The Limits of 13F Data: What Filings Don't Tell You

Quarterly 13F filings are the best free window into what large investment managers own. They are also easy to over-read. The form was created by Congress in 1975 to give regulators visibility into institutional holdings, not to help retail investors clone portfolios, and its blind spots are large enough that a filing can sometimes suggest the opposite of what a manager actually believes. If you already know the basics of how 13Fs work, this guide covers what the form leaves out and how to work around it.

Short positions never appear

Form 13F reports long positions in so-called 13F securities and nothing else. Short sales are not disclosed, on this form or anywhere else in routine US filings. That has a sharper consequence than most people appreciate: a manager can be net short a company while their 13F shows a long position in it.

Merger arbitrage is the cleanest example of the pattern. An arbitrageur who is long the target and short the acquirer files a 13F that shows only the long leg, so the acquirer position simply vanishes and the target position looks like a conviction bet on the business rather than a bet on a deal spread. Similarly, a fund can hold common shares while carrying a larger bearish exposure through swaps or single-stock futures, none of which show up. When a filing from a long-short fund looks bullish on a name, the honest reading is that you can see half the trade at most.

A long position may be one leg of something else

This is the standard caveat attached to almost every serious discussion of 13F data, and it deserves repeating because it applies even to positions that are genuinely long. A reported holding may be one side of a pairs trade, a hedge against a related exposure, or part of a capital-structure trade such as convertible arbitrage, where the manager holds a convertible security and shorts the underlying stock. In each case the position exists to offset something you cannot see. The filing gives you the instrument, not the thesis.

You cannot see cash

Cash, money market funds, and Treasury bills are not 13F securities, so they are absent from the filing entirely. Portfolio percentages on any 13F tracker, including this one, are calculated against the reported securities only. A manager who has quietly moved 40 percent of assets into T-bills looks exactly as fully invested as one running zero cash. There is no way to measure how defensive a fund is from its 13F, and a shrinking filing total can mean redemptions, losses, or a deliberate retreat to cash, with no way to tell which from the form itself.

Bonds and foreign listings are mostly missing

The official 13F securities list covers US exchange-traded stocks, ADRs, certain ETFs and closed-end funds, some convertible bonds, and exchange-listed options on those securities. Treasuries, corporate bonds, municipal bonds, private holdings, and shares that trade only on foreign exchanges are all outside it.

The practical effect is that managers with large foreign books look smaller than they are. A global fund with most of its capital in Tokyo, London, or Mumbai listings files a 13F covering only its US-listed sleeve, which may be a modest and unrepresentative slice of the real portfolio. Comparing that fund's reported total to a domestic-only manager on the manager list tells you about their US exposure, not their size or their best ideas. Foreign positions held through ADRs do appear, which is why some international holdings show up while economically similar ones do not.

The 45-day lag and the snapshot problem

Filings are due 45 days after quarter end, and most managers file close to the deadline. By the time you read a new filing, the positions are between 45 and 135 days old depending on when in the quarter they were established, and there is no requirement to still hold them. For a low-turnover value investor that lag barely matters. For a fund that trades weekly, the filing describes a portfolio that may no longer exist.

The deeper issue is that a 13F is a single-day photograph. Anything bought and sold within the quarter never appears at all, so a manager could take a large position in January, exit in March, and leave no trace. Quarter-end snapshots also invite window dressing: some funds tidy the portfolio before the photo, adding recent winners or trimming embarrassing losers so the filing reads better than the quarter actually went. You cannot detect this from one filing, which is one reason position changes tracked across quarters in the activity feed are more informative than any single snapshot.

Options rows need care

Listed options on 13F securities are reported in an unusual way: the filer lists the underlying stock, flags the row as PUT or CALL, and reports the number of underlying shares and the market value of those underlying shares, not the option premium. A fund that spent a few million dollars on puts can therefore show a row that looks like a nine-figure stock position.

Two mistakes follow from this. The first is adding option share-equivalents to common stock and calling the sum a position size, which wildly overstates the capital actually at risk. The second is reading a PUT row as bullish because it appears in the portfolio like any other holding, when it usually reflects a bearish or protective stance. There is also an asymmetry worth knowing: only options the fund holds are reported. Options the fund has written, such as covered calls or cash-secured puts, do not appear, so even the options picture is one-sided.

Amendments, restatements, and confidential treatment

Filings are not always final. Managers file amendments to correct errors or add omitted holdings, and an amendment can restate the entire report, meaning the version you analyzed at the deadline may differ from the version that stands as the record. Data errors are also more common than you might expect: wrong identifiers, misplaced decimal points, and unit mistakes, a problem the SEC's 2023 switch from reporting values in thousands of dollars to whole dollars did not fully retire.

Separately, a manager can ask the SEC for confidential treatment of a position while it is being accumulated. When granted, the position is omitted from the public filing and disclosed months later in an amendment. This means the most interesting position in a portfolio may be exactly the one that is missing, and a well-known holding can appear to materialize retroactively in a quarter you already studied.

What you see and what you don't

Visible in a 13F	Invisible in a 13F
Long positions in US-listed stocks and ADRs	Short positions of any kind
Held puts and calls, as underlying share equivalents	Written options, swaps, futures
Certain ETFs, closed-end funds, convertibles	Cash, Treasuries, most bonds
Quarter-end holdings, 45 days late	Intra-quarter trades, foreign-only listings, private stakes

How to use the data anyway

None of this makes 13F data useless. It makes it data that rewards the right questions. A few habits help.

• Favor long-only, low-turnover managers. For a concentrated stock picker who rarely trades and runs little or no short book, the 13F is a close approximation of the real portfolio. Berkshire Hathaway's portfolio is the canonical case: what you see is very nearly what they own in US equities. For multi-strategy and quantitative funds, per-name readings are close to meaningless.
• Read across quarters, not within one. A position held and added to over several filings signals conviction in a way one snapshot cannot. It also filters out window dressing, which is hard to sustain for a year.
• Weight buys above sells, and both lightly. A new purchase is a deliberate act. A sale might be a trim, a redemption, or rebalancing, so a name appearing among the most sold stocks deserves investigation before interpretation.
• Treat every filing as a starting point. The useful output of 13F research is a watchlist of ideas that intelligent investors found worth owning at some point in the recent past. The valuation work, and the decision, still belong to you.

Held with that kind of skepticism, 13F data is genuinely valuable: a free, structured record of where serious capital has been going. Just remember that the form shows one side of one slice of a portfolio, six to nineteen weeks after the fact. Nothing in it is a recommendation, from the filer or from anyone summarizing them.