Do Hedge Funds Beat the Market? What the Numbers Say
Hedge fund investors' actual returns trailed the S&P 500 after adjusting for cash-flow timing. Peer-reviewed evidence, Buffett's ten-year bet figures, and Jensen's alpha for seven famous investors' 13F portfolios.
Not for the people invested in them. Hedge fund investors' actual returns fell below the S&P 500 after adjusting for cash-flow timing, per Dichev and Yu in the Journal of Financial Economics (2011). Buffett's ten-year bet pointed the same way. Among US large-cap equity funds, 92.89% trailed the S&P 500 over the 20 years to December 31, 2025 (SPIVA, Year-End 2025). The question is whether a public filing tells you enough to spot the few who do.
- Hedge fund investors earned 3 to 7 percentage points less per year than the buy-and-hold returns of the funds they held, once dollar-weighted returns were measured, trailing the S&P 500 (Dichev and Yu, Journal of Financial Economics, 2011).
- Buffett's ten-year bet ended with the S&P 500 index fund at +125.8% cumulative and the best fund-of-funds basket at +87.7% (Berkshire Hathaway, 2017 annual letter, January 2008 to December 2017).
- 92.89% of US large-cap equity funds trailed the S&P 500 over 20 years ending December 2025 (SPIVA Scorecard, Year-End 2025).
What percentage of active managers beat the market?
A minority at every horizon SPIVA measures, and fewer than one in six over five years or longer. S&P Dow Jones Indices publishes the SPIVA Scorecard twice a year, comparing actively managed US equity mutual funds and ETFs against their benchmarks. The Year-End 2025 edition, with data as of December 31, 2025, reports these underperformance rates for US large-cap equity funds measured against the S&P 500 (Report 1a; the table leaves out the three-year rate, 66.84%):
| Horizon | Funds that trailed the S&P 500 |
|---|---|
| 1 year | 78.78% |
| 5 years | 88.96% |
| 10 years | 85.59% |
| 15 years | 89.93% |
| 20 years | 92.89% |
Fewer than one in ten US large-cap equity funds beat the S&P 500 over 20 years ending December 2025, and the underperformance rate stays above 85% at the five-, ten-, 15- and 20-year horizons (Year-End 2025, data as of December 31, 2025). The ten-year rate (85.59%) dips below the five-year (88.96%), so the staircase is not perfectly clean, but the direction holds.
SPIVA accounts for survivorship bias by including funds that closed or merged during the period. These numbers cover US equity mutual funds and ETFs and do not directly measure hedge funds. For the funds they do cover, the verdict is plain: after fees, beating the index for long is rare.
Did Warren Buffett prove the point?
He put money on it. In 2007, Buffett wagered that a Vanguard S&P 500 index fund would outperform any selection of at least five hedge funds over the following decade. Ted Seides of Protégé Partners accepted and chose five funds of hedge funds (Berkshire Hathaway's 2016 annual letter). The bet ran from January 2008 through December 2017.
Buffett published the final results in Berkshire Hathaway's 2017 annual letter:
| Cumulative gain, 2008 to 2017 | |
|---|---|
| S&P 500 index fund | +125.8% |
| Fund of funds C (best) | +87.7% |
| Fund of funds B | +42.3% |
| Fund of funds E | +27.0% |
| Fund of funds A | +21.7% |
| Fund of funds D | +2.8%* |
*Fund D was liquidated in 2017, so its 2.8% cumulative gain covers approximately nine years before the bet ended.
The index fund returned +125.8% over the decade, beating the best of five fund-of-funds baskets by 38 percentage points (Berkshire Hathaway, 2017 annual letter, cumulative returns January 2008 to December 2017). Each basket carried two layers of fees that compounded against the investor over the same period that returns compounded for the index.
The five baskets' individual annualized returns ranged from 0.3% to 6.5%, against the index fund's 8.5%. The fees worked like a toll booth with a second toll booth behind it: one to the underlying hedge funds and one to the fund-of-funds manager above them. All investing involves risk, including possible loss of principal, and past performance does not guarantee future results.
The bet settled that a low-cost index beat a diversified selection of hedge funds over a decade that included the 2008 crisis. It did not settle whether a concentrated stock-picker, paying one layer of fees rather than two, can sustain an edge.
Can anyone actually beat the market consistently?
Some have, measured in hindsight. Berkshire Hathaway's compounded annual gain in per-share market value was 19.7% from 1965 through 2025, against 10.5% for the S&P 500 with dividends, per Berkshire Hathaway's 2025 annual letter (dated February 28, 2026).
The trouble is telling skill from survivorship. Many other managers started over those same decades. Survivorship bias works like a basketball tryout: the players who get cut never appear in the final roster, so the roster's shooting average always looks better than the gym's. SPIVA includes defunct funds for exactly that reason; a screen of today's funds leaves them out.
The most direct measurement comes from Dichev and Yu's 2011 study in the Journal of Financial Economics, covering hedge funds from 1980 onward. They used dollar-weighted returns, weighting each year's return by how much money was in the fund at the time. The gap was 3 to 7 percentage points a year between the funds' buy-and-hold returns and what investors earned, because capital chased performance: it flowed in after strong years and out after weak ones. After that adjustment, hedge fund investors' annual returns fell below the S&P 500 and barely exceeded the risk-free rate as of the end of 2008. Reported hedge fund performance overstates the experience of the people who wrote the checks.
Hedge funds also resist a single benchmark. A market-neutral fund that targets uncorrelated returns can trail the S&P 500 in a bull run and still do the job its investors hired it for, so one answer covering a stock-picker and an arbitrage fund hides more than it reveals.
What can a 13F filing actually tell you?
