Skip to content
DilutionScan

Small-cap trading mechanics

Short Interest vs Short Float: Which Number Matters for a Squeeze

Short interest is a FINRA share count; short float is that count divided by a float. Why the two differ, which dates they carry, and how to read each.

6 min readLast verified

The short version

Short interestShort float (short % of float)
What it isShares sold short and not yet coveredShort interest divided by a float figure
UnitSharesPercent
Who produces itBroker-dealers report it, FINRA compiles and publishes itCalculated by each data provider
How oftenTwice a month, as of a settlement dateWhenever the provider recalculates
Official definitionYes, FINRA Rule 4560No; the denominator varies by provider
Main weaknessAbout one to two weeks old when publishedTwo numbers from two different dates
AnswersHow many shares are short?How large is that position relative to supply?

Short interest is a count of shares. Short float is a ratio, and its denominator is the part people disagree about. When a scanner says a stock is "25% short", it is showing short interest divided by whatever float that scanner uses.

For squeeze research the ratio is the more comparable figure across stocks, because 5 million shares short means something very different for a company with 20 million tradeable shares than for one with 500 million. But the ratio is only as good as both of its inputs, and they are rarely measured on the same day.

What short interest measures

Under FINRA Rule 4560, broker-dealers report the short positions in customer and proprietary accounts for every exchange-listed and OTC equity security, twice a month:

  1. Settlement date. FINRA designates a mid-month and an end-of-month settlement date. Positions are measured as of that date.
  2. Due date. Firms report by 6:00 p.m. ET on the second business day after the settlement date.
  3. Publication. FINRA publishes the compiled figures on a date set in its published calendar, typically about a week after the due date. For the January 15, 2026 settlement date, reports were due January 20 and FINRA's calendar set publication for January 27.

So a published short interest number describes positions that are at least a week and a half old, and the next update is about two weeks away. It counts open positions, not trading activity. FINRA's daily short sale volume files are a separate data set and are not short interest. For the full collection process, see short interest, days to cover and borrow.

What short float measures

Short float = short interest ÷ float × 100.

FINRA publishes the numerator. Nobody publishes an official denominator for this purpose, so each provider picks one:

  • Shares outstanding. Every issued share, including insiders' shares. This gives the smallest percentage. Strictly it is "short % of shares outstanding", but some sites still label it float.
  • Non-affiliate float. Shares outstanding minus shares held by officers, directors and controlling holders. This is close to the SEC's public float concept.
  • Free or tradeable float. Non-affiliate float minus restricted, locked-up or strategic holdings. This gives the largest percentage, and every vendor computes it differently.

Float vs shares outstanding walks through all three with one hypothetical company.

A worked example with real numbers

GameStop (GME), checked on Signal8 on November 2, 2026:

InputValueAs of
Short interest (FINRA, via Signal8)48,031,618 sharesSettlement date October 15, 2026
Class A shares outstanding (10-Q cover page)504,500,990September 3, 2026
Class A shares outstanding (earlier 10-Q cover page)448,691,257June 5, 2026
Vendor free float409,099,854October 2, 2026

One short position produced three figures between 9.52% and 11.74%, depending only on which denominator sat underneath it. The free float is about 95.4 million shares below the September share count. The vendor does not state which holdings it removed, so the gap cannot be split between insider holdings, other excluded blocks and timing.

The timing point is visible in the filings alone. The share count rose by 55,809,733 between the June and September cover pages. A provider still dividing by the June figure would show a higher percentage than one using September's, for the same short position.

Why the two numbers move differently

Because the ratio has two inputs, it can change when no short seller has done anything:

  • New shares. An offering, a convertible conversion or a warrant exercise raises shares outstanding and usually the float. Short float falls, while short interest is unchanged.
  • Reverse splits. Both inputs fall by the split ratio. If the short interest figure predates the split and the float is post-split, the ratio is wrong by the full ratio. Check the dates.
  • Restricted shares becoming tradeable. A resale registration going effective or a lockup ending can raise a tradeable float without changing shares outstanding.
  • Covering or new shorting. Only this changes short interest itself, and you see it at the next settlement date.

