Hedge funds boost tech bets to record highs on AI optimism: Goldman Sachs

Hedge funds bought technology stocks at the fastest pace in nearly three months, with semiconductor and chipmakers drawing the most interest and global IT bets reaching record highs since 2016. Buying was led by North America and Asia emerging markets, while funds sold communications equipment and IT services. The positioning suggests continued enthusiasm for AI-linked names despite broader geopolitical and macro pressure.

Analysis

This is less a simple “tech rally” than a crowdedness reset in the most liquid growth factor. When hedge funds add to semis/software while cutting hardware and IT services, they are effectively expressing a narrow view of who monetizes AI: owners of scarce compute, IP, and pricing power should continue to outperform the lower-margin implementers that are more exposed to budget scrutiny. The second-order effect is that supplier bottlenecks and procurement discipline become more important than top-line AI enthusiasm; names that cannot prove incremental margin expansion may lag even if end-demand stays healthy.

Positioning is now the bigger story than fundamentals in the near term. Record-long exposure means incremental inflows can still push the group higher, but it also means the trade is increasingly vulnerable to any macro shock that forces de-grossing, especially if rates back up or geopolitics widen. The fastest reversal risk is not a single earnings miss; it is a broad factor unwind that hits the most crowded semis and mega-cap software names first, with the higher-beta AI basket likely moving 2-3x the index on any tape break.

The contrarian read is that the market may be late-cycle chasing the same AI winners, while underpricing how much good news is already in the tape. If semis continue to rerate without a corresponding increase in deployment ROI, capex discipline at hyperscalers could eventually slow orders, and that would hit the entire ecosystem with a 1-2 quarter lag. In other words, the trade works until the market stops rewarding narrative and starts demanding monetization cadence; that transition is where crowded longs get punished most violently.

Key Decisions for Investors

• Fade the crowded basket tactically: short SMH or SOXX via 2-4 week put spreads into strength; risk/reward favors a 1:2 premium-defined structure because positioning is extended and the unwind can be fast.
• Pair long selected AI infrastructure winners vs short IT services/hardware enablers: e.g., long NVDA/AVGO, short an equal-dollar basket of lower-margin services names; aim for 1-3 month relative outperformance as AI capex concentration persists.
• Add a hedge against factor de-grossing: buy QQQ downside via 1-2 month put spreads financed partially by covered calls on existing tech longs; this protects against a crowded-position flush without fully exiting the theme.
• For a cleaner contrarian entry, wait for a 5-7% pullback in semis before adding exposure; if the drawdown is flow-driven rather than fundamental, the next leg up can still deliver 15-20% over 3-6 months.
• If owning GS as a read-through on prime-brokerage flow, keep it as a small satellite long rather than a core; the benefit is durable but the upside is capped unless tech breadth broadens beyond the current narrow winners.