Part of the picture. The SEC requires institutional investment managers with at least $100 million in qualifying assets to file Form 13F quarterly, disclosing their long US-listed equity positions as of the end of each calendar quarter. The filing is public and due within 45 days of the quarter's close, and it shows position sizes in shares and market value.
What it shows: long positions in securities on the SEC's official 13(f) list, mainly US-listed stocks and ETFs, with share counts and dollar values. Long put and call options on listed securities are also reportable, per the SEC's Form 13F FAQ (accessed October 8, 2026). What it omits: short positions, cash, most fixed income, and shares listed only outside the US. A 13F is a quarterly photograph of one room in a house; the bedrooms, the basement and the debts are out of frame.
The filing lag sharpens the limitation. A 13F dated March 31 can be filed as late as mid-May, and the positions are older still by the time you read it. The positions may have been sold, hedged, or doubled. A manager who bought and sold a position between two quarter-ends looks identical, in both filings, to one who never touched it.
How can you check your own portfolio?
Type your tickers and share counts into a free PortLens scan to see your portfolio's return against the SPY ETF's total-return series, a proxy for the S&P 500, alongside its beta and diversification. No account is required. The comparison uses a trailing window; its scoring and risk-adjustment methods are documented in our methodology.
A scan cannot measure your money-weighted return, because it does not know when you bought or added money. What it measures is the thing a 13F cannot show: the holdings you enter, as they stand today, with the concentration, cost and risk they carry. The Does the S&P 500 Return Include Dividends? post explains why the benchmark choice changes the reading.
Did seven famous investors' June 13F portfolios beat the S&P 500?
One did. We took the 13F filings of the seven investors PortLens follows on its investor pages for the quarter ending June 30, 2026, held each portfolio unchanged, and measured it against the S&P 500 total-return index from June 30 to October 9, 2026. Only the portfolio in Bill Ackman's Pershing Square filing beat the index on return and on alpha. Berkshire Hathaway finished a fraction behind the index but carried less market risk, so its alpha came out just above zero. The other five trailed.
Alpha here is Jensen's alpha: the return left over after allowing for the market risk the portfolio carried. A portfolio with a beta of 1.5 is expected to move about one and a half times as far as the market, so matching the index is not enough. The calculation takes the one-year Treasury yield (4.47% on October 9, 2026, per the US Treasury, scaled to the 101-day window) off both returns, multiplies the market's remainder by the portfolio's beta, and reports what is left in percentage points.
| Investor (filer) | Jensen's alpha | Beta | Portfolio return | Return minus S&P 500 TR |
|---|---|---|---|---|
| Bill Ackman (Pershing Square) | +3.93 | 1.18 | +8.99% | +4.51 |
| Berkshire Hathaway | +0.16 | 0.89 | +4.30% | −0.18 |
| Li Lu (Himalaya Capital) | −2.82 | 0.80 | +1.01% | −3.47 |
| Chase Coleman (Tiger Global) | −5.26 | 1.52 | +0.89% | −3.59 |
| David Tepper (Appaloosa) | −8.68 | 1.59 | −2.30% | −6.78 |
| Chris Hohn (TCI Fund Management) | −8.80 | 1.14 | −3.85% | −8.32 |
| Gates Foundation Trust | −9.78 | 0.86 | −5.76% | −10.24 |
Window: June 30 to October 9, 2026. Benchmark: S&P 500 total-return index, +4.48% over the window. Holdings: each filer's 13F for the quarter ending June 30, 2026, filed August 14, 2026, weighted by reported market value. Returns use dividend-adjusted prices.
Beta explains part of the spread. The filed Tepper and Coleman portfolios were about one and a half times as sensitive as the market, so they needed far more than the index to score positive alpha; held unchanged, one lost money and the other barely gained.
These figures measure dated snapshots. About three months is far too short to separate skill from luck, each portfolio is frozen at June 30, and every 13F limit described above applies. Two of Tepper's put options (3.3% of his reported value, per the filing) are left out, because a 13F reports an option at the value of its underlying shares rather than its cost (Form 13F instructions). Pershing Square Capital Management filed only a notice (Form 13F-NT) for the June quarter; the holdings were reported by Pershing Square Inc. Beta is the provider-published beta for each holding at the time of this calculation, weighted by position. We computed the figures with PortLens's own return and alpha code, documented in our methodology.
All investing involves risk, including possible loss of principal, and past performance does not guarantee future results. A positive alpha over one quarter says nothing about any manager's skill going forward, and nothing about whether to copy or avoid their holdings.
Key takeaways
- Hedge fund investors earned 3 to 7 percentage points less per year than the buy-and-hold returns of the funds they held, once cash-flow timing was measured, and their actual returns fell below the S&P 500, per Dichev and Yu in the Journal of Financial Economics (2011).
- Buffett's ten-year bet ended with the index fund at +125.8% and the best fund-of-funds basket at +87.7% (cumulative, January 2008 to December 2017).
- 92.89% of US large-cap equity funds trailed the S&P 500 over 20 years ending December 2025, per the SPIVA Scorecard. These are mutual funds and ETFs, not hedge funds; the hedge fund evidence above comes from Dichev and Yu and from Buffett's bet.
- Of seven famous investors' June 30, 2026 13F portfolios, only Pershing Square's beat the S&P 500 total return from June 30 to October 9, 2026 (alpha +3.93 percentage points; see our methodology). The other six trailed it on return, and five had negative alpha.
- 13F filings show long positions in US-listed stocks, ETFs and listed options as of each quarter end, are due 45 days later, and omit shorts and cash.
This article is for information and education only and is not investment advice. Analytics referenced are computed as described in our methodology; see our disclosures.
