Short interest and its derived ratios are the standard metrics for gauging how heavily a stock is sold short — i.e., how large and crowded the bearish bet against it is. Short interest is the total number of shares that have been sold short and not yet bought back (covered). On its own it's just a share count; its analytical value comes from two normalizations: short interest as a percent of float (the squeeze-potential gauge, scaling the short position against tradable supply) and days to cover, also called the short interest ratio (scaling the short position against daily liquidity). The core tension of these metrics is that they cut two ways at once — a heavy short position is simultaneously fuel for an upside short squeeze and a signal that informed, often-correct sellers are betting against the stock. The data is also stale by construction, which limits how literally it can be read.

How it's calculated

• Short interest (SI): raw count of shares sold short and still open. Reported by broker-dealers, aggregated by FINRA and the exchanges.
• Short interest % of float: SI ÷ free float × 100. Float is the freely-tradable share count (shares outstanding minus restricted/insider/closely-held shares). This measures what fraction of the actually-available supply is short. Some data vendors report % of shares outstanding instead of % of float — these differ, sometimes materially, for companies with large insider holdings. Check which denominator a source uses.
• Days to cover (short interest ratio): SI ÷ average daily trading volume. The result is a number of days — an estimate of how long it would take all shorts to buy back their positions at normal trading volume. The average-volume window is usually ~20–30 days; the exact window varies by vendor, which changes the number.

Both ratios rise with the short position, but they answer different questions: % of float measures crowdedness relative to supply; days to cover measures how trapped shorts are relative to liquidity. A stock can be high on one and low on the other (e.g. a high-float, very-liquid mega-cap can have a large % short but low days to cover).

How to read it

Per StockTitan, Schwab and similar desks, common (informal, not regulatory) reference bands:

• Short interest % of float: under ~5% is unremarkable; ~10%+ is elevated; ~20%+ is high; 30%+ is rare and flags significant squeeze risk.
• Days to cover: under ~1–2 is easily covered; ~5+ is elevated; ~10+ is high (shorts would need many days of normal volume to exit, so any forced buying compounds).

These thresholds are conventions, not edges — treat them as a screen, not a signal. The combination that flags maximum squeeze potential is high % of float and high days to cover together: a large position that is also hard to exit. That is the bullish-on-a-catalyst read. The same reading is also, plainly, evidence of crowded bearish conviction — see the evidence section.

How it's used in practice

• Squeeze setup screening (long). Traders screen for high SI% of float + high days to cover, then wait for an upside catalyst (earnings beat, positive news, technical breakout) to force shorts to cover into thin supply, amplifying the move. The metrics identify fuel; they do not provide the spark — without a catalyst, a high short position can persist for many months. Cross-link the Short Squeezes node for the mechanics of how this actually plays out.
• Sentiment / crowding gauge. Rising short interest signals growing bearish conviction (or growing hedging demand — see caveats); falling short interest signals covering. Some traders watch the change in SI more than the level.
• Contrarian and confirmation reads. Bulls treat extreme shorting as a contrarian squeeze setup; bears treat it as confirmation that sophisticated money is positioned against the stock. Both readings use the same number — which is correct depends on context the metric alone cannot supply.
• Liquidity/risk awareness. High days to cover warns that exits (in either direction) can be violent and gappy, which matters for position sizing and stop placement.

Data source, cadence & the staleness problem

In the U.S., short interest is reported by broker-dealers to FINRA and published twice a month (a mid-month settlement reading around the 15th and an end-of-month reading), per FINRA. Critically, the data is reported and disseminated with a lag: positions are reported by the second business day after the settlement date, and FINRA makes the consolidated data available for publication on roughly the 7th business day after the reporting settlement date. The practical consequence: by the time you read a short-interest figure it can be a week-plus old and reflect positioning from up to ~two weeks prior — and it can be badly wrong intramonth, especially in fast-moving names where shorts cover (or pile in) between reporting dates. This is the single most important caveat: short interest is a lagging snapshot, not real-time positioning. (Some vendors estimate daily short data from securities-lending feeds; those are proxies, not the official FINRA figure.)

