What Is Days to Cover? Short Interest Ratio

Days to cover is short interest divided by average daily trading volume: how many typical days of a stock's entire volume it would take for every short seller to buy back their shares. It also goes by a second name, the short interest ratio, and the two mean the same thing. It is the standard gauge of how crowded a short position is relative to the exit. This page shows how to read it, who acts on it, and what it does not tell you, using the latest exchange-reported settlement.

How is days to cover calculated?

Take the reported short interest, every share currently sold short, and divide by the stock's average daily volume. A company with 20 million shares short that trades 10 million shares a day carries 2.0 days to cover. The number is a time estimate: if short sellers unwound and absorbed all of a typical day's trading, that is how long the exit would take. Real unwinding never works that cleanly, but the ratio's power is comparative, a name at 10 days is far more crowded than one at 1.

Four familiar names from the June 30, 2026 settlement:

SELECT ticker,
       round(short_interest / 1e6, 1) AS shares_short_m,
       round(avg_daily_volume / 1e6, 1) AS avg_daily_volume_m,
       round(days_to_cover, 2) AS days_to_cover
FROM global_markets.stocks_short_interest
WHERE settlement_date = '2026-06-30'
  AND ticker IN ('AAPL', 'GME', 'MU', 'TSLA')
ORDER BY ticker

Look at the shape, not the levels. AAPL carries 140.5 million shares short, more than double GME's 55.9 million, yet its days to cover is only 1.73: at 81.1 million shares of daily volume, that short position is a rounding error of the tape. GME is the reverse, a smaller position over 5.2 million shares a day puts it at 10.75 days, and the name of the 2021 short squeeze is again the crowded one. Raw short interest counts shares; days to cover counts exit time.

What counts as high days to cover?

Calibrate against the market. Across every liquid name (5M+ shares of average daily volume) at the June 30 settlement:

SELECT multiIf(days_to_cover <= 1, '1 (the reported floor)',
               days_to_cover < 2, '>1 to 2 days',
               days_to_cover < 5, '2 to 5 days',
               days_to_cover < 10, '5 to 10 days',
               '10+ days') AS bucket,
       count() AS names
FROM global_markets.stocks_short_interest
WHERE settlement_date = '2026-06-30'
  AND avg_daily_volume >= 5000000
  AND days_to_cover IS NOT NULL
GROUP BY bucket
ORDER BY min(days_to_cover)

The mass sits low: 323 names print at the reported floor and 188 more sit between one and two days, most liquid stocks would clear their short book inside two sessions. The middle bucket, two to five days, holds 340 names. Above that the count falls off a cliff: 91 in the five-to-ten band, just 5 at ten days or more. Five days is uncommon; double digits is rare air.

One data note: the feed floors the ratio at 1.0, so a mega-cap whose true arithmetic works out to half a day still prints 1.0. Read "1" as "one day or less."

Is high days to cover normal for small caps?

Volume sits in the denominator, so a thin stock can look crowded on a position a liquid stock would swallow whole. Splitting the settlement file into four volume tiers separates liquidity from crowding:

SELECT multiIf(avg_daily_volume < 1000000, 'under 1M shares/day',
               avg_daily_volume < 5000000, '1M to 5M',
               avg_daily_volume < 20000000, '5M to 20M',
               '20M+ shares/day') AS adv_tier,
       count() AS names,
       if(count() < 1000,
          toString(count()),
          concat(toString(intDiv(count(), 1000)), ',', leftPad(toString(count() % 1000), 3, '0'))) AS names_fmt,
       round(min(days_to_cover), 2) AS lowest_reported,
       round(quantileDeterministic(0.5)(days_to_cover, cityHash64(ticker)), 2) AS median_days_to_cover,
       round(countIf(days_to_cover >= 5) / count() * 100, 1) AS pct_above_5_days,
       round(countIf(days_to_cover = 1) / count() * 100, 1) AS pct_at_floor
FROM global_markets.stocks_short_interest
WHERE settlement_date = '2026-06-30'
  AND days_to_cover IS NOT NULL
  AND avg_daily_volume > 0
GROUP BY adv_tier
ORDER BY min(avg_daily_volume)

The share of names above five days falls with every step up the volume ladder: 23.9% of the 16,020 names trading under a million shares a day, 19.1% in the 1M–5M tier, 12.1% in the 5M–20M tier, and only 3.6% of the 220 heaviest-traded names. Six days is ordinary in a micro-cap and genuinely unusual in a stock trading 20 million shares a day.

