Wins and Losses of the Giants

1. Report Card: The Giants Together vs the Market

Q1 1999–Q2 2026, 109 quarters: each quarter, the 13F holdings of the giants counted that quarter are combined, the quarter’s change in quarter-end prices is applied, and the result is compared with the market (SPY). Price change only; neither side includes dividends or fees.

Q1 1999 = 100; price change only, both sides exclude dividends and fees.

By Period: Not Ahead All the Way

• Of the 62 giants counted for at least 5 years (20 quarters), 23 had 13F holdings that beat SPY over the quarters they were counted; the median giant was 0.9 percentage points a year behind SPY. No names, no rankings.
• The overall lead comes mainly from 2000–2007; most giants were counted after 2013.

Year by Year

Method

• Each quarter starts from the long US stock holdings filed by the giants on the roster at the end of the previous quarter and applies that quarter’s change in quarter-end prices (split-adjusted). The roster is recomputed every quarter with the same rules (13 giants in Q1 1999, 58 now).
• Composite: one vote per giant (the same method as the site’s composite weight). Money-weighted: big portfolios weigh more; Berkshire once made up as much as 65%.
• SPY: from Q2 2013, the SPY quarter-end price on the giants’ own 13Fs; before that few giants held SPY, so the SPY price on the 13Fs of three large banks (Northern Trust, Bank of America, JPMorgan Chase) is used, cross-checked every quarter.
• A few positions lack a price (for example when every giant sold out that quarter, or a split is unconfirmed) and are assumed to move with the rest of the portfolio; treating them as unchanged instead gives the composite +9.0% a year.

Limits

• Price change only, without dividends on either side; no fees or taxes.
• These are 13F holdings, not actual fund results: short positions, cash, fees, foreign shares and trades within the quarter are not visible.
• Nobody could have copied this: 13Fs arrive up to 45 days after the quarter ends.
• Survivorship bias: the roster consists of value investors who are still well known and still filing today; those who called themselves value investors back then and later closed are missing, so the earlier the year, the rosier the result.
• Holdings tables for 1999–2013 are plain text; after parsing, about 74% add up to the total on their cover page.

2. More than One Benchmark

Sources differ in period and method (with or without dividends and fees, time-weighted or money-weighted), so they are not plotted on one chart; each row gives its source and period.

• SPIVA counts funds that closed during the period; this site’s roster contains only giants still around today, so survivorship works in the opposite direction and flatters the giants.
• For 2006–2025, the S&P 500 with dividends and this site’s price-only SPY differ by about 2 percentage points, mainly dividends.

3. The Stocks the Giants Sold: Gain or Loss?

Scope: positions in which a giant invested an estimated $100 million or more while counted, and later sold completely. One stretch from first purchase to final sale counts as one position; buying again later starts a new one.

• Estimated return = estimated proceeds ÷ estimated amount invested − 1; both use (start + end price) ÷ 2, without dividends, fees or taxes, and round trips within a quarter are invisible.
• Positions already held in the first quarter of data take Q1 1999 (or the giant’s earliest electronic filing) as their starting cost, which is not what the giant paid, so they are shown separately.
• 128 positions that cannot be estimated (unconfirmed split, possible change of security code, no price in the quarter of sale) are left out.

4. What Became of Companies the Giants Once Held Heavily?

Scope: 1,093 companies that the giants together held $100 million or more of at some point since 1999, and that no giant holds in the latest quarter. Outcomes come from each company’s own SEC filings (for example 8-K item 1.03 = bankruptcy filing; 2.01/5.01 = acquisition; Form 25/15 = delisting).

Delisting is not failure: an acquisition often comes at a premium. A bankruptcy filing must also be checked case by case: parent or subsidiary, and whether shareholders were wiped out.

5. Case Studies: Successes, Failures and One Stock with Different Outcomes

• Only positions the giants have fully sold, in companies no giant holds in the latest quarter; being included is neither a recommendation nor a verdict.
• Every statement comes from 13F figures or the linked SEC filing (checked 2026-10-11). Amounts are estimates: invested = shares added each quarter × (start price + end price) ÷ 2, proceeds use the same estimated price, without dividends, fees or taxes; “SPY, same period” is estimated the same way.
• Shares of a company spun off by the company appear as a separate holding on a 13F and are not counted in the proceeds; prices in the charts are split-adjusted, and dots are estimated trade prices (buys and sells in this site’s up and down colours).

Failures and Risks (6)

Successes (6)

Same Stock, Different Outcomes (3)

6. The Boundaries of This Feature

• It covers only positions already closed and company events that have already happened, never stocks the giants hold now.
• It does not calculate or rank any individual giant’s overall performance, and offers no “what if you had followed them” back-tests.
• An “estimated loss” is only an estimate from 13F data, not the giant’s overall loss: hedges, short positions, bonds and foreign shares are not visible.
• More: Method and data, Disclaimer & data notice.