Days to cover is a third measure built on the same numerator: short interest divided by average daily volume. A volume spike, measured with relative volume, lowers it without any covering.

Which number to use for which question

QuestionBetter inputWhy
Did shorts add or cover since last report?Short interest, compared across settlement datesSame unit, same source, no denominator
How crowded is the short relative to supply?Short float, with the float definition statedComparable across companies of different sizes
How long might covering take at recent volume?Days to coverUses volume, not float
Is the figure current?Neither alone; read both datesShort interest and float are dated separately

Short float above 100% is possible: a borrowed share can be sold, bought and lent again, so each short position counts separately. A stale or narrow float definition can also push the ratio past 100%.

How to check short data on Signal8

  • The Shorts tab on a company page, for example GME's Shorts tab, brings short interest and related data into one view. Check the settlement date beside any figure.
  • The Squeeze scanner lists stocks with at least 15% short interest, so the denominator question above applies to every row.
  • The Low Float scanner and the guide to finding low float stocks cover the float side.
  • In the Terminal, open a symbol to see its quote and volume alongside the short figures.

FINRA publishes its reporting calendar on its short interest reporting page.

FAQ

Is short float the same as short interest?

No. Short interest is a number of shares: the open short positions brokers reported to FINRA as of a settlement date. Short float is a percentage: that share count divided by a float figure chosen by the data provider. Different providers can show the same short interest and different short float percentages, because they divide by different floats or by floats from different dates.

Does short float include shares sold short today?

No. Short interest, and therefore short float, is measured as of FINRA's mid-month or end-of-month settlement date and published about a week and a half later. Shorting that happened after the settlement date shows up only in the next report. FINRA's daily short sale volume files count trades marked short on a given day, which is a different measure and is not a running short interest total.

What denominator should short float use?

There is no official one. Dividing by shares outstanding gives the lowest figure and is the easiest to verify, because the count is on every 10-Q and 10-K cover page. Dividing by a tradeable float gives a higher figure that is closer to the supply that can trade, but every vendor builds it differently. Whichever is used, state it, and check that its date is close to the short interest settlement date.

What is considered a high short float?

There is no regulatory threshold. Screens set their own cut-offs; Signal8's Squeeze scanner, for example, starts at 15% of float. A percentage is only comparable between two stocks when both use the same float definition and similar dates. A high figure on a stale or narrow float can overstate how much of the tradeable supply is short.

See it on DilutionScan

Companies DilutionScan covers, with the instruments read from their SEC filings. Open one to see each instrument and the filing it came from.

Terms in this guide

Days to cover
Days to cover is short interest divided by average daily trading volume, a rough estimate of how many days of normal trading it would take short sellers to buy back their shares.
Public float
Public float is the shares held by investors other than a company's officers, directors and controlling holders, or their market value; the SEC uses the dollar figure for form eligibility.
Non-affiliate float
Non-affiliate float is the shares, or their market value, held by anyone who is not an affiliate of the company; it is the SEC's float measure and includes restricted shares held by non-affiliates.
Relative volume (RVOL)
Relative volume (RVOL) is a stock's trading volume divided by its typical volume for the same period, so a reading of 3.0 means three times the usual activity.

Sources

  • SEC accession 0001326380-26-000055 GameStop 10-Q, filed 2026-09-09, cover page: 504,500,990 Class A shares outstanding as of September 3, 2026
  • SEC accession 0001326380-26-000025 GameStop 10-Q, filed 2026-06-11, cover page: 448,691,257 Class A shares outstanding as of June 5, 2026

Check any ticker for dilution

DilutionScan reads each covered company's SEC filings and lists its warrants, convertibles, ATM programs, shelves and S-1s on one page, with a dilution rating and cash runway.

This guide is from the Signal8 Academy, the research team behind DilutionScan. Educational content only. DilutionScan and Signal8 are not broker-dealers or investment advisers, and nothing here is a recommendation to buy or sell any security.

DilutionScan ProEvery instrument in full · $59/mo