Strengths & limitations

Strengths. Cheap, standardized, broadly available; genuinely useful for flagging crowding and squeeze fuel; the change in SI is a reasonable sentiment proxy.

Key limitations and the #1 misuse:

1. Staleness (above). Trading a two-week-old snapshot as if it's live positioning is the most common error. 2. High SI is not purely bearish/directional. A large short position can be a hedge, not a bet that the stock falls. Common non-directional sources: convertible-bond arbitrage (long the convertible, short the underlying stock as a delta hedge), merger/risk arbitrage (short the acquirer in a stock-for-stock deal), and ETF/index creation-redemption and basket hedging. For such names, high short interest carries little directional information. Always ask why a stock is heavily shorted before reading it as a bearish signal. 3. Squeeze ≠ likely. High SI% + high days to cover is necessary but not sufficient for a squeeze; most heavily-shorted stocks never squeeze, because the catalyst never arrives and the shorts are often right.

Standing & evidence — the two-sided record

Both of the following are well-documented, and they coexist; an honest read holds both at once:

• The squeeze story (real but rare). When a crowded short with high days to cover meets an upside catalyst, forced covering can drive sharp, outsized rallies (the GameStop episode of January 2021 being the canonical extreme). These events are real but infrequent relative to the universe of heavily-shorted stocks.
• The academic record points the other way on average. A substantial body of peer-reviewed research finds that high short interest predicts negative future returns — i.e., on average heavily-shorted stocks subsequently underperform, consistent with short sellers being informed traders. Asquith, Pathak & Ritter (2005) document underperformance of high-short-interest stocks; Boehmer, Jones & Zhang and related work find short sellers are informed; Rapach, Ringgenberg & Zhou (2016) find aggregate short interest is among the strongest predictors of market returns. The information content is stronger around news and negative-news days (Engelberg, Reed & Ringgenberg).

The reconciliation: the average heavily-shorted stock drifts down (informed shorts win the base case), while a small minority squeeze violently upward (the tail the squeeze trader is hunting). A high short-interest reading is therefore not a directional buy signal — it is a flag that both a negative-drift base case and a low-probability/high-magnitude squeeze tail are in play, and external context (catalyst, the reason for the shorting) decides which dominates.

System relevance

This is a definition/metric node. The squeeze mechanics and trade construction live in the Short Squeezes node — cross-link there rather than duplicating. For any consuming system (e.g. the Augustus trade-setup agent): treat a high SI% / high days-to-cover reading as a context flag, not a signal, always tag it with its as-of date (and never assume it's current), and check whether the short position is plausibly a hedge (convertible/merger-arb/ETF) before reading it directionally. Pair it with a live catalyst before treating it as a bullish squeeze setup; absent a catalyst, the academic base rate (negative drift) is the more likely path.

Sources

• FINRA — Equity Short Interest Data and Short Interest Reporting (reporting cadence: twice monthly; ~7-business-day publication lag). finra.org
• StockTitan — Short Interest Explained: What It Is and Isn't (% of float vs % of shares outstanding; non-directional/hedging sources of short interest).
• Charles Schwab — Short Interest Monitor Explained (threshold bands for % short and days to cover).
• Forex.com — What is the short interest ratio? (days-to-cover = SI ÷ avg daily volume definition).
• Asquith, Pathak & Ritter (2005), Short Interest, Institutional Ownership, and Stock Returns (Journal of Financial Economics) — high short interest → negative forward returns.
• Boehmer, Jones & Zhang; Diether, Lee & Werner; Engelberg, Reed & Ringgenberg — short sellers are informed; effect concentrated around news.
• Rapach, Ringgenberg & Zhou (2016), Short Interest and Aggregate Stock Returns (Journal of Financial Economics) — aggregate short interest predicts market returns.
• AlphaArchitect / Quantpedia — summaries of the "short interest effect" anomaly (high SI → underperformance).
• WallStreetMojo; Mergers & Inquisitions — convertible arbitrage as a delta-hedged long-convertible/short-stock source of non-directional short interest.

Note: precise band thresholds (10% / 20% / 5 days, etc.) are widely-cited industry conventions, not regulatory or statistically-validated cutoffs — treat as screening heuristics.