The 10 Hedge Fund Managers Whose 13Fs Actually Predict Returns (And Who Owns What Right Now)

Of 2,302 distinct stocks held by eleven elite portfolio managers as of the Q4 2025 13F cycle, exactly one is held by seven of them: Amazon. Five more are held by four. The other 2,296 are held by three or fewer. The "consensus pick" of the world's most-watched 13Fs is not a list. It is a single name, and a thin tier of follow-ons.

The academic record on whether copying these 13Fs generates alpha is messier than the marketing implies. It exists. It has decayed. The decay is roughly proportional to how many people read the same filing. This post answers the question the title raises — which managers, and which holdings — and grounds it in what the peer-reviewed literature actually shows.

Executive summary

Eleven elite portfolio managers — Berkshire Hathaway, Pershing Square, Baupost, Third Point, Soros, Point72, Lone Pine, Greenlight, Tiger Global, Elliott, Duquesne — filed Q4 2025 13Fs by February 17, 2026. Across every common-stock position, only Amazon is held by seven of them. Five issuers are held by four: Alphabet, Taiwan Semiconductor, Microsoft, Restaurant Brands International, and Kenvue. Documented 13F alpha is real but small.

Why "predict returns" is the wrong frame, and what the evidence actually shows

The phrase "13Fs that predict returns" overstates the literature. Three findings frame what is and is not true.

First, Brav, Jiang, Partnoy and Thomas (2008), the foundational study of activist hedge fund Schedule 13D filings, document an average abnormal return of approximately 7% in the (–20, +20) window around a 13D announcement. Across roughly 888 hedge fund 13D events from 2001 to 2006, this announcement-window return declined from 15.9% in 2001 to 3.4% in 2006 as activist competition intensified. Bebchuk, Brav and Jiang (2015) extend the sample and find no long-term reversal — the announcement gain persists. Activist 13D filings still carry signal. The signal is real and the signal is shrinking.

Second, Cohen, Polk and Silli (2010) "Best Ideas" document that the single largest active position in a manager's portfolio outperforms the market by 2.8% to 4.5% per year after risk adjustment. Antón, Cohen and Polk (2021) extend the result to hedge funds and find that small-fund best ideas outperform large-fund best ideas by approximately 15% per year. This is the most important and least-quoted finding in the 13F-copying literature: the alpha is concentrated in conviction, not coverage.

Third, Angelini, Iqbal and Jivraj (2019), "Systematic 13F Hedge Fund Alpha", test a portfolio-tilt strategy assuming a realistic 47-day lag after quarter-end before rebalancing. The strategy delivers approximately 3.80% annual outperformance at a Sharpe ratio of 0.69. This is the cleanest published estimate of what a 13F-copying strategy nets after disclosure-delay decay. It is positive, modest, and far below the marketing claims.

The honest reading: 13Fs from elite managers contain alpha, but the alpha is concentrated in high-conviction positions (which require knowing which manager you are watching) and decays as the strategy is adopted. The post below is a screen, not a model.

The eleven elite managers tracked

The 13dwatch pipeline ingests the full 13F filing of each manager from SEC EDGAR (https://www.sec.gov/edgar) within hours of disclosure. Filers are selected by a documented track record: hedge fund activism with measurable post-13D abnormal returns (Pershing Square, Third Point, Elliott), best-idea concentration (Berkshire Hathaway, Baupost, Lone Pine, Greenlight), or institutional reputation as bellwether discretionary capital (Soros, Point72, Tiger Global, Duquesne). Inclusion is a track record claim, not a forward-looking endorsement. Performance is not guaranteed by past returns and the past returns themselves are contested.

Source: live query of institutional_holdings table, 13dwatch data pipeline, 2026-05-07.

