US large-cap active mutual fund positioning and benchmark-relative tilts Report Interpretation

Overview

This report tracks Q2 2026 holdings, performance and flows across large-cap active US mutual funds. It finds that benchmark changes materially reshaped measured tilts, while managers remained structurally underweight mega-cap technology and AI exposure and increased relative exposure to Financials, Industrials, Health Care and Consumer Staples.

Core views

Goldman Sachs analyzes quarter-end positioning across 504 large-cap active mutual funds with $4.6 trillion in equity assets. Fund performance was somewhat better than usual in aggregate: 41% of large-cap funds outperformed their benchmarks year to date, above the 37% historical average. Growth funds were the strongest relative group, with 74% outperforming, versus 42% of core funds and only 9% of value funds. Value funds had the best absolute return at 24.0% YTD but lagged their Russell 1000 Value benchmark, which returned 17.3%, illustrating how a rising benchmark made relative outperformance difficult. Cash balances were 1.2% of assets at June-end, near the 1.1% record low in December 2025. US equity mutual funds and ETFs had received roughly $400 billion of YTD inflows, but active mutual funds remained on track for a 12th consecutive year of outflows; 8% of funds in the sample held ETFs at Q2-end. The June 2026 Russell reconstitution was a central driver of reported changes in style-manager tilts. The rebalance took effect after the June 26 close and produced historically large constituent-weight changes in Russell 1000 Growth and Value, partly reflecting AI's growing effect on company fundamentals. The standard deviation of constituent-weight changes was 0.5 percentage points for Russell 1000 Growth and 0.3 percentage points for Russell 1000 Value—five times and three times, respectively, the S&P 500's change. Apple, Amazon, Microsoft, Alphabet and Micron experienced major style-index membership or weight changes. Goldman Sachs notes that fund portfolio weights themselves moved by less than 0.1 percentage points in standard deviation for both growth and value managers, so benchmark movement rather than active trading accounted for most large observed tilt changes. AI-exposed equities became a larger part of portfolios but lagged their benchmark representation. Excluding AMZN, AVGO, GOOGL, META, MSFT and NVDA to limit the effect of diversification restrictions, AI-exposed stocks accounted for 13% of the average large-cap fund at the start of Q3, more than twice their weight at the start of 2024. Their benchmark weight reached 15%, however, leaving the average fund 175 bp underweight. Managers were underweight Micron, AMD and SanDisk but these were the three most popular broad increases in Q2; CrowdStrike, Datadog and Palo Alto Networks also ranked among popular additions. The report separately notes a 9 bp average overweight in Seagate, with a net 24 funds adding it. Funds remained 724 bp underweight the Magnificent 7 at the end of Q2, essentially unchanged from a 723 bp underweight in Q1. Goldman Sachs attributes part of the persistent gap to diversification restrictions. On a share-ownership basis, a majority of funds reduced every Magnificent 7 position, with the largest reductions in Alphabet and Meta. The report states that the Magnificent 7 lagged the broader market YTD and that only NVIDIA outperformed the equal-weight S&P 500 among the group. At the sector level, Financials was the largest average fund overweight at +231 bp, after increasing by 41 bp in Q2. Industrials (+204 bp) and Health Care (+197 bp) were also at 10-year highs, while Information Technology was the largest underweight at -660 bp and Communication Services was also at a 10-year-low tilt. Funds increased Consumer Staples by 45 bp, their largest sector increase, and became overweight the sector for the first time since tracking began in 2012 despite its +11% YTD return versus +15% for the S&P 500. Funds cut Information Technology by 182 bp, with Semiconductors accounting for the largest industry-level reduction, partly because of rapid share-price appreciation. Across core, growth and value styles, funds were overweight Financials, Industrials, Health Care, Energy and Materials, and all styles were underweight Information Technology. Goldman Sachs constructs equal-weighted, non-sector-neutral 50-stock baskets from the most overweight and most underweight Russell 1000 positions relative to a blended benchmark. The overweight basket returned +9% YTD, trailing both the equal-weight S&P 500 at +17% and the underweight basket at +25%. The rebalanced overweight basket added 13 names, including UnitedHealth, Amphenol and SpaceX; the underweight basket added seven, including Marvell, Lam Research and Fortinet. The biggest net share increases across Russell 3000 stocks were AMD, Intel and SanDisk, while Alphabet, Meta and JPMorgan Chase recorded the biggest net decreases. Goldman Sachs emphasizes that this share-based analysis is intended to isolate actual buying and selling, because overweight and underweight tilts can be distorted by stock-price moves and benchmark-weight changes.

