Short Volume
Obtain the daily short sale volume of each company from FINRA’s Reg SHO files: the number of shares sold short, of which short exempt, the total number of shares reported to FINRA, the short volume ratio (short volume divided by total volume, as a decimal) and the z-score of that ratio against the window trading days before it. A ratio well above its usual level means an unusually large share of the day’s selling was short selling.
This is short sale VOLUME, the shares sold short during a day, not short interest, the short positions still open (which FINRA publishes twice a month through api.finra.org). Most short sales are closed again within days, and market makers sell short to fill buy orders, so a ratio around 0.4 to 0.5 is normal for liquid stocks. The total volume is only the volume reported to FINRA’s facilities (the Trade Reporting Facilities, where off-exchange trades are reported, and the Alternative Display Facility), not the consolidated volume of all exchanges, so it is lower than the volume of the historical data. The short volume includes the short exempt volume, and volumes include fractional shares.
FINRA publishes one file per trading day, from August 2018, covering securities listed in the United States only: a ticker with an exchange suffix (such as ASML.AS) has no data. Each day is a separate download of about half a megabyte, so only the history_days before the Toolkit’s end date are retrieved, by default about three months (some 65 files, a few seconds). The days are cached per ticker, so a rerun only downloads the days that are new.
The z-score is (ratio − mean) / standard deviation (ddof=1) of the window trading days before each day, so a day is not part of its own comparison, and it is empty for the first window days. Research finds that heavy short selling predicts lower returns in the following weeks, see Boehmer, Jones and Zhang (2008), “Which Shorts Are Informed?”, Journal of Finance 63(2), and Diether, Lee and Werner (2009), “Short-Sale Strategies and Return Predictability”, Review of Financial Studies 22(2).
Also known as: short sale volume, short volume ratio, daily short volume, Reg SHO short volume.
Short Volume in Python or via the MCP
Install the Finance Toolkit with:
pip install financetoolkit -U
Then call get_short_volume as shown below.
from financetoolkit import Toolkit
toolkit = Toolkit(
["GME"],
api_key="FINANCIAL_MODELING_PREP_KEY",
start_date="2026-09-01",
end_date="2026-10-30",
)
toolkit.get_short_volume().drop(columns="Short Exempt Volume").tail()
Or, after setting up the Finance Toolkit MCP, ask:
What were GameStop's short volume, total volume and short volume ratio on the last five trading days of October 2026?
Which returns:
| Date | Short Volume | Total Volume | Short Volume Ratio | Short Volume Ratio Z-Score |
|---|---|---|---|---|
| 2026-10-26 | 1.64656e+06 | 3.75988e+06 | 0.4379 | -2.0905 |
| 2026-10-27 | 1.34555e+06 | 2.42539e+06 | 0.5548 | -0.5315 |
| 2026-10-28 | 1.51079e+06 | 2.53413e+06 | 0.5962 | 0.0279 |
| 2026-10-29 | 1.34307e+06 | 2.34151e+06 | 0.5736 | -0.2147 |
| 2026-10-30 | 2.39821e+06 | 4.14535e+06 | 0.5785 | -0.1479 |
Parameters
get_short_volume accepts the following parameters.
| Parameter | Type | Description | Example |
|---|---|---|---|
history_days | int | None | The number of calendar days before the Toolkit’s end date to retrieve, never before its start date. Defaults to 92, about three months. None retrieves the Toolkit’s whole date range (one download per trading day). | |
window | int | The number of earlier trading days the z-score compares with. Defaults to 20, about a month. | 20 |
rounding | int | None | The number of decimals to round the results to. Defaults to None, which uses the rounding of the Toolkit. | 2 |
show_columns | list[str] | None | A list of column names to keep in the result. Invalid names are reported and ignored. Defaults to None, which keeps every column. | None, which keeps every column |