Choose the measure that answers your question
Short volume is a flow; short interest is a stock of positions. The word stock here means an amount measured at one point in time. Neither measure identifies every trader's motive. The SEC describes short sales used for speculation, hedging and market making, so a short-sale label alone does not establish a directional bet.
- FINRA Rule 4560 reports gross short positions by account, rather than netting every investor against a long holder. Reportable short positions must have settled or reached settlement date by the snapshot close.
- A percentage of shares outstanding is a different statistic from a percentage of float. Keep the denominator in the label.
- Do not add daily short volume to yesterday's short interest. The daily file does not identify all purchases that close shorts or provide an account-level position reconciliation.
| Measure | Question answered | Time and denominator |
|---|
| Short sale volume | How many shares traded with a short-sale designation? | Shares during a trade date; divide by matching total volume for a percentage. |
| Short interest | How many reported short shares remained outstanding? | Shares at the reporting settlement-date snapshot; no denominator for the raw count. |
| Short float | How large is that outstanding balance relative to float? | Short interest / float; save the float definition and effective date. |
| Days to cover | How large is that balance relative to usual trading activity? | Short interest / average daily share volume; save the averaging window. |
Read the FINRA file before calculating a percentage
FINRA daily files cover publicly disseminated trades reported to its facilities during normal market hours. For listed shares, these are off-exchange reports. The Consolidated TRF/ADF file combines those facilities; it is not a consolidated total of all exchange and off-exchange trading. The ORF file covers a separate OTC reporting population.
- Use either the consolidated TRF/ADF file or its component facility files for the same listed security and date. Adding both counts overlapping reports twice.
- FINRA's file layout defines ShortVolume as including short-exempt trades. ShortExemptVolume is a subset: do not add it again. To study non-exempt short sales, subtract it and label that different numerator.
- Use TotalVolume from the same coverage and session as the numerator. A chart-volume denominator may have a different feed, venue set or session.
- To extend coverage, add non-overlapping exchange data to both numerator and denominator after checking definitions. Missing venues mean a partial-market result, even if the filename says consolidated.
- FINRA explains that certain offsetting customer transactions are not publicly disseminated. Their absence can make reported short volume look high without revealing a lasting short position.
Worked example: 70% short volume, no additional short position
This invented one-day ledger demonstrates the mechanism, not a real stock. A dealer begins flat, sells 7,000 shares short in public off-exchange trades, then buys 7,000 shares from customers in offsetting transactions that are not publicly disseminated under the assumed reporting treatment. Other public long sales total 3,000 shares. Assume no other dealer activity and no change in other participants' short positions.
- Excluding the 200 short-exempt shares gives 6,800 / 10,000 × 100 = 68% non-exempt short volume. Adding them to ShortVolume would incorrectly produce 72%.
- The position conclusion comes from the hypothetical ledger, not from the 70% statistic. A public volume file alone cannot reconstruct that ledger.
- Even where both legs are publicly reported, selling short and covering on the same day can leave no new short balance. The volume counts executions while the snapshot counts positions still open.
| Ledger item | Shares | Effect in this example |
|---|
| Public short sales | 7,000 | Included in ShortVolume and TotalVolume. |
| Of those sales, short exempt | 200 | Already part of the 7,000, not additional volume. |
| Other public long sales | 3,000 | Included in TotalVolume only. |
| Offsetting dealer purchases | 7,000 | Close the dealer position; excluded from the assumed public file. |
| Reported short-volume percentage | 7,000 / 10,000 × 100 | 70%. |
| Dealer ending short position | 0 + 7,000 - 7,000 | 0 shares; no additional outstanding short position. |
Combine venue volumes, then calculate the ratio
Continue the hypothetical example with disjoint exchange data for the same security, trade date and normal-hours session: 4,000 short shares out of 20,000 total shares. Assume these two datasets cover the entire chosen study universe. This is a controlled arithmetic example, not a claim that two real files cover the whole US market.
- Correct: (7,000 + 4,000) / (10,000 + 20,000) × 100 = 36.67%. Averaging 70% and 20% equally gives 45%, which ignores different volumes.
- Using only the off-exchange numerator with combined volume gives 7,000 / 30,000 = 23.33%. That is not the combined short-volume percentage.
