Every allocator has a line item marked "passive equity" — the boring bucket that doesn't eat into the active-risk budget. Doug Kramer and Maarten Nederlof of Neuberger want you to look at it again1. "We argue that rising exposure to the AI theme has rendered broad equity benchmarks passive in name only, and that equity indexing now amounts to a deliberate, one-size-fits-all bet on AI," they write. That's a thesis, and the numbers back it up.
The Concentration Nobody Budgeted For
Start with the Magnificent Seven. In 2015 they were 11% of the S&P 500. Today they're pushing 35% — a familiar statistic by now, one Kramer and Nederlof concede "sounds like a tired refrain." So they reframe it in the currency that actually matters to risk managers: volatility contribution, not weight. "On a contribution-to-risk basis, we find these seven stocks recently accounted for nearly half of the index's volatility." Weight tells you what you own. Risk contribution tells you what can hurt you. Those are different numbers, and the second one is the scarier one.
The Math Gets Worse Abroad
This isn't a Nasdaq problem you can diversify away by going global. The authors find that "an AI basket of roughly 120 companies drives more than half the volatility of the MSCI All-Country World Index, a passive global benchmark spanning more than 3,000 names." Three thousand tickers, and the risk still boils down to a basket you could fit in a mid-cap fund.
The structural drift shows up in drawdown history, too. In the dotcom 1990s, the Nasdaq 100 suffered three drawdowns greater than 15% against just one for the S&P 500. In the 2020s, "both indices have suffered four such drawdowns thus far." The "effective number of constituents" — a measure that discounts small names that never move the needle — tells the same story: the S&P 500's historical average is 117 effective stocks. "We are now approaching 46." Kramer and Nederlof don't soften it: "That's not passive diversification, in our view. It's an active bet with material risk."
How the Sausage Gets Made
The concentration isn't just organic drift — it's getting an assist from the index providers themselves. SpaceX "loudly joined the Nasdaq 100 Index, the Russell 1000 Index and various MSCI indices" this summer, with OpenAI and Anthropic likely next, even though "those three AI titans are still a far cry from meeting established index-selection criteria, including profitability and free-float requirements." Their read: "We believe these exceptions represent active decisions made by the index providers."
Why would S&P Global, FTSE Russell and MSCI bend their own rules? Because "index companies, remember, answer to their own shareholders — not to the asset managers who license their benchmarks, and not explicitly to the end investor." The result: "these moves amount to active decisions that stand to benefit the index providers, passive managers and public stock exchanges in the near term," even as "none of these parties has taken a fiduciary stand like an active manager would be required to do."
What Allocators Should Do
Kramer and Nederlof are careful not to overreach. "We are not saying to sell the broader equity index, nor are we arguing that AI exposure is inherently bad." Their case is narrower and sharper: "we do not believe AI is an all-in bet," and the benchmark shouldn't make that call for you silently. Their advice: build a true AI risk budget across every asset class rather than trusting "simple GICS classifications," and force the conversation upstairs. "If you told them an index had 50% of its risk in a single theme, that would be worth a conversation, in our view." Their closing line sums it up: "Don't make a big active bet by accident."
Five Takeaways for Advisors and Investors
1. Measure AI exposure by risk contribution, not headline weight — the Mag 7 carry nearly half the S&P 500's volatility, well above their ~35% index weight.
2. Global diversification doesn't dilute this risk; roughly 120 AI-linked names drive over half the MSCI ACWI's volatility across 3,000+ constituents.
3. Structural diversification has thinned: the S&P 500's effective constituent count has fallen from a historical average of 117 to roughly 46.
4. Index inclusion rules are being bent for high-profile AI names (SpaceX, likely OpenAI and Anthropic), an active choice by providers who answer to their own shareholders, not to end investors.
5. Treat "passive" equity as a live risk budget: quantify true AI exposure across the whole portfolio and revisit liquidity and diversification assumptions before a drawdown forces the issue.
Footnote:
1 Kramer, Doug, and Maarten Nederlof. "Passive in Name Only: The Active Bet Within Your Equity Index." Neuberger, 30 July 2026, www.nb.com/insights/article-passive-in-name-only-the-active-bet-within-your-equity-index.