The thin tier is split at both ends, though: 46.5% of sub-1M names print exactly at the floor, a tiny short book over a tiny volume, dragging their median to 1.21 days, below the 2.77 of the 1M–5M tier. The floor shows in every tier: the lowest value reported in each is 1.

What's crowded right now?

The ratio's practical use is a screen. The most crowded liquid names, 5M+ average daily volume, so no thin-tape artifacts, at the newest settlement on file:

SELECT ticker,
       round(days_to_cover, 2) AS days_to_cover,
       round(short_interest / 1e6, 1) AS shares_short_m,
       round(avg_daily_volume / 1e6, 1) AS avg_daily_volume_m
FROM global_markets.stocks_short_interest
WHERE settlement_date = '2026-06-30'
  AND avg_daily_volume >= 5000000
  AND days_to_cover IS NOT NULL
  AND ticker NOT IN ('SPCX')
ORDER BY days_to_cover DESC, ticker ASC
LIMIT 12

MPT tops the list at 16.93 days, 138.1 million shares short against 8.2 million shares of daily volume. Every one of the 12 names clears 8.39 days, and GME is among them. Note what is absent: no mega-cap. This crowding lives in mid-caps and heavily-traded small caps, the population the most shorted stocks screens surface. Mind the freshness caveats below before reading any such list as today's positioning.

Both halves of the fraction move

The subtlety most explainers skip: the denominator is the twitchy half. A burst of volume, an earnings week, an index add, a viral day, can halve a stock's days to cover while not one short seller closes a position. Volume drying up inflates the ratio with no new shorting. GME's run into the 2021 squeeze is the textbook:

SELECT settlement_date,
       round(days_to_cover, 2) AS days_to_cover,
       round(max(days_to_cover) OVER (ORDER BY settlement_date), 2) AS peak_days_to_cover,
       round(short_interest / 1e6, 1) AS shares_short_m,
       round(avg_daily_volume / 1e6, 1) AS avg_daily_volume_m
FROM global_markets.stocks_short_interest
WHERE ticker = 'GME'
  AND settlement_date >= '2020-06-30'
  AND settlement_date <= '2021-01-29'
ORDER BY settlement_date

At the window's first settlement, 54.6 million shares were short against just 3.2 million shares of daily volume, 17.03 days to cover. The high-water mark over the stretch reached 25.82 days, and the chart shows the ratio swinging between single and double digits through the autumn while the short position barely moved. Then the squeeze: at the January 29, 2021 settlement, average daily volume had exploded to 96.8 million shares and days to cover printed 1, the floor, with 21.4 million shares still short. Some shorts did close; the collapse in the ratio is mostly the denominator.

Days to cover vs. short interest as a percent of float

The other crowding measure is short interest as a percent of float, shares short divided by the tradable share count, not by daily volume. Percent of float measures how much of the ownership pie is sold short; days to cover measures how narrow the exit door is. They rank the same stocks differently (the denominator below is total shares outstanding from the fundamentals snapshot; a true float excludes insider and restricted holdings, so vendor percent-of-float figures run higher):

WITH share_counts AS (
    SELECT ticker,
           argMax(market_cap / price, date) AS shares_outstanding,
           argMax(date, date) AS as_of
    FROM global_markets.stocks_ratios
    WHERE ticker IN ('AAPL', 'MSFT', 'NVDA', 'TSLA', 'MU', 'GME')
      AND market_cap > 0
      AND price > 0
    GROUP BY ticker
)
SELECT si.ticker AS ticker,
       round(si.short_interest / 1e6, 1) AS shares_short_m,
       round(si.short_interest / sc.shares_outstanding * 100, 2) AS pct_of_shares_out,
       round(si.days_to_cover, 2) AS days_to_cover,
       formatDateTimeInJodaSyntax(sc.as_of, 'MMMM d, yyyy') AS share_count_as_of
FROM global_markets.stocks_short_interest AS si
INNER JOIN share_counts AS sc ON sc.ticker = si.ticker
WHERE si.settlement_date = '2026-06-30'
ORDER BY pct_of_shares_out DESC

Ranked by percent of shares outstanding (share counts as of July 24, 2026), GME leads at 12.45%, and leads on days to cover too, at 10.75. Then they part company. MU ranks second on percent shorted (2.8%) yet sits dead last on days to cover, at the 1 floor: plenty of the company is sold short, and the volume is deep enough that the exit is one day wide. AAPL is the mirror image, the lowest percent shorted of the six at 0.96%, yet its 1.73 days puts it mid-pack. Two screens, two "most crowded" answers.