Analysis framework

Goldman Sachs compares individual fund holdings with blended benchmarks across core, growth and value managers, then separates apparent positioning changes caused by benchmark weights from changes in shares owned. It also tracks fund performance, cash and flows, sector tilts, AI and Magnificent 7 exposure, and rebalances equal-weighted baskets of the most overweight and underweight Russell 1000 positions.

Methodology notes

• Quantitative, Factor, and Portfolio Theory

  Benchmark-relative portfolio tilt and share-based ownership analysis

  The report measures each fund's holdings against a blended benchmark to identify overweights and underweights, while separately using changes in shares owned to distinguish actual trading from moves caused by stock prices or benchmark reconstitution.

• Quantitative, Factor, and Portfolio Theory

  Equal-weighted mutual-fund positioning baskets

  Goldman Sachs creates 50-stock baskets from the most overweight and most underweight Russell 1000 positions. The baskets are equal-weighted and not sector-neutral, so their returns reflect positioning signals as well as sector composition.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• AMD (AMD)

  One of the largest net mutual-fund share increases in Q2 despite an average underweight position.

  Strengths

  Ranked among the three biggest net increases across Russell 3000 stocks.

  Weaknesses

  Funds remained underweight on average.

  Comparison

  Alongside Intel, SanDisk and Micron as a prominent semiconductor buying area.

• Intel (INTC)

  One of the three largest net mutual-fund share increases in Q2.

  Strengths

  Large number of funds increased exposure.

  Weaknesses

  Funds remained underweight on average.

  Comparison

  Grouped with AMD and SanDisk as leading net increases.

• SanDisk (SNDK)

  One of the three largest net mutual-fund share increases in Q2.

  Strengths

  Popular increase despite average fund underweight.

  Weaknesses

  Funds remained underweight on average.

  Comparison

  Grouped with AMD and Intel as leading net increases.

• Alphabet (GOOGL)

  A major net decrease and a Magnificent 7 position reduced by funds.

  Weaknesses

  Largest reduction in ownership among Magnificent 7 names; 110 more fund decreases than increases.

  Comparison

  Alongside Meta as the largest Magnificent 7 exposure reduction.

• Meta Platforms (META)

  A major net decrease and a Magnificent 7 position reduced by funds.

  Weaknesses

  89 more fund decreases than increases.

  Comparison

  Alongside Alphabet as the largest Magnificent 7 exposure reduction.

• GSTHMFOW / GSTHMFUW baskets

  Goldman Sachs baskets of the most mutual-fund-overweight and underweight Russell 1000 stocks.

  Strengths

  The underweight basket returned +25% YTD.

  Weaknesses

  The overweight basket returned +9% YTD, trailing the equal-weight S&P 500's +17%.

  Comparison

  GSTHMFUW outperformed GSTHMFOW by 16 percentage points YTD.

  Risks

  Basket tradability depends on market conditions, including liquidity and borrow constraints.

Key data

• Fund sample504 funds; $4.6 trillionLarge-cap active mutual funds and equity assets analyzed.
• Funds outperforming YTD41%Versus a 37% historical average; 74% of growth funds, 42% of core funds and 9% of value funds outperformed.
• Mutual fund cash balance1.2% of assetsAt end-June 2026, near the 1.1% low in December 2025.
• AI-exposed equity weight13% of average fund portfolio; 175 bp underweightBenchmark weight was 15%, excluding six largest AI-linked stocks.
• Magnificent 7 positioning-724 bpAverage large-cap mutual fund underweight at Q2-end, versus -723 bp in Q1.
• Sector tiltsFinancials +231 bp; Industrials +204 bp; Health Care +197 bp; Information Technology -660 bpFinancials increased 41 bp in Q2; Information Technology fell 182 bp.
• Positioning-basket performanceGSTHMFOW +9%; equal-weight S&P 500 +17%; GSTHMFUW +25%Year-to-date performance.

Impact & implications

The report portrays active managers as remaining benchmark-underweight in large technology and AI-linked stocks despite selective buying in semiconductors and software. Measured style tilts should be interpreted carefully because the Russell reconstitution altered benchmarks far more than managers altered portfolio weights. The sector data indicate a broad relative preference for Financials, Industrials, Health Care and, newly, Consumer Staples over Information Technology and Communication Services.

What to watch

• Changes in Russell benchmark weights and style-index membership following the June 2026 reconstitution.
• Whether AI-exposed portfolio weights close the 175 bp gap to benchmark weights.
• Further changes in mutual-fund ownership of the Magnificent 7, particularly Alphabet and Meta.
• Whether the rotation toward Financials, Industrials, Health Care and Consumer Staples persists.
• Performance and rebalancing effects of the mutual-fund-overweight and underweight baskets.