There is a gap opening up in US equity markets, and it is costing mutual fund managers dearly. Large-cap active mutual funds are now sitting roughly 107 basis points underweight in AI-related equities relative to their benchmarks, according to a Goldman Sachs analysis covering 504 actively managed mutual funds with a combined equity asset value of approximately $4.6 trillion.
To put that in plain terms: fund managers have been buying AI stocks, just not fast enough to keep up with how much those stocks have grown inside the benchmarks they are trying to beat.
Nvidia is the headline problem
Nvidia sits at the center of this story. The chipmaker is the single largest underweight among all AI equities tracked by Goldman, clocking in at around 100 basis points of underweighting on its own.
Other names are adding to the drag. AMD sits at approximately 60 basis points underweight, Alphabet at around 70 basis points, and Microsoft at roughly 50 basis points. The pattern is consistent: large-cap mutual funds have been gradual accumulators of AI infrastructure and hyperscaler stocks, but gradual has not been fast enough.
There is one exception worth flagging. Micron Technology shows up as an overweight of about 40 basis points, suggesting fund managers were selectively comfortable with parts of the semiconductor value chain.
The performance math is brutal
The positioning has a direct cost. Only 36% of large-cap core and growth funds managed to outperform their style benchmarks in the period Goldman examined.
Goldman’s analysis covered not just the 504 mutual funds but also 991 hedge funds, with hedge fund equity assets totaling approximately $5.4 trillion. Hedge funds have maintained deeper exposure to AI themes, adjusting holdings more aggressively as the AI-driven rally in megacap technology unfolded.
Hedge funds operate with fewer constraints on position sizing and benchmark tracking, giving them room to run concentrated bets on high-conviction themes. Mutual funds, particularly those benchmarked to large-cap indexes, face a different problem: the bigger a stock gets in the index, the harder it is to justify holding less of it without an explicit contrarian thesis.
Why active managers are behind
The underweighting reflects a deliberate tilt that many large-cap active managers have maintained toward cyclical and value-oriented investments, even as AI momentum has accelerated.
There is also a structural dynamic at play. As benchmark weights shift to reflect a stock’s rising market capitalization, active managers who were already underweight face compounding pressure. Every quarter that Nvidia’s weighting in the index grows, the underweight position effectively gets larger even if the manager does not sell a single share. Keeping pace requires active buying, not just holding.