What Is a 13F?
7 min read·Reviewed by the StockTools.ai Research Team
- ▸Institutional managers with $100 million or more in US-listed stocks must file Form 13F within 45 days of each quarter end, so first-quarter positions surface around May 15.
- ▸A 13F lists long positions with share counts and quarter-end market values only: no short positions, no entry prices, no trade dates.
- ▸Compare share counts across quarters, never dollar values — price moves change a position's value without a single share being traded.
- ▸A disclosed long can be one visible leg of a hedged trade whose offsetting short never appears, so conviction is easy to misread.
- ▸By the time a filing is public, the data is 45 to 135 days old and the position may already be gone — 13Fs generate ideas, not entry signals.
The rule: who files, what, and when
Form 13F is a quarterly SEC disclosure required of institutional investment managers who exercise discretion over $100 million or more in '13F securities' — mostly US exchange-listed stocks and ETFs, plus certain options and convertibles on the SEC's official list. Hedge funds, mutual fund advisers, pensions, insurers, and large family offices all cross the threshold; the filing goes to EDGAR, where anyone can read it free.
The deadline is 45 days after quarter end. A quarter ending March 31 must be reported by May 15, June 30 by August 14, and so on. That lag is not a formality — it is the defining feature of the document. Everything in a 13F was true as of a quarter-end snapshot at least a month and a half before you see it, and possibly much longer for positions built early in the quarter.
The rule exists for market transparency, not for copy trading. Congress added Section 13(f) in 1975 so regulators and the public could see where institutional money was concentrated. The fact that it became the raw material for whale watching is a side effect, and reading one well means staying aware of what the form was never designed to show.
What is inside — and what never shows up
Each line of a 13F carries an issuer name, the class of security, the number of shares, and the position's market value at the quarter-end close. Options positions in covered names appear flagged as puts or calls, reported by the share count of the underlying. That is the entire inventory: a list of longs with sizes, priced at one moment.
The list of absences is longer than the list of contents. No short positions of any kind. No cash balances, no bonds (with narrow exceptions), no foreign-listed shares, no futures, no swaps, no currency positions. No entry prices, no trade dates, no profit or loss — you learn a fund held 2 million shares on March 31, not when it bought them, at what price, or whether it was buying or unloading on April 1. The SEC can also grant confidential treatment, letting a manager delay disclosure of a position still being accumulated.
Framing matters here: a 13F is an 'as filed' quarter-end snapshot of one side of one slice of a portfolio. For a long-only concentrated manager, that slice may be nearly the whole book. For a multi-strategy fund running shorts, swaps, and international positions, the 13F might describe a third of the portfolio with the hedges amputated.
Reading position changes across quarters
The single most common reading error is comparing dollar values across quarters. Values move with price whether or not the manager traded, so the share count is the only line that records decisions. Take a fund that reported 2,000,000 shares of a stock at $45 on December 31, a $90.0 million position, and then 2,600,000 shares at $40 on March 31, a $104.0 million position.
The value grew $14 million, which reads like modest adding. The share count tells the real story: the fund bought 600,000 more shares, a 30% increase, while the price fell 11% — deliberate accumulation into weakness, a far stronger statement than the dollar change suggests. The illusion runs the other way too: a position can grow 20% in value while the share count shrinks, meaning the manager was selling into a rally that outran the trimming.
The loudest signals are the edges: brand-new positions, complete exits, and doublings, especially from managers who run few names and hold for years. Sequences matter more than single quarters — three consecutive quarters of adding is a thesis; one quarter is an ambiguous data point. Watch also for amended filings (13F/A) that restate a quarter, and for 13D or 13G filings, which are separate forms triggered by crossing 5% ownership and arrive far faster than the 13F cycle.
Longs without their shorts
Because shorts are invisible, a 13F long is not automatically a bullish bet. Merger arbitrage is the cleanest illustration. Suppose an acquirer offers 0.5 of its own shares for each share of a target, and a fund buys 1,000,000 shares of the target while shorting 500,000 shares of the acquirer. The 13F dutifully reports a large new long in the target — which every screener will present as conviction in that company.
