Hedge Funds and Mutual Funds Diverge on AI but Both Raise Financials Tilts to Record Highs

Overview

Using institutional holdings from the beginning of the third quarter of 2026, the report compares the performance, market exposure, AI positioning, sector tilts, and shared top holdings of U.S. hedge funds and large-cap active mutual funds. The key conclusion is that both types of institutions have delivered respectable overall performance this year, and their current equity exposure is below recent peaks but still above multi-year norms. They diverge markedly on the AI trade but agree on the Financials and Health Care sectors.

Core views

Goldman Sachs' comparison covers $5.4 trillion in gross equity positions held by 991 hedge funds and $4.6 trillion in equity assets held by 504 large-cap active mutual funds, measured at the beginning of the third quarter of 2026. Both types of institutions have generated relatively solid returns this year, but their most popular holdings have not performed particularly well. 42% of large-cap mutual funds have outperformed their respective benchmarks year to date, above the historical average of 37%. However, the mutual fund overweight basket GSTHMFOW has risen only 9%, trailing the 16% return of the equal-weight S&P 500 Index and the 24% return of the mutual fund underweight basket GSTHMFUW. U.S. equity long/short hedge funds had returned approximately 10% year to date through August 20; their popular long basket GSTHHVIP rose 12%, while the concentrated short basket GSCBMSAL rose 38%. This indicates that overall fund returns were not driven primarily by the most crowded and popular holdings. Equity market exposure for both groups is below recent peaks but remains above typical levels of the past several years. During the second quarter, hedge funds experienced simultaneous increases in returns, leverage, and crowding, followed in July by one of the sharpest deleveraging episodes of the past decade. Their net leverage ranks in the 27th percentile relative to the past 12 months but remains in the 62nd percentile relative to the past five years. Gross and net leverage are both below their averages for the past 12 months but remain above levels of recent years. Mutual fund cash represented 1.2% of assets at the end of June, only slightly above the record low of 1.1% reached in December 2025, and remains among the lowest levels on record. Therefore, although current institutional risk exposure has retreated from its highs, it is not low. AI is the most prominent area of divergence between the two types of institutions. Hedge fund portfolios and their popular holdings have maintained a high correlation with fluctuations in the AI trade in recent months and, overall, remain more deeply engaged in the AI theme than mutual funds. Mutual funds have substantially increased their weighting in AI infrastructure stocks this year, but their purchases have not kept pace with the increase in benchmark weights, leaving them significantly underweight the theme. During the second-quarter AI rally, hedge funds sold several large AI stocks and most mega-cap stocks but increased holdings in MSFT and AMZN; mutual funds sold both stocks during the quarter. Conversely, most mutual funds increased holdings in AMD, MU, and SNDK, while hedge funds reduced them. The main text also lists 12 AI infrastructure stocks purchased by both types of institutions: AEP, AXTI, BE, CRWV, FLEX, LGN, NI, SANM, SITM, S, TLN, and XEL, showing that the two groups do not hold opposing views on every AI subsector. Outside AI, the sector tilts of the two types of institutions are generally aligned. Using the Russell 3000 as the reference for hedge fund net sector positions, both groups are substantially overweight Health Care and substantially underweight TMT. The clearest sector divergence appears in Consumer sectors: hedge funds are overweight Consumer Discretionary and underweight Consumer Staples, while mutual funds hold the opposite positioning. This indicates that the two groups have different views on AI and the internal structure of Consumer sectors, while still sharing substantial directional overlap across several major sectors. Financials is the strongest shared positioning theme. Hedge funds and mutual funds were simultaneously overweight Financials for only the third quarter in Goldman Sachs' historical data, and the Financials tilts of both groups rose to the highest levels in their respective sample histories. Hedge funds increased their net Financials tilt by more than 300 basis points in the second quarter, creating their largest Financials position since before the Global Financial Crisis. Mutual funds also increased their Financials tilt in the second quarter, and their current overweight is the highest since at least 2012. Financial stocks purchased by both types of institutions include COF, CPAY, FISV, and IBKR. The report identifies six “shared favorite” stocks included in both the hedge fund VIP basket and the mutual fund overweight basket: BA, COF, MA, SPCX, TMO, and V. COF, SPCX, and TMO entered the list this quarter, while MRVL exited. The rolling shared favorites portfolio has returned 29% year to date, compared with 16% for the equal-weight S&P 500 Index, an outperformance of 13 percentage points, and has also outperformed the most popular holdings of each institutional group. Since 2013, the portfolio has generated a 17% annualized return with a return standard deviation of 22%, meaning its historical excess performance has come at the cost of higher volatility. The median P/E ratio of the shared favorite stocks is 25x, substantially above the 19x median for S&P 500 constituents, indicating that the institutional consensus portfolio also carries a higher valuation.

