Diversification Now an Active Decision

• Rising concentration and overlapping holdings across styles can leave seemingly diversified portfolios dependent on the same handful of mega cap companies.
• Genuine diversification requires looking at the underlying companies, position sizes and shared investment themes across an entire portfolio.
• Deliberate security selection and limits on company and thematic exposure can enable active managers to provide the breadth that concentrated benchmarks and the passive vehicles that track them increasingly lack.

Over the last three decades, investors have gravitated to passive investment products for broad equity market exposure. Active products have been used as supplements, to express a deliberate view or obtain idiosyncratic exposure. This approach rests on the assumption that the passive index products are diversified; that is no longer the case. The S&P 500 Index, the Russell 1000 Growth Index and the Russell 1000 Value Index — among the most widely used benchmarks in the U.S. — have all become concentrated (Exhibit 1). Moreover, the indexes no longer look different — they have converged on the same companies.

Because success breeds complacency, and because the markets have done so well for so long, most investors are overlooking this risk. Allocating 50% to the S&P 500, 30% to the Russell 1000 Growth and 20% to the Russell 1000 Value, for example, yields a portfolio where just five companies tied to the single theme of AI — Nvidia, Apple, Alphabet, Microsoft and Amazon — represent 31% of the total portfolio.

Exhibit 1: Top 10 Index Constituents Over Time

The S&P 500 Index is the archetype for a broad market index. Yet four companies — Nvidia, Apple, Microsoft and Alphabet — now account for 27% of the benchmark. It takes just 20 companies to account for half the index weight by market cap.¹ Nvidia, on its own, is larger than the energy, utilities and real estate sectors combined; Nvidia, Apple and Microsoft together outweigh six of the index’s 11 sectors combined.²

Today, the 10 largest holdings represent 40.6% of the S&P 500 (Exhibit 2). For context, that figure hovered around 20% for most of the last 35 years, breaching 25% only at the height of the dot-com bubble before rising sharply beginning in 2020.

Exhibit 2: Top 10 Holdings in S&P 500, 1990-2026

Growth stocks are even less diversified. In the Russell 1000 Growth, two companies — Nvidia and Alphabet — make up a quarter of the index and seven make up half.³ The top 10 make up 60.5%. Nvidia alone is 15.7% — roughly one dollar in six. Approximately a third of the index is semiconductors, which have grown in weight dramatically here as they have in the S&P 500 (Exhibit 3).

Exhibit 3: Semiconductor Weight in S&P 500 and Russell 1000 Growth

The Russell 1000 Value has been less concentrated than the other two benchmarks since 2018, and remains the least concentrated of the three, but it has changed the fastest since the end of 2024. The weight of its top 10 stocks went from 17% at the end of 2024 to 30% today. Its exposure to mega cap technology and internet companies (Amazon, Apple, Microsoft and Meta) went from zero at the end of 2024 to about 18% today, most of it added in June 2026 (Exhibit 4).⁴

Exhibit 4: Russell 1000 Value, Top 10 Weight and Mega Cap Technology Exposure

With Amazon, Apple and Microsoft its three largest holdings, the value benchmark is no longer a reliable way to step away from the companies that dominate the core and growth benchmarks.

Set the three indexes side by side and the comparison is shocking: Apple and Microsoft each exceed 5% of all three benchmarks simultaneously (Exhibit 5). Nvidia and Alphabet each exceed 5% of two.

Exhibit 5: Three Indexes, Same Companies

While there are roughly 500 stocks in the S&P 500, it does not diversify like an equal-weighted portfolio of 500 stocks. As a capitalization weighted index, a 10% position exerts as much force in the index as 50 names that are 20 basis points each. For the roughly 500 constituents in the S&P 500 today, the effective number of holdings is 41 (Exhibit 6). In other words, the S&P 500 diversifies like a portfolio of 41. In 2016 it diversified like a portfolio of 138. The growth and value benchmarks also diversify much less than their names would suggest.

Exhibit 6: Composition of the Three Major U.S. Large Cap Benchmarks

As indexes become more concentrated, they become more volatile, as we saw in the selloff following 2026’s midterm elections. As of the end of the third quarter, a 10% move up or down in Nvidia — which it has matched or exceeded in about a third of the months since March 2025 — moves the S&P 500 by 0.84 points and the Russell 1000 Growth by 1.57 points. A 25% decline across the 10 largest S&P 500 holdings would cost the index 10.1%. At 2010 weights the same decline would have cost about 4.7%.⁵ Traditional risk statistics obscure this dynamic because the beta of the benchmark is always 1.0, no matter how volatile it is.

Traditional asset allocation has become inverted. Passive was supposed to be the diversified, broad, risk-spread component — the beta. Active was the concentrated version, bought for alpha. Today the index is the concentrated position and most investors are unaware of the risks embedded in their “diversified” portfolios. Thoughtful investors must move beyond style boxes and away from concentrated benchmarks and increase their active exposure.