The entire US stock market rally has a very specific center of gravity, and it answers to the ticker NVDA. Goldman Sachs is now telling clients what many have been quietly thinking: the AI trade is working, but the lack of diversification underneath it is a real problem.
The bank’s latest guidance boils down to a five-word thesis: stay long AI, but buy insurance.
The concentration problem
Nvidia has been the single largest contributor to US index gains. Goldman identifies a familiar cast of characters doing the heavy lifting alongside Nvidia: Microsoft, Alphabet, Meta, and Broadcom. These mega-cap AI names have become the market’s load-bearing walls.
The specific concern centers on what Goldman describes as lower-quality AI-related stocks. These are the companies riding the AI narrative without the earnings power of an Nvidia or a Microsoft. Goldman’s view is that these names are most vulnerable to a momentum unwind. The mega-caps might dip; the lower-quality cohort could face significant drawdowns.
Hedge funds already learned this lesson
Equity hedge funds recently experienced their worst performance stretch in nearly a year during a tech-led AI selloff. Some managers were down close to 3%. The culprit was crowded positioning. Forced liquidations from long positions in AI and tech stocks compounded the problem.
Goldman’s machine-learning hedging strategies attempt to identify when momentum trades are most vulnerable to reversal. The approach involves maintaining core AI exposure while simultaneously purchasing downside protection through options structures or relative value trades that profit from a momentum unwind.
Why crypto traders are paying attention
Crypto traders are increasingly adopting similar hedging techniques as those suggested for traditional equities, recognizing that AI equity momentum has become a meaningful driver of crypto sentiment. Crypto-native traders are applying Goldman’s risk management playbook by hedging directional crypto positions against AI equity momentum shifts.