Analysis framework

The report first uses fund holdings from the beginning of the third quarter to compare mutual fund performance relative to benchmarks with aggregate hedge fund returns, and then assesses market exposure using cash ratios, gross leverage, net leverage, and historical percentiles. At the sector level, it compares mutual fund deviations from benchmarks with hedge fund net positions relative to the Russell 3000, then tracks purchases and sales of specific stocks during the second quarter to identify AI divergence, Financials consensus, and differences in Consumer sectors. Finally, it constructs a “shared favorites” portfolio from the intersection of Goldman Sachs fund holding baskets and measures its performance and characteristics using historical returns, volatility, and P/E ratios.

Methodology notes

• Event-Driven Strategies and Behavioral FinanceFund Flow/Positioning Analysis

  Comparison of institutional holdings, purchases and sales, and benchmark tilts

  The report identifies fund preferences through quarterly fund holdings and changes in stock ownership, and compares mutual fund overweights and underweights relative to benchmarks with hedge fund net sector positions to distinguish consensus trades from divergent trades.

• Quantitative/Factor/Portfolio Theory

  Weighted aggregate performance estimate for anonymous equity long/short funds

  Hedge fund returns are calculated by Goldman Sachs Prime Services as a weighted average based on anonymously aggregated positions of equity long/short clients; the report also cautions that these data should not be viewed as representative of the entire market.

• Valuation MethodologyPE/PEG valuation

  P/E ratio comparison for shared favorite stocks

  The report compares the 25x median P/E ratio of shared favorite stocks with the 19x median for S&P 500 constituents to illustrate the valuation premium of the institutional consensus portfolio.

• Quantitative/Factor/Portfolio Theory

  Ratio of return to realized volatility

  The report measures risk-adjusted performance by dividing returns by realized volatility, allowing the returns of different assets or portfolios to be compared with their degree of volatility on a common scale.

• Valuation MethodologyDDM Dividend Discount Model

  Implied equity risk premium from a multi-stage S&P 500 dividend discount model

  The report calculates the implied equity risk premium by solving for the discount rate that equates the fair value from a multi-stage S&P 500 dividend discount model with the current market price, then subtracting the 10-year U.S. Treasury yield.

Asset mapping & comparison

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

• Hedge Fund VIP Basket (GSTHHVIP)

  Represents the most popular hedge fund long positions, with a year-to-date return of 12%, below the 38% return of the concentrated short basket.

  Strengths

  Hedge funds overall have still generated a positive year-to-date return of approximately 10%.

  Weaknesses

  Popular long positions have not been the strongest source of hedge fund returns this year.

  Comparison

  Below the 38% return of the concentrated short basket GSCBMSAL.

  Risks

  The portfolio maintains a high correlation with fluctuations in the AI trade.

• Mutual Fund Overweight Basket (GSTHMFOW)

  Represents the most significant overweight positions of large-cap active mutual funds, with a year-to-date return of 9%.

  Strengths

  42% of large-cap mutual funds overall have outperformed their benchmarks, above the historical average.

  Weaknesses

  The basket's return trails the equal-weight S&P 500 Index and the mutual fund underweight basket.

  Comparison

  GSTHMFOW returned 9%, the equal-weight S&P 500 returned 16%, and GSTHMFUW returned 24%.

• Shared Favorite Stocks (BA, COF, MA, SPCX, TMO, V)

  Included in both the hedge fund VIP basket and the mutual fund overweight basket; the rolling portfolio has returned 29% year to date.

  Strengths

  It has generated a 17% annualized return since 2013 and has outperformed the equal-weight S&P 500 Index by 13 percentage points year to date.

