A 13F filing is the SEC form every institutional investor managing over $100M must file each quarter. It lists the equity positions they held on the last day of the quarter. It's free, public, and one of the most-used (and most-misunderstood) signals in retail investing.
This post explains the format, the what-13Fs-don't-show (which is more important than what they do), and how to use the data with the appropriate skepticism. invest-like surfaces 13F data on every stock page at /institutional/[ticker]/ and aggregates 13Fs by institution at /institutional/.
What a 13F filing actually reports
Each 13F entry is one institutional investor + one stock. The investor lists:
- Investor name (the institution: Berkshire Hathaway, BlackRock, Citadel, etc.)
- Stock CUSIP and ticker
- Number of shares held on the last day of the quarter
- Market value of position (shares × end-of-quarter close)
- Sole or shared voting authority
- Type of security (common stock, ETF, derivative, etc. - but 13F only covers long equity positions; see "What 13F doesn't show" below)
That's it. The filing is a snapshot of long equity holdings at a single point in time.
What 13F doesn't show (the important part)
Here's what most retail investors miss:
-
Short positions. 13F is long only. A hedge fund may report a 3% Apple long position while simultaneously holding a 4% Apple short via a swap or options structure. The 13F shows the long; you don't see the offset. Net exposure can be wildly different from the gross reported.
-
Cash and bonds. 13F covers only listed equities. A fund could have 70% cash and you'd never know. Berkshire's 13F shows ~$300B of equity holdings; their $300B+ cash position is invisible to 13F.
-
Foreign listings. 13F requires US-listed equities only. A fund's positions in TSMC (Taiwan), ASML (Netherlands), Nestlé (Swiss), or Toyota (Japan) typically don't show up unless held via ADR.
-
Private positions. PE-style positions are invisible.
-
Derivatives. Options, swaps, and structured products generally don't appear on 13Fs (with some exceptions for cash-settled options reported by some filers).
-
The 45-day lag. 13Fs must be filed within 45 days of the quarter end. By the time you read a 13F, the data is at minimum 6 weeks old. The institutional investor may have completely changed the position since the snapshot.
-
Filer's own bias on reporting. Some institutions request "confidential treatment" on certain positions (typically while they're still building or unwinding a major position). These positions appear later, retroactively, in an amended 13F.
The cleanest single sentence on 13F limitations: a 13F tells you what the institution owned 6 weeks ago, long only, US equities only, with no hedges or cash visible.
How to actually use 13F data
Despite all the limitations, 13Fs are still useful for three specific use cases:
Use case 1: Tracking value-investing-aligned institutions
If you respect Berkshire Hathaway's investing philosophy, knowing what Berkshire owned six weeks ago is more useful than knowing what some no-name asset manager owned, because Berkshire's positions are large, slow-turning, and value-investing-aligned. The 6-week lag matters less when the typical holding period is 5+ years.
Useful institutions to track:
- Berkshire Hathaway (BRK.B 13F)
- Markel Group (MKL - Buffett-style philosophy at smaller scale)
- Fundsmith (Terry Smith - UK filing, slightly different format)
- Stochastic Risk Premia and Pure Quant funds (smaller institutions)
- Pat Dorsey's Dorsey Asset Management (wide-moat focus)
For tracking opportunistic fast-turnover hedge funds (Citadel, Renaissance, Two Sigma, etc.) - 13F is much less useful because their turnover is high and the 6-week lag means the data is stale.
Use case 2: Identifying conviction positions
When a 13F shows an institution with a > 5% position in a single stock, that's a high-conviction signal. The institution has done extensive research and concentrated meaningful capital. Combine this with the 7-framework consensus and you have a powerful filter.
Worked example: Markel's 13F shows ~8% of its equity book in Berkshire Hathaway itself, ~6% in Diageo, ~5% in Brookfield Asset Management. Cross-reference with our 7-framework consensus: Diageo passes 5-of-7, BAM passes 5-of-7. Conviction × framework agreement = real signal.
Use case 3: Identifying capitulation / exit signals
When a famous investor exits a position, that's often more informative than them entering one. Buffett's 13F has historically signalled major macro turn points when he aggressively reduces equity weight (e.g., dramatic position trimming in late 2022 / early 2023 was a leading indicator of his 2024 reduction in Apple).
When an institution with a 5-year average holding period suddenly exits a position, ask why.
Common 13F traps
-
Mistaking a 13F entry for "the fund is bullish right now". The data is 6 weeks old. The fund may have already sold.
-
Copying small positions from a "famous" investor. A 0.5% position in a fund's equity book is not high-conviction. A 5%+ position is.
-
Ignoring the cash and short side. A fund's 13F long equity could be tiny relative to its overall positioning.
-
Comparing different filers' 13Fs to each other directly. A 5% position in Berkshire's book has different meaning than a 5% position in Citadel's book - Berkshire concentrates, Citadel hedges.
-
Treating new positions as "buy signals". A hedge fund's new position could be a hedge for an offsetting short, a Rule-Specific basis trade, or a position they're already exiting in real time.
Where invest-like surfaces this data
Every stock page on invest-like shows:
- Institutional holders of that ticker (top 25 by position size)
- Filing date of each institution's 13F (so you can see the lag)
- Change since previous quarter (new position / increased / reduced / closed)
For aggregated institutional tracking:
Disclosure
Educational tool. 13F data is sourced from EDGAR (SEC). All limitations described above apply to every 13F. Past 13F filings do not predict future returns; institutions can completely change positions in the 6-week reporting lag. Berkshire Hathaway, Markel Group, and Fundsmith are named as analytic examples, not investment recommendations.
Author: Zaid Ghazal, founder of invest-like, indie SaaS, Kiel, Germany.