Three caveats are visible in the table itself. Greenlight's most recent ingested 13F is dated 2023-12-31 — Einhorn's firm has reportedly reduced its 13F-eligible exposure or filed under an exemption; the data is stale and we flag it explicitly. Tiger Global's single position reflects the firm's well-documented pivot toward private holdings. Baupost's $11M figure is a partial-sleeve disclosure — the headline AUM is far larger but most positions are not 13F-reportable. The lesson is that AUM-weighted analyses of a 13F panel are misleading; holder count is the more honest measure.

The cross-PM consensus picks

Aggregating every common-stock position across the eleven managers' latest 13Fs (excluding options and derivatives) yields 2,302 distinct issuer/CUSIP keys. The distribution of holder counts:

A single name is held by seven of eleven elite managers. Five more by four.

The one stock held by 7 of 11 elite PMs: Amazon

Amazon (NASDAQ: AMZN, CUSIP 023135106) is held by Baupost Group, Berkshire Hathaway, Lone Pine Capital, Pershing Square, Point72, Soros Fund Management, and Third Point. Total disclosed value across the seven filings is approximately $6.32 billion representing roughly 29.76 million shares. Pershing Square's stake is among the largest single 13F-disclosed Amazon positions in the panel; Berkshire's stake originated under Todd Combs, not Buffett directly. The position appears across activist (Third Point), best-idea-concentrated (Lone Pine, Pershing Square), and value (Baupost) styles — a rare instance of cross-style consensus.

Held by 4 elite PMs

Note: Alphabet Class C (CUSIP 02079K107) appears separately in our consensus rollup at three holders ($2.02B). Combining share classes pushes Alphabet to a five-of-eleven consensus, second only to Amazon.

Held by 3 elite PMs (top 10 by combined $-value)

The Apple line is dominated by Berkshire's stake. Bank of America similarly. The list is mega-cap heavy because three filers in the panel — Berkshire, Point72, and Soros — hold the vast majority of the U.S. mega-cap complex by default. The interesting consensus picks are the smaller-AUM names (Restaurant Brands, Kenvue, Brookfield) where the overlap is independent rather than mechanical.

How to read this list as a hedge fund analyst

Three operating heuristics, ranked by what the literature supports.

Treat consensus as a screen, not a portfolio. Cohen-Polk-Silli show that high-conviction individual positions are where alpha lives. Cross-PM consensus narrows the universe but does not by itself match the conviction signal. The right next step is to pull each consensus name's filer-by-filer concentration: a 5% Pershing Square position carries more signal than a 0.05% Point72 position.

Discount the post-disclosure window. The Form 13F filing deadline is 45 days after quarter-end (SEC Rule 13f-1(a)(1)). For Q4 2025, that deadline was Feb 17, 2026. By the time you read the consensus list above, three months have passed and the position may have been exited. Angelini-Iqbal-Jivraj's 47-day-lag backtest is the relevant benchmark; assume something close to their 3.80% annual outperformance, not the gross figure that ignores delay.

Watch for cross-source confirmation. Activist 13D filings have their own announcement-window alpha (~7% historically, declining). When a stock appears on the elite-PM consensus and is the subject of a recent 13D filing and shows insider buying, the cluster compounds — see our activist + insider cluster study for the database evidence.

Frequently Asked Questions

Do hedge fund 13Fs predict returns?

The peer-reviewed evidence is mixed. Activist 13D filings (Schedule 13D, not Form 13F) have a documented announcement-window abnormal return averaging 7% over 2001–2006, declining to roughly 3% by the end of that period (Brav, Jiang, Partnoy and Thomas 2008). Quarterly Form 13F portfolios from elite managers contain alpha primarily in their highest-conviction positions, where Cohen, Polk and Silli (2010) document 2.8%–4.5% annual outperformance for "best ideas." Naive copying of full 13Fs at a 47-day lag yields approximately 3.80% annual outperformance (Angelini, Iqbal and Jivraj 2019).

What is the deadline for filing Form 13F?