  Weaknesses

  Its historical standard deviation is 22%, with higher returns accompanied by higher volatility.

  Comparison

  Its median P/E ratio is 25x, above the 19x median for S&P 500 constituents.

  Risks

  Higher volatility and a significant valuation premium relative to the S&P 500.

• Divergent AI Stocks (MSFT, AMZN, AMD, MU, SNDK)

  In the second quarter, hedge funds increased holdings in MSFT and AMZN and reduced holdings in AMD, MU, and SNDK; mutual funds moved in the opposite direction.

  Weaknesses

  The two types of institutions have not formed a consensus positioning direction.

  Comparison

  Hedge funds overall are more deeply engaged in the AI trade, while mutual funds remain significantly underweight the AI theme.

• AI Infrastructure Stocks Purchased by Both Groups (AEP, AXTI, BE, CRWV, FLEX, LGN, NI, SANM, SITM, S, TLN, XEL)

  The main text identifies them as 12 AI infrastructure stocks purchased by both hedge funds and mutual funds in the second quarter of 2026.

  Strengths

  Both types of institutions increased their holdings.

  Comparison

  This contrasts with their opposing actions in MSFT, AMZN, AMD, MU, and SNDK.

• Financial Stocks Purchased by Both Groups (COF, CPAY, FISV, IBKR)

  Both types of institutions increased their holdings in the second quarter, reflecting their shared overweight in Financials.

  Strengths

  The Financials tilts of both hedge funds and mutual funds reached the highest levels in their respective historical datasets.

  Comparison

  Hedge funds increased their net Financials tilt by more than 300 basis points; the mutual fund overweight is the highest since at least 2012.

Key data

• Total Scale of Holdings AnalyzedApproximately $10 trillionCombined gross hedge fund equity positions and large-cap active mutual fund equity assets at the beginning of the third quarter of 2026
• Hedge Fund Sample991 funds; $5.4 trillionGross equity positions
• Mutual Fund Sample504 funds; $4.6 trillionEquity assets of large-cap active mutual funds
• Share of Mutual Funds Outperforming Benchmarks42%Year to date, above the historical average of 37%
• U.S. Equity Long/Short Hedge Fund Return+10%Year-to-date return through August 20, 2026
• Fund Basket Year-to-Date ReturnsGSTHMFOW 9%; GSTHMFUW 24%; GSTHHVIP 12%; GSCBMSAL 38%Respectively, the mutual fund overweight basket, mutual fund underweight basket, hedge fund popular long basket, and concentrated short basket
• Mutual Fund Cash Ratio1.2%As of the end of June 2026; the record low in December 2025 was 1.1%
• Hedge Fund Net Leverage Percentile27th percentile over the past 12 months; 62nd percentile over the past 5 yearsIt has retreated from recent highs but remains elevated relative to multi-year levels
• Change in Net Financials TiltMore than 300 basis pointsIncrease for hedge funds in the second quarter of 2026
• Shared Favorites Portfolio Year-to-Date Return29%The equal-weight S&P 500 Index returned 16%, with the shared favorites portfolio outperforming by 13 percentage points
• Long-Term Performance of the Shared Favorites Portfolio17% annualized return; 22% standard deviationSince 2013, higher returns have been accompanied by higher volatility
• Valuation of Shared Favorite Stocks25x P/EMedian; the median for S&P 500 constituents is 19x
• Shared Favorite StocksBA, COF, MA, SPCX, TMO, VCOF, SPCX, and TMO entered this quarter, while MRVL exited

Impact & implications

The report shows that the effectiveness of popular holdings cannot be judged solely from overall fund returns: hedge funds and mutual funds have delivered solid overall performance this year, but the most crowded long or overweight baskets have trailed the underweight and concentrated short baskets. The AI trade remains the primary source of divergence between the two groups, while Financials has become a rare shared overweight at historically elevated levels. The shared favorites portfolio has generated strong historical returns, but its long-term volatility of 22% and median P/E ratio of 25x indicate that these institutional consensus holdings also carry higher volatility and valuation characteristics.

Risks

• The historical excess returns of shared favorite stocks have come at the cost of higher volatility: their return standard deviation since 2013 is 22%, and their median P/E ratio of 25x is also above the 19x median for S&P 500 constituents.