Goldman Sachs is sounding the alarm on what it sees as a dangerous concentration in the artificial intelligence-driven stock market rally.
In a May 15 report, Goldman Sachs strategist Ben Snider said the S&P 500 has climbed 10% this year, but the gains have become heavily concentrated in technology and AI-linked companies.
Technology stocks alone accounted for 85% of the index’s return, while the S&P 500 excluding technology rose just 3%. Nvidia (NASDAQ: NVDA), which now represents 9% of the index by market value, contributed 20% of the total return this year on its own.
Related: Goldman Sachs analyst has a no-nonsense approach on inflation
The ‘insensitive portfolio’ solution
Goldman described the market as increasingly dominated by a single theme, with AI enthusiasm and momentum trading becoming tightly linked.
The bank noted that periods of narrow leadership and rapid momentum gains have historically preceded higher volatility and weaker equity returns. Goldman pointed to similar episodes in 1998, 1999, 2015 and 2021.
To navigate this, Goldman published an “insensitive portfolio” of stocks with positive earnings revisions but low sensitivity to AI trading and macro-growth shifts. Consumer staples, healthcare, and real estate showed the lowest correlations with the AI trade.
The list thus includes Eli Lilly (NYSE: LLY), Reddit (NYSE: RDDT), Newmont (NYSE: NEM), Archer-Daniels-Midland (NYSE: ADM), and Casey’s General Stores (NASDAQ: CASY).
More from TheStreet Roundtable
Why crypto investors should care
The implications for crypto are direct. May 2026 has seen the lines between crypto and AI continue to blur.
TeraWulf, originally a Bitcoin (BTC) miner, hit a 52-week high this week after acquiring a hyperscale AI data center campus in Kentucky. The pivot is supported by Alphabet’s Google, which holds roughly a 14% equity stake.
TeraWulf is among the growing number of publicly traded Bitcoin mining companies priced as AI plays rather than pure crypto plays.
Within crypto itself, the AI-token category, which includes Fetch.ai, Render, and Bittensor, has tracked equity-market AI sentiment more closely than Bitcoin fundamentals. If the AI trade unwinds, the assets most exposed will not just be chip stocks; they will be crypto infrastructure plays, AI tokens, and leveraged treasury vehicles riding the same wave.
But not all hope is lost. Veteran investor and Bitwise advisor Jeffrey Park has identified what he believes could become the next “NVIDIA,” and his answer is crypto.
