Institutional Ownership Percentage, Explained
What "X% institutionally owned" really measures, how 13F aggregation produces the number, and why different screeners disagree on the same stock.
The short version
A stock's institutional ownership percentage is the fraction of its shares outstanding reported as held by professional investment managers, and it is built almost entirely by aggregating SEC Form 13F filings. Any manager with $100 million or more in covered US equities must disclose its long positions quarterly, within 45 days of quarter end. Sum every filer's reported shares of a stock, divide by shares outstanding, and you have the headline number. Sources disagree on the same stock because each one makes different choices about the denominator, amendments, option legs, and double counting.
What "X% institutionally owned" actually claims
When a screener labels NVIDIA as, say, two-thirds institutionally owned, it is making a much narrower claim than most readers assume. "Institutional" here has a precise regulatory meaning: an institutional investment manager exercising investment discretion over at least $100 million in 13F securities (mostly US-listed equities and listed options), which obligates it to file Form 13F with the SEC every quarter. That bucket covers hedge funds, mutual fund complexes, pension managers, bank trust departments, insurance companies, and large registered investment advisers.
The numerator of the percentage is the sum of shares those filers reported holding at the most recent quarter end. The denominator is usually the company's shares outstanding, sometimes its public float. Neither side of the fraction is live data. The numerator is a quarter-end snapshot disclosed up to 45 days later, the denominator is whatever share count the vendor chose, and the two are often measured on different dates.
Three things the number does not include: short positions (13Fs disclose longs only), holdings of managers below the $100 million threshold, and most non-US institutions with no filing obligation. Insider stakes appear only if the insider also happens to be a 13F filer; otherwise founders and executives sit outside the figure entirely, in a separate disclosure regime built on Forms 3, 4, and 5. So a stock described as "70% institutional" is really "70% of the share count we picked was reported, as of a past quarter end, by managers large enough to file."
Where the number comes from: 13F aggregation
There is no central feed of institutional ownership. The figure is manufactured, by every vendor independently, from thousands of individual filings on SEC EDGAR. The pipeline looks like this: each quarter, every 13F filer submits an information table listing security name, CUSIP, share count, and market value for each long position. The 2026 deadlines are February 17, May 15, August 14, and November 16, which is quarter end plus 45 days adjusted for weekends and holidays (full calendar in 13F filing deadlines for 2026).
An aggregator then groups every filer's rows by security identifier, almost always the CUSIP, because the issuer-name field is free text and unreliable. Sum the shares across filers, fetch a shares-outstanding figure for the company, divide. The SEC's own Form 13F FAQ documents the filing rules; everything past that point (the cleaning, the joining, the amendment handling, the deduplication) is vendor work, and vendors do it differently.
The scale of that work is the point. Arkolith's Q1 2026 dataset spans 1,824 institutional filers reporting 1.87 million long positions worth $53.7 trillion. A single mega-cap stock appears in a large share of those filings, each with its own spelling of the issuer name, its own amendment history, and occasionally its own unit errors. The institutional ownership percentage you see on any site is the output of all those merge decisions, which is exactly why the next section exists.
Why sources disagree on the same stock
Pull up one ticker on three screeners and you will routinely get three different institutional ownership percentages, sometimes spread surprisingly wide. None of them is necessarily wrong; they answered different questions. The main forks in the pipeline:
| Pipeline decision | Common choice A | Common choice B | Effect on the percentage |
|---|---|---|---|
| Denominator | Shares outstanding | Public float | Float-based figures run higher |
| Amendments | Effective post-amendment filing only | Originals plus amendments | Double counting inflates the number |
| Option legs | Excluded from share sums | Notional shares included | Puts and calls masquerade as ownership |
| Filer overlap | Related filer entities deduplicated | Every filer counted separately | The same shares counted twice |
| Quarter alignment | One quarter for all filers | Latest available per filer | A blended, mixed-date snapshot |
| Share count vintage | Point-in-time shares outstanding | Today's share count | Buybacks and issuance skew history |
A few of these deserve emphasis. 13F filings include listed put and call positions reported with the notional share count of the underlying, so pipelines that sum option rows into the numerator turn a bearish put into apparent "ownership." Amendments keep arriving for weeks after each deadline, and a vendor that keeps both an original and its restatement counts an entire portfolio twice. Large asset managers also file through multiple related entities, and whether those get rolled up changes the total. We have written about these failure modes in detail in how accurate is 13F data, because we hit most of them while building our own pipeline.