Institutional investment managers exercising discretion over $100 million or more in Section 13(f) securities must file Form 13F within 45 days after the end of each calendar quarter, per SEC Rule 13f-1(a)(1). For 2026, the four filing deadlines are February 17, May 15, August 14, and November 16.

How do you copy a hedge fund's 13F portfolio?

Three constraints apply. First, the 45-day disclosure lag means the portfolio you copy is at least 45 calendar days stale and possibly more. Second, derivative positions, short positions, and non-U.S. equity holdings are not reported on Form 13F, so the disclosure is incomplete. Third, the academic literature suggests the alpha is concentrated in the manager's highest-conviction position rather than the full portfolio. The most realistic implementation is a portfolio-tilt rebalanced on a 45–60-day lag using a small basket of high-conviction or high-consensus names.

Which hedge fund manager has the best 13F track record?

There is no single answer. By documented activist-filing announcement returns, Pershing Square (Ackman), Third Point (Loeb), and Elliott Investment Management (Singer) have measurable track records. By concentrated best-idea performance, Berkshire Hathaway, Baupost, and Lone Pine have multi-decade records. By Sharpe-adjusted long-equity performance, Renaissance Technologies (Medallion is closed), Citadel and Millennium are widely cited but file less interpretable 13Fs because of position counts and short discretion.

Why is Amazon held by 7 of 11 elite portfolio managers?

Amazon is the only stock held by seven of the eleven elite managers tracked here. The convergence reflects three independent drivers: Pershing Square's concentrated activist-style position; Berkshire Hathaway's stake (originated by Todd Combs); and Point72, Soros and Third Point's market-neutral or thematic exposure. Cross-style independent convergence (activist, value, multi-strat) on a single name is rare. Whether this constitutes a forward-looking buy signal is contested — see Cohen, Polk and Silli (2010) for the case that conviction matters more than overlap.

What is the difference between Form 13F and Schedule 13D?

Form 13F is a quarterly portfolio disclosure required of institutional managers with $100 million or more in Section 13(f) securities, due 45 days after each quarter-end. Schedule 13D is an event-driven disclosure required when an investor acquires beneficial ownership exceeding 5% of a registered class of equity with control intent, due within 5 business days of crossing the threshold (per the SEC's October 2023 amendments). 13F discloses positions; 13D discloses intent. See our 13D vs 13G vs 13F explainer.

Methodology and limitations

Source data is the SEC EDGAR database (https://www.sec.gov/edgar). 13F-HR (and 13F-HR/A amendment) filings for each tracked manager are parsed from the official informationtable.xml document. The 13dwatch pipeline aggregates positions by (cusip, title_of_class) to handle multi-manager consolidated 13Fs. Options and derivative positions (putCall populated) are excluded from the consensus aggregation; only common-stock positions count. The consensus is computed from the latest 13F period for each manager — for Berkshire, Point72, Pershing Square, Soros, Third Point, Elliott, and Baupost, the latest period is 2025-12-31. Lone Pine and Duquesne were last filed for 2025-06-30. Tiger Global's most recent filing covers 2025-09-30. Greenlight's most recent ingested 13F is 2023-12-31 — likely reflecting reduced 13F-eligible exposure or confidential treatment; this is a known data gap.

Position values reflect manager-reported value field at quarter-end mark. Manager AUM as a whole exceeds the disclosed 13F value because many strategies (private investments, short books, non-U.S. equity) are not 13F-eligible. AUM-weighting the consensus is therefore misleading; we use unweighted holder count.

This is research, not investment advice. Past performance does not predict future returns. Long Street Consulting LLC publishes 13dwatch as institutional intelligence; it is not a registered investment advisor and does not make recommendations.

Get the live elite-PM consensus feed

13dwatch publishes the cross-PM consensus list, refreshed within hours of each new 13F drop, alongside live activist filings, insider transactions, and short-interest cross-references. The next 13F deadline is August 14, 2026 for Q2 2026 data. Request a pilot — $1,500/mo or a 20-minute call.