Analysis

Days to Cover: How to Calculate Short Squeeze Risk

Days to cover divides shares sold short by average daily volume to show how crowded a short trade is. Here is the formula, FINRA's twice-monthly data schedule and worked 2026 examples using SoundHound, Jack in the Box and Bloom Energy.

Days to Cover: How to Calculate Short Squeeze Risk

Days to cover is short interest divided by average daily trading volume: it estimates how many sessions it would take short sellers to buy back every borrowed share if they captured all of a normal day's volume. SoundHound AI ($SOUN) scored 12.44 days in FINRA's October 30, 2026 data, Jack in the Box ($JACK) 12.74 and Bloom Energy ($BE) just 1.06. The number measures how crowded the exit is. It does not say when anyone will head for it.

The days to cover formula

The math fits on one line:

Days to cover = shares sold short ÷ average daily volume

A stock with 10 million shares short that trades 2 million shares a day has a days-to-cover reading of 5 (hypothetical numbers, used only to show the formula). Traders also call it the short interest ratio. A reading near 1 means a single average session could, in theory, absorb every short position. A reading above 10 means shorts own a position that would take two trading weeks of total volume to unwind, and they would never get all of it, because buyers, sellers and market makers are trading too.

That is why squeeze watchers care. When a crowded short meets good news, the shorts who want out are competing for a limited daily supply of shares, and the lower the volume relative to the position, the more that competition can push the price.

Where short interest data comes from

US short interest is collected by FINRA from its member broker-dealers. Its short interest reporting page states: "FINRA requires firms to report short interest positions in all customer and proprietary accounts in all equity securities twice a month." FINRA Rule 4560 adds that reports "shall be received by FINRA no later than the second business day after the reporting settlement date designated by FINRA."

Three dates matter for every cycle:

  • Settlement date: the snapshot. Only short positions from sales that have settled or reached settlement by that day count.
  • Due date: brokers file by 6 p.m. Eastern, two business days later.
  • Publication date: FINRA releases the aggregate per-security figures.

On FINRA's 2026 schedule, the November 13 settlement is due November 17 and publishes November 24. So on November 19, the freshest public snapshot is still October 30, roughly three weeks old. FINRA's equity short interest data publishes the short position, an average daily volume figure and its own days-to-cover number for each security. Its field notes say the volume figure "Excludes non-media trades." Nasdaq's per-ticker short interest pages show the same figures.

Days to cover vs. short interest % of float

The two numbers answer different questions. Short interest as a percentage of float (short shares divided by shares available to trade) measures how much of the tradable stock is borrowed and sold. Days to cover measures how hard it would be to buy those shares back, given how actively the stock trades.

A stock can score high on one and low on the other. A heavily traded name can carry a big short position and still show a low days-to-cover reading, while a sleepy small cap with a modest short position can post a high one. Squeeze screens usually want both: a large share of the float sold short and a thin market to buy it back in.

Float is also not a standard number. Data vendors subtract insider, strategic and restricted holdings in different ways. The cleanest public denominator is shares outstanding from the cover page of a company's latest 10-Q, which is what the table below uses. Percent-of-float figures will run higher.

Worked example: SoundHound, Jack in the Box and Bloom Energy

TTFY has covered squeeze talk in all three: SoundHound in August, Jack in the Box in June and Bloom Energy in July. Here is the October 30, 2026 settlement data from FINRA, with each ratio recomputed.

StockShort interestFINRA avg daily volumeDays to cover
$SOUN163,774,72813,163,206163,774,728 ÷ 13,163,206 = 12.44
$JACK6,033,733473,6656,033,733 ÷ 473,665 = 12.74
$BE17,686,32816,752,80717,686,328 ÷ 16,752,807 = 1.06

Next, the volume side, rebuilt from Yahoo Finance daily data. The 11 sessions from October 16 through October 30 (the stretch since the prior settlement date) land within 1% of FINRA's figure. The last 10 completed sessions, November 5 through November 18, show how far the ratio has drifted since.

StockOct 16–30 avg volumeDays to coverNov 5–18 avg volumeDays to cover
$SOUN146,023,100 ÷ 11 = 13,274,82712.34226,821,000 ÷ 10 = 22,682,1007.22
$JACK5,258,400 ÷ 11 = 478,03612.623,720,000 ÷ 10 = 372,00016.22
$BE184,280,700 ÷ 11 = 16,752,7911.06179,620,700 ÷ 10 = 17,962,0700.98

Same short position, different volume window, very different story. SoundHound's reading falls from about 12 days to about 7 because November volume jumped. Jack in the Box's rises to about 16 because volume thinned. Both recent figures still divide by the October 30 short count, the only one published.

Finally, the share of stock sold short, using 10-Q cover-page share counts:

StockShares outstanding (10-Q cover)Short % of outstanding
$SOUN455,858,001 Class A (Nov 5, 2026)163,774,728 ÷ 455,858,001 = 35.93%
$JACK19,179,330 (Aug 5, 2026)6,033,733 ÷ 19,179,330 = 31.46%
$BE295,656,689 (Oct 19, 2026)17,686,328 ÷ 295,656,689 = 5.98%

SoundHound also has 32,535,408 unlisted Class B shares. Folding them in drops the ratio to 163,774,728 ÷ 488,393,409 = 33.53%, a reminder that the denominator choice moves the answer.