- Zero total volume makes the percentage undefined. A missing file is unknown, not zero trading. Retain corrections as new versions rather than combining an original file with its replacement.
| Coverage | Short shares | Total shares | Short-volume percentage |
|---|
| Off-exchange sample | 7,000 | 10,000 | 70% |
| Exchange sample | 4,000 | 20,000 | 20% |
| Combined study universe | 11,000 | 30,000 | 36.67% |
Calculate short float and days to cover separately
For an independent hypothetical settlement-date snapshot, assume short interest is 1,200,000 shares, float is 8,000,000 shares and shares outstanding are 10,000,000. Define a research average using 20 completed trading sessions totaling 6,000,000 shares in the chosen market coverage. All inputs use compatible share units and split adjustments.
- The 20-session window is an explicit research choice. Nasdaq's report defines its own average-volume window and display conventions; a custom result need not match a vendor field.
- Four days does not predict that shorts will close within four days. Covering demand, available liquidity and trading volume can change, and ordinary volume is not reserved for short covering.
- If average daily volume or float is zero or missing, mark the corresponding ratio unavailable. Do not substitute a tiny denominator or silently use shares outstanding as float.
- With the same balance and average volume of 600,000 shares, days to cover falls to 2 while short float stays 15%. A falling days-to-cover ratio alone does not prove shorts covered.
| Calculation | Arithmetic | Result |
|---|
| Short interest as a percentage of float | 1,200,000 / 8,000,000 × 100 | 15% |
| Short interest as a percentage of shares outstanding | 1,200,000 / 10,000,000 × 100 | 12% |
| 20-session average daily volume | 6,000,000 / 20 | 300,000 shares per session |
| Research days to cover | 1,200,000 / 300,000 | 4 days |
Keep settlement, publication and availability dates apart
FINRA requires short-interest reporting twice monthly. Its calendar distinguishes the settlement snapshot, the broker reporting deadline and public publication. For example, the September 30, 2026 snapshot has an October 2 reporting deadline and October 9 publication date. The reporting deadline is not public availability.
- Nasdaq states that its short-interest release occurs after 4 p.m. Eastern Time on the dissemination date. An October 9 regular-session decision cannot assume that release was already known.
- FINRA states that daily short-volume files are posted by 6 p.m. ET on their trade date, with occasional later corrections. A same-date label does not make the completed daily total available at the morning open.
- Save the actual release or retrieval timestamp when available. For a reproducible decision rule, use the later of public availability and your recorded receipt, then the next eligible decision time.
- For a regular-session study using the September 30 snapshot, start eligibility at the next regular session after the October 9 release, subject to actual receipt. Keep September 30 as the observation label; do not backfill it as information known that day.
Apply the record to a TradingView chart
Open the exchange-qualified price symbol in TradingView and record its session, interval and timezone. Use an external source record for the short data unless you have verified a matching accessible series. This workflow does not assume a universal TradingView short-interest feed or attempt to infer positions from price candles.
- Set the displayed chart timezone to New York for this US example and retain the original timestamp with its timezone. TradingView documents that chart timezone changes the display; it does not change Pine script time calculations.
- Place a clearly labeled observation-date annotation and a separate availability annotation. For historical decisions, inspect only bars at or after the eligible decision timestamp.
- If using a vendor indicator, check its source, venue coverage, date semantics and revisions. A plotted historical value may represent an observation date rather than when users first received it.
- Save the screenshot with its symbol, interval and timestamp visible. Keep the source file separately so chart annotations can be checked against the underlying numbers.
| Record field | What to preserve |
|---|
| Source identity | Provider, URL or filename, version, security identifier and retrieval timestamp. |
| Coverage | Facilities or venues, session, short-exempt treatment and whether any venues are missing. |
| Dates | Trade date or settlement date; publication timestamp; received timestamp; first eligible decision. |
| Inputs | ShortVolume, TotalVolume, short interest, float source/date, average-volume window and units. |
| Review | Hypothesis written before the outcome, chart screenshot URL, later observation and alternative explanations. |
Save the evidence before interpreting the price move
Pineify Trading Journal session notes can hold the record fields, assumptions and screenshot links. Add a dated follow-up after the observation window rather than rewriting the original hypothesis. AI Chart Analysis accepts a chart screenshot for review of visible chart context; it cannot recover unshown FINRA reports, identify who covered, or verify a float denominator from candles.
- Keep the input files and arithmetic with the record. A chart interpretation should remain distinguishable from a verified short-interest observation.
- A high short float or days-to-cover ratio can motivate further research, but it does not establish a squeeze, a covering date or a future price direction. Borrow availability and fees are separate data inputs.
