Skip to main content
The 13F Filings page ingests quarterly SEC 13F-HR filings from a curated list of 34 funds — long-only equity shops, hedge funds, multi-strats, and quants — and turns them into three views: what the crowd is doing (Consensus), what each manager is doing (By Fund), and how each manager’s book has actually performed since they reported it (Performance). Tracked managers include Tiger Global, Coatue, Lone Pine, Stanley Druckenmiller’s Duquesne, Berkshire, Pershing Square, Third Point, Greenlight, Hound Partners, Whale Rock, Citadel, Renaissance, Two Sigma, and more.

Consensus view

The default view. Two panels side-by-side.

Cluster Buys & Sells

Cross-fund signal: which CUSIPs the most managers piled into (or out of) last quarter. Filter tabs:
  • Top Buys — net buyer count and dollar flow positive
  • Top Sells — net seller count and dollar flow negative
  • New Positions — at least 3 funds initiated this quarter
Columns: The Max % column is the one to watch when buyer count is misleading. 8 buyers might sound like consensus — but if the max conviction is only 0.1%, nobody really cares. If max is 3.9% and the rest are 0.15%, it’s a single-fund bet wearing a consensus mask. Click any row to expand a drill-down showing each fund’s position, change direction, weight, and dollar delta. The expanded table is sortable.

Most Held

Same universe, different question: which names are held by the most funds — and where the most aggregate dollars sit. Less about flow, more about consensus exposure.

Q1 2026 Highlights panel (top of view)

Above the tables, a summary panel shows:
  • Top buys — best aggregate-flow names with lead funds named
  • Top sells — mirror logic
  • Cluster new positions — names ≥3 funds just initiated
Lead funds shown are filtered for conviction: positions ≥0.25% portfolio weight, with quant/multi-strat managers held to a higher 1.0% bar so basis-point market-neutral noise doesn’t dominate the picture.

By Fund view

Switch the top-of-page toggle to By Fund. Left sidebar lists all 34 funds; clicking one shows its drill-down:
  • Q/Q highlights — top new positions, biggest increases, biggest decreases, exits
  • Full position table — every name in the fund’s filing, sortable by weight, value, share change, etc.
  • Per-row history — click a position for a multi-quarter history of that fund × ticker

Performance view

Switch the top-of-page toggle to Performance. This treats each fund’s reported long-equity book as a hypothetical value-weighted, buy-and-hold clone and tracks how it has done — marked to the latest close — anchored at two dates that sit on every filing:
  • Effective date — the quarter-end the holdings reflect. What the book has done since the snapshot.
  • Filing date — when the 13F actually hit EDGAR (~30–45 days later). What a follower could realistically have captured, since the holdings weren’t public until then.
The gap between the two is lag decay — the edge that’s already gone by the time you can act on a public filing. That contrast is the whole point of this view.

Leaderboard

Every tracked fund, ranked. Sortable; click a row to load that fund’s detail below. Filter pills: All / Beating / Lagging the benchmark. The Benchmark toggle (top-right of the panel) switches the whole view between SPY and QQQ — the vs-column, the filter, the detail tiles, and the chart line all re-rank live.

Fund detail

Clicking a fund opens:
  • Stat tiles — Since Effective, Since Filing, alpha vs the benchmark, max drawdown, hit rate (share of names up), and best/worst contributor.
  • Equity curve — the clone rebased to 100 at the effective date vs the benchmark, with E (effective) and F (filed) flag markers and an underwater drawdown sub-panel.
  • Holdings contribution — each name’s weight, price at filing, latest price, return, and contribution (weight × return). Filter to winners or losers.
  • Quarter chips — jump between the fund’s recent filings to see how older books played out.

Methodology & caveats

A deliberately honest, hypothetical replica — read it as direction and decay, not a live track record:
  • Value-weighted, buy-and-hold — weights are each name’s reported dollar value; no rebalancing.
  • Long US equity only — no shorts, options, leverage, or foreign listings; gross of fees and taxes.
  • Coverage % is shown because it matters — names we can’t price (delisted, foreign, deep tail) are excluded, so concentrated funds read ~95%+ and sprawling multi-strats lower.
  • Prices come from Polygon and are de-spiked against bad single-day ticks; the leaderboard re-marks to the latest close daily.

Period selector

Top-right dropdown — defaults to the latest quarter where most funds have filed. Drop back to prior quarters to compare flows over time. (In the Performance view, use the per-fund quarter chips instead.)

Daily refresh

A Railway cron service polls EDGAR daily for new 13F-HR filings from the tracked CIK list. New filings parsed, CUSIPs resolved (via OpenFIGI), and the views update automatically. There’s typically a 30-45 day lag between quarter-end and the bulk of filings appearing, then a long tail of stragglers and amendments.

What 13F is and isn’t

  • Is: long equity positions ≥ $100M AUM, filed within 45 days of quarter-end.
  • Is not: shorts (not required), options exposures (only US-listed derivs are partial), foreign holdings (US-only), or current positioning — there’s always a quarter-end lag.
Treat 13Fs as direction signal, not timing. A position you see today was held at some point in the last 4-7 weeks; the manager may have already adjusted.

What you can do with it

Find conviction names

Sort Cluster Buys by Max % to see what funds really care about.

Compare manager styles

By Fund view side-by-side reveals growth vs value vs macro vs quant signatures.

Spot rotation themes

Cluster New Positions surfaces the early-stage consensus before it shows up in price.

Track one manager over time

Per-position history reveals when they sized up or trimmed.

See whose picks actually worked

The Performance view ranks every fund’s clone return and shows how much edge decays in the reporting lag.