What the numbers did and did not show

SoundHound's history is the cleanest lesson. On July 31, 2026, FINRA showed 162,745,724 shares short, average volume of 38,950,100 and days to cover of 4.18. By October 30, short interest had barely moved: up 1,029,004 shares, or 0.63% (1,029,004 ÷ 162,745,724). Average volume had fallen 66.2% ((38,950,100 − 13,163,206) ÷ 38,950,100). The ratio nearly tripled because trading dried up, not because shorts piled in. Meanwhile, $SOUN closed at $4.22 on October 30 and $3.01 at the November 18, 2026 close, down 28.7% ((3.01 − 4.22) ÷ 4.22). A high reading did not stop the decline.

Bloom Energy shows the reverse. In FINRA's June 30, 2026 cycle, around the squeeze TTFY covered, Bloom's short interest fell 35.03%, from 29,780,167 to 19,348,027 shares, with days to cover at 1.11. When the ratio is near one, shorts can exit in about a day of normal trading, so there is little scarcity to squeeze. $BE went from $203.17 on October 30 to $170.98 at the November 18 close, down 15.8% ((170.98 − 203.17) ÷ 203.17).

Jack in the Box shows the thin-volume trap. Its days to cover was 5.00 on June 30, when FINRA's average volume was 1,346,893 shares. October's average was about 35% of that (473,665 ÷ 1,346,893 = 0.35), so the reading climbed to 12.74 even though short interest fell from 6,740,848 to 6,033,733 shares. $JACK rose from $13.02 to $14.79 at the November 18 close, up 13.6% (1.77 ÷ 13.02), but whether shorts covered into that move will not be visible until the November 13 data publishes on November 24.

Common mistakes with days to cover

  • Treating stale data as live. Every figure is a settlement-date snapshot published roughly a week and a half later. Shorts may already have covered or doubled down.
  • Ignoring the volume window. A single headline, even an unconfirmed one like the report TTFY covered on Nvidia, can bring a burst of volume that drags the average up and the ratio down. Know whether a site uses the settlement period, 20 days or 30 days.
  • Mixing float definitions. Compare percent-of-float figures only from the same provider, or compute your own from the 10-Q.
  • Calling a high reading a trigger. Days to cover measures fuel, not ignition. A squeeze needs a catalyst that forces buying: news, a borrow recall or rising borrow costs. SoundHound's fall shows a crowded short can keep working.
  • Forgetting deal and hedge positions. Short interest includes arbitrage and hedging shorts that have no reason to panic-cover.

Days to cover checklist

  1. Pull the latest settlement date from FINRA and note how old it is.
  2. Recompute short interest ÷ average daily volume yourself.
  3. Rerun it with the last 10 sessions of volume to see the drift.
  4. Divide short interest by 10-Q shares outstanding, listed class only.
  5. Compare against two or three prior cycles: is the short count or the volume driving the change?
  6. Look for a dated catalyst before calling anything a squeeze setup.
  7. Mark the next publication date and recheck then.

This is general information, not investment advice. For more short squeeze and deal coverage, see the Trading Tips For You homepage. Short interest is an aggregate number with no names attached; to see who holds large long stakes, learn to read a Schedule 13D.

Bull/Bear Verdict

Bull Case: Days to cover tends to favor holders when a high reading, a large short share of the float and a real catalyst line up, because shorts then compete for scarce daily volume. Computed fresh, it is a free, standardized way to rank how crowded a short is.

Bear Case: Each snapshot is more than three weeks old by the time the next one publishes, it swings with whatever volume window is used, and SoundHound's 28.7% slide with a 12-plus reading shows that crowded shorts can be right. It flags fuel, never timing.

FAQ

What is a high days to cover number?

There is no official cutoff. Readings near 1 mean shorts can exit in about a day of normal trading; double-digit readings, like SoundHound's 12.44 and Jack in the Box's 12.74 on October 30, 2026, mean the exit would take weeks of total volume.

How often is short interest updated?

Twice a month. FINRA sets settlement dates around mid-month and month-end, brokers report two business days later, and FINRA publishes about a week after that. The November 13, 2026 snapshot publishes November 24.

Is days to cover the same as the short interest ratio?

Yes. Both terms mean short interest divided by average daily volume. Short interest as a percentage of float is a different measure.

Why does my broker show a different days to cover figure?

Providers use different volume windows and update on different days. Recomputing $SOUN with 10 recent sessions instead of the settlement period moved its reading from 12.34 to 7.22.

Share X LinkedIn Email
Disclaimer: The information provided is for informational purposes only and is not intended as financial, legal, or tax advice. Trading around earnings involves significant risk and increased volatility. Past performance is not indicative of future results. No strategy can guarantee profits or protect against loss. Consult a professional advisor before acting on any information provided.

More in Analysis · Full archive