A headline calls Harbor Coffee, our fictional coffee company, a squeeze candidate. One screen shows short interest at 20% of float. Another says 16% of shares outstanding. A third says days to cover has halved.
Have short sellers started leaving?
You cannot tell from those numbers alone. The percentages can describe the same short position, and days to cover can fall while that count stays put. The missing piece is what each number is divided by.
Count positions, then name the base
Short interest counts shares in open short positions on a given date. These are short sales that have not yet been closed.
Short interest as a percentage of float compares that count with float, the shares available for public trading. That base can be smaller than the company's total shares outstanding.
Harbor reports 50 million shares outstanding at the end of year 3. Suppose 40 million are in the float and 8 million are short. Those two inputs, and the trading volumes below, are invented. With share counts in millions:
- Of float: 8 ÷ 40 × 100 = 20%.
- Of shares outstanding: 8 ÷ 50 × 100 = 16%.
Same short position, different base. Neither screen contradicts the other.
In the US, brokerage firms report short interest to FINRA twice a month. On FINRA's reporting calendar, the settlement date tells you when positions were measured; the later publication date tells you when the report was released. Freshly published does not mean freshly measured.
Days to cover is a volume ratio
Days to cover divides shares short by average daily share volume. Here the average covers the 20 trading days ending at each snapshot; other providers can use different windows.
At 2 million shares traded a day on average, 8 million shares short gives 8 ÷ 2 = 4 days. The open short position equals four days of average trading volume.
Snapshot A is year-3 end. B is the next reporting snapshot. Only average volume changes; shares short, float and shares outstanding stay fixed. Share counts and daily volumes are in millions (m).
| Snapshot | Short (m) | Daily avg. (m) | Days to cover |
|---|---|---|---|
| A | 8 | 2 | 4 |
| B | 8 | 4 | 2 |
The ratio halves: 8 ÷ 4 = 2 days. The reported short position has not budged.
Four days to cover gives nobody a four-day deadline. It does not promise an exit in four days, either. It compares a position with the pace of trading volume. Short sellers compete with other buyers; past volume is not a pile of shares reserved for them at a fixed price.
What a crowded position tells you
Short interest at 20% of float is more crowded than 2% of a comparably measured float. It does not tell you which side is right. No percentage turns crowding into an automatic buy or sell signal.
Harbor's history is useful only if the percentages use the same definition of float and days to cover uses the same averaging method. Otherwise a data change can look like traders changing their minds.
A short squeeze happens when purchases to close short positions help push the price higher, prompting further covering. Traders who expected a decline become buyers under pressure. Rising prices bring losses, and brokers may demand more collateral.
Other buyers can strengthen the loop; willing sellers can absorb the demand. A rally alone does not reveal how much buying came from shorts under pressure. The GameStop case study examines the evidence behind its 2021 rally.
Read the label before the signal
Daily short-sale volume counts shares sold short during a day in the markets a dataset covers. A trader who shorts 1 million shares and buys them back that afternoon adds 1 million to short-sale volume, but leaves no open position for the next snapshot.
Adding up daily short sales cannot tell you how many shares remain short. Nor does a steady position count prove nobody covered: new shorts can replace closed ones.
A high short-interest figure can reflect doubts about the business, a hedge against another holding, or speculation. The count does not reveal motives or tell you what the stock is worth. Borrow fees and available stock loans are separate data.
Read a display's date, denominator and volume window before interpreting the change. If the snapshot date or averaging window is missing, the comparison has to wait. Harbor's sheet gives us all three:
- Date: A measures year-3 end; B measures the next reporting snapshot.
- Percentage: 8 million short shares equal 20% of float or 16% of shares outstanding. Both screens are right.
- Volume window: The 20-trading-day average rises from 2 million shares to 4 million. Days to cover falls from four to two; total short interest stays at 8 million shares.
Nothing in the sheet shows a fall in total short interest. Whether Harbor is worth buying still depends on the business and its price. These figures settle the arithmetic, not the investment case.
Borrowing shares creates a debt. The next lesson turns to options, whose buyers pay for a right with an expiration date.
In short
- Short interest counts open short positions at a dated snapshot.
- A short-interest percentage needs a stated float or shares-outstanding denominator.
- Days to cover is a volume ratio, not a repayment clock.
- Daily short-sale volume includes shorts that were closed the same day.
- Crowding can fuel a squeeze but does not predict one.