The fund has no view on the target's business at all. It is betting the deal closes: if it does, the long and short converge and the spread is captured regardless of where either stock trades. If the deal breaks, the position loses even if the target's stock merely drifts. An investor who cloned the visible leg took a completely different trade than the fund did — directional exposure the fund deliberately built the trade to avoid.
The same distortion appears wherever a long is half a structure: convertible arbitrage funds short shares against convertible bonds (bonds that mostly do not appear in the filing), pairs traders hold the long leg visibly and the short leg invisibly, and index-hedged books show their single-name longs without the offsetting futures. The bigger and more multi-strategy the filer, the less any individual line means on its own.
Why cloning is harder than it looks
Staleness has a price you can compute. A manager who bought in mid-January at $30 files on May 15, when the stock might trade at $42 — the cloner pays 40% more than the fund did, with no way to know the fund's basis or whether the position survived April. Studies of 13F-cloning strategies show the approach can work for slow-turnover managers precisely because their positions are still there two quarters later; for anyone who trades actively, the filing describes a portfolio that no longer exists.
Position sizing does not translate either. A $50 billion fund's 2% position is $1 billion parked in a liquid large cap it may hold for five years, sized against its liquidity constraints and its other 40 positions — none of which describes your account. The filing also carries no stop, no thesis, and no exit plan, so a cloner who follows a whale in has nothing to follow on the way out except a filing that arrives 45 days after the whale already left.
Two further distortions: quarter-end window dressing, where some managers tidy the book right before the snapshot, and survivorship in whale lists — the famous filers are famous because they performed, and the many funds whose 13Fs would have cloned badly are not on anyone's dashboard. None of this makes the data worthless. It makes the data an input, with error bars the form itself cannot shrink.
What 13Fs are actually good for
Used as an idea filter rather than a buy list, the 13F earns its keep. The strongest reads come from concentrated, low-turnover, long-only managers, where the filing approximates the real book and a new position reflects months of research by a team with resources most individuals lack. Multi-quarter accumulation by several unrelated managers of that type is a genuinely interesting flag; a single quarter's line at a 200-position multi-strat is closer to noise.
The mechanics of following along are free. Filings sit on EDGAR, and StockTools' Whale Portfolios tracker assembles them into readable portfolios with quarter-over-quarter changes, so watching share counts move across filings takes minutes instead of a spreadsheet session.
The honest workflow: treat a whale's position as a lead, do the work the filing cannot do for you — the business, the valuation, the reason the position might already be stale — and if the idea survives, size it for your own account rather than mirroring theirs. The position size calculator handles that last step; the 13F was only ever the tip.
FAQ
When are 13F filings published?
Within 45 days after each calendar quarter ends: mid-May for Q1, mid-August for Q2, mid-November for Q3, and mid-February for Q4. Most large funds file near the deadline, so the interesting reading arrives in a burst around those dates on EDGAR.
Do 13Fs show short positions?
No. The form covers long positions in US-listed 13F securities only. Shorts, swaps, futures, cash, and most bonds are absent, which means a visible long can be one leg of a hedged trade — a merger-arb fund's target-company long, for example — rather than a directional bet.
Can I see what price a fund paid for its shares?
No. A 13F reports share counts and market value at the quarter-end close, nothing about cost basis or trade timing. A position filed in May could have been bought at any point during the first quarter, at prices possibly 30-40% away from where the stock trades when you read the filing.
Why did a fund's position value change if the share count stayed flat?
Price movement. Market value is shares times the quarter-end price, so an untouched 2,000,000-share position moves from $90 million to $104 million if the stock rises from $45 to $52 — hence share counts, not dollar values, are the line that records actual buying and selling.
Is copying 13F portfolios a good strategy?
Copying works best, when it works at all, with concentrated low-turnover managers whose filings still resemble their live books after the 45-day lag. It fails with active traders and hedged funds, where the disclosed long may be stale or half a structure. Better use: treat filings as a research pipeline, verify the thesis yourself, and size positions to your own account.
Put it to work
More to learn
Educational only — not financial advice. Concepts simplified for clarity; markets are messier than definitions.