The practical rule: never compare institutional ownership percentages across sources, only within one source over time, and prefer a source that states exactly which choices it made.
What the percentage can and cannot tell you
Used carefully, the level of institutional ownership tells you about a stock's audience. Very high percentages in large caps mostly reflect index inclusion: the biggest 13F filers are passive managers who hold a stock because it sits in a benchmark, not because anyone underwrote the thesis. Very low institutional ownership in a liquid name is often more interesting, flagging companies that are too small, too new, or too restricted for institutional mandates. Academic work has generally found that ownership levels track size, liquidity, and index membership more than subsequent returns, so the raw level is a weak signal on its own.
The changes are where the information lives. Which funds initiated, added, trimmed, or exited a position quarter over quarter is a far richer dataset than one percentage, and it is exactly what the underlying filings contain once you know how to read a 13F filing. You can browse that fund-by-fund view across the whole filer universe at /investors.
Know the blind spots, too. The 45-day lag means a Q1 snapshot can describe positions long since sold. Long-only reporting means a fund that looks aggressively bullish on paper may be hedged flat in instruments you cannot see. And 13F is the slow channel: a holder crossing 5% with activist intent must file Schedule 13D within 5 business days, and corporate insiders report trades on Form 4 within 2 business days (Arkolith tracks 51,000+ of those insider transactions alongside the 13F spine). The full comparison lives in 13F vs 13D vs 13G vs Form 4.
Pulling the number apart programmatically
If you are building an agent or a quantitative process on top of institutional ownership, query the filings rather than the precomputed percentage, so the merge decisions are yours and visible. Arkolith serves the aggregated 13F dataset through a REST API and an MCP server behind a single key. The recipe for a defensible ownership figure: resolve the security, pull holdings across filers, aggregate share counts by CUSIP from each filer's effective post-amendment filing with option legs excluded, then divide by a shares-outstanding figure dated to the same quarter end.
Resolve a security or filer first:
curl -H "Authorization: Bearer YOUR_KEY" "https://arkolith.com/api/v1/search?q=NVIDIA"
List the covered institutional filers:
curl -H "Authorization: Bearer YOUR_KEY" "https://arkolith.com/api/v1/funds"
Pull one manager's reported holdings by CIK (Berkshire Hathaway here):
curl -H "Authorization: Bearer YOUR_KEY" "https://arkolith.com/api/v1/funds/1067983/holdings"
Every datapoint carries provenance back to the SEC EDGAR accession number of the filing it came from, so an agent can cite the exact source document instead of asserting a number from memory. When two agents disagree about an ownership figure, they can settle it by diffing source filings rather than by arguing. Setup takes a few minutes; start at the quickstart.
Frequently asked questions about institutional ownership percentage
Can institutional ownership exceed 100%?
Yes, and it usually signals a data artifact rather than anything sinister. Double-counted amendments, overlapping filer entities, option notional summed as shares, or a stale shares-outstanding figure after buybacks can all push the computed percentage past 100. Shares on loan to short sellers can also be reported as held in more than one account at once, which inflates aggregate holdings.
Does the percentage include index funds and ETFs?
Yes. The managers behind the major index funds and ETFs are among the largest 13F filers, so a large-cap stock's institutional ownership is dominated by passive holdings. That is the main reason a high percentage says more about index membership than about active conviction.
How often does institutional ownership update?
Quarterly at best. 13F filings are due 45 days after each quarter end, and amendments keep arriving afterward, so the number can describe positioning that is two to four and a half months old. Faster disclosures exist for different events: Schedule 13D within 5 business days and Form 4 within 2 business days.
Is a high institutional ownership percentage bullish?
Not by itself. Academic work has generally found the level correlates with company size, liquidity, and index inclusion more than with future returns. Quarter-over-quarter changes, especially clustered initiations or exits by active managers, are the more informative slice of the same data.
This article explains public filings and data concepts. It is not investment advice.
Keep reading
Measuring Portfolio Turnover From 13F Filings
How fast does a manager really trade? A practical method for estimating portfolio turnover from quarterly 13F snapshots, what high and low readings imply, and the caveats that bite.
Hedge Fund Clone Portfolios: Sizing, Timing, Caveats
How to build a hedge fund clone portfolio from 13F filings: manager selection, three weighting schemes, the quarterly rebalance calendar, and the caveats that compound.
Family Office 13F Rules: Why Some Investors Vanish
Some of the most famous investors file no 13F at all. What the family office exemption actually covers, what still gets disclosed, and where the record goes dark.