Who uses days to cover, and how?

Short sellers sizing the exit. A short position's real risk is not just the price, it is the door. In a name at ten days to cover, if everyone reaches for the exit at once the tape cannot absorb the covering at anything like normal prices. Position size gets cut accordingly.

Squeeze hunters screening for setups. Momentum traders run screens like the panel above, high days to cover, filtered for tradable liquidity, for names where a rally could meet crowded shorts. It is the precondition they screen for, not the event.

Risk desks flagging crowding. Funds track the measure across a whole book: if several positions are also somebody else's crowded short, the book carries a correlated exit risk no single stock's chart shows. Days to cover, alongside daily short volume and borrow costs, is how that gets measured.

Does high days to cover predict a short squeeze?

No, it describes crowding, it does not forecast it. A squeeze runs through the borrow: the cost to borrow shares climbs, lenders recall shares they had lent out, brokers issue buy-ins that close a position at the market, and margin clerks demand more collateral as a rising price makes each short share costlier to hold. Short sellers buying back into a thin tape push the price higher, and the next round begins. Days to cover measures the buying queued behind that door; it says nothing about whether anyone will open it.

At this one settlement 91 liquid names sat between five and ten days and 5 above ten, far more than ever produce a headline squeeze. High days to cover is common; squeezes are not. The ratio also lags: GME's had already collapsed to 1 days at the January 29, 2021 settlement, mid-squeeze.

How fresh is the number?

Days to cover is not a daily statistic. Short interest is exchange-reported twice a month, and the file arrives well after the settlement date it covers:

SELECT settlement_date,
       formatDateTimeInJodaSyntax(settlement_date, 'MMMM d, yyyy') AS settlement_label,
       formatDateTime(toDate(min(_ingest_time)), '%Y-%m-%d') AS arrived_in_warehouse,
       dateDiff('day', settlement_date, toDate(min(_ingest_time))) AS lag_days,
       count() AS names_reported,
       if(count() < 1000,
          toString(count()),
          concat(toString(intDiv(count(), 1000)), ',', leftPad(toString(count() % 1000), 3, '0'))) AS names_reported_fmt
FROM global_markets.stocks_short_interest
WHERE settlement_date >= '2026-03-01'
  AND settlement_date <= '2026-06-30'
GROUP BY settlement_date
ORDER BY settlement_date

8 settlements across four months, two a month, never daily. The June 30, 2026 file covering 22,207 names landed 11 days after its settlement date; the March 13, 2026 file took 19. The "current" days to cover on any website, every number here included, is a photograph of positioning from one to several weeks ago. Why short interest is two weeks old walks through the calendar.

FAQ

What is a good or high days to cover?

There is no official threshold; the market's distribution is the guide. At the June 30, 2026 settlement 323 liquid names sat at the reported floor of one day, 91 reached five to ten days, 5 exceeded ten. Judge against the volume tier too: five-plus days occurs in 23.9% of sub-1M-volume names, 3.6% of the heaviest-traded.

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

Yes, two names for one calculation: short interest divided by average daily volume. Some sources abbreviate it SIR. Vendors differ in the volume window they average, so ratios vary slightly site to site.

Does a high days to cover mean a short squeeze is coming?

No. It describes how crowded the short side is relative to daily volume, a precondition, not a prediction. Squeezes run on borrow costs, share recalls, buy-ins and margin calls; far more names carry high days to cover than ever squeeze, and GME's own ratio had already fallen to 1 days by the January 29, 2021 settlement, mid-squeeze.

Is days to cover updated daily?

No. Short interest is reported twice a month and published on a lag, so days to cover updates roughly every two weeks. March through June 2026 brought 8 settlements; the June 30 file reached this warehouse 11 days after settlement.

Can days to cover fall without short sellers covering?

Yes, constantly: a volume spike grows the denominator and shrinks the ratio with no change in the short position. GME printed 1 days at the January 29, 2021 settlement with 21.4 million shares still short, the squeeze's enormous volume, not a completed unwind, accounts for most of the drop.

Every figure above is a stored query over the exchange-reported short interest file, expand any panel's SQL, or pull days to cover for your own watchlist on the Strasmore terminal.