companies with no obvious AI branding may still rely heavily on the sector's continued spending and growth.
For individual investors, the lesson is practical. Checking a portfolio for AI exposure means looking past fund names and ticker symbols to the businesses that depend on continued AI investment.
The fastest hedge-fund deleveraging in a decade also signals that professional money is reducing risk rather than adding to it.
With the audit still underway, the key things to watch are whether the MSCI infrastructure index stabilizes after its 22% slide, whether hedge-fund deleveraging slows, and whether institutional rebalancing remains modest or accelerates.
This article is based on reporting by CryptoBriefing, edited by Diego Almada Lopez.