How Short Interest Is Reported Around the World

Short interest — the number of shares that have been sold short and not yet bought back — is one of the most requested numbers in investing, and one of the least straightforward to obtain. No exchange or regulator anywhere in the world publishes a complete, real-time count. What regulators do publish differs enormously by country, and understanding those differences is the key to reading short interest data correctly.

📘
Key takeaways

No market publishes complete, real-time short interest. Every official number is either delayed or incomplete — usually one or the other by design.

Countries fall into two families: delayed full-market counts (US, Canada, Australia, Hong Kong) and threshold disclosures that are timely but only show large positions (EU, UK, Japan).

Threshold data is a floor: true short interest is at least the published figure, and often meaningfully more.

ORTEX publishes an Estimated Short Interest that bridges these gaps — anchored to the official figures and updated daily and intraday using securities lending market activity.

Why isn't short interest just... published?

When you short a stock, you don't file a public notice saying so. Your broker knows, and depending on where the stock is listed, a regulator will eventually require some form of report — but what gets reported, by whom, how often, and above what size varies by jurisdiction. The result is that "the" short interest figure for a stock is really a patchwork of national reporting regimes, each with its own blind spots.

Broadly, every major market uses one of two systems.

System 1: Everyone reports — but you find out later

In the United States, Canada, Australia, and Hong Kong, regulators collect short positions from across the whole market and publish a per-stock total. Think of it as a census: nearly everyone is counted, but the results take time to compile and publish. By the time the number is public, it describes a snapshot from days or weeks earlier — and positions have moved since.

Market	Published by	How often	Delay	Worth knowing
🇺🇸 United States	FINRA	Twice a month (mid-month and month-end)	~7–9 business days after the snapshot date	The most-watched short interest print in the world, and still up to three weeks stale on arrival
🇨🇦 Canada	CIRO	Twice a month	~4 business days	Covers TSX and TSXV listings
🇦🇺 Australia	ASIC	Daily	~4 business days	The most complete regime globally — a near-complete count every trading day
🇭🇰 Hong Kong	SFC	Weekly	~5 business days	Positions below a small threshold (HKD 30 million or 0.02% of issued shares) are excluded, so the total is a slight undercount

The defining trade-off of this family is completeness at the cost of timeliness. A twice-monthly figure with a week's processing delay means that for most of the month, the latest official number is one to three weeks old. Anything can have happened in between — which is exactly the gap the ORTEX estimate exists to fill.

System 2: You find out quickly — but only about the big positions

The European Union and EEA (under the ESMA Short Selling Regulation), the United Kingdom (FCA), and Japan (FSA) work differently. Instead of counting everyone, they require individual short sellers to publicly disclose their position once it crosses a size threshold — in the EU and UK, 0.5% of a company's issued share capital, updated at each 0.1% step up or down.

Think of it as an iceberg: the disclosures show you what's above the waterline, promptly and by name. Everything below the threshold is invisible.

This has three consequences worth internalizing:

The published total is a floor, not a count. Summing the public disclosures tells you the minimum short interest, contributed by large holders. A stock shorted by many funds each holding 0.4% could show zero flagged shorts while carrying substantial real short interest.
Positions "disappear" without closing. When a holder drops below 0.5%, they leave the public register — the position looks closed, but the holder may still be short up to just under the threshold. A sudden drop in aggregate flagged shorts often means a holder dipped below the line, not that they bought everything back.
You get names. Unlike the census-style regimes, threshold registers identify who is short — useful information the US system never provides.

A notable variation: Sweden additionally publishes a per-stock aggregate of all positions above 0.1%, giving a partial view into the range that's invisible elsewhere in Europe. ORTEX incorporates this automatically, always using the most complete figure available. Japan's regime follows the same holder-disclosure model with its own thresholds.

How ORTEX bridges the gaps

Both systems leave the same question unanswered: what is short interest today? ORTEX publishes an Estimated Short Interest to answer it — a daily figure, plus live intraday updates as securities lending activity is recorded through the day:

In delayed markets (US, Canada, Australia, Hong Kong), the estimate is anchored to the most recent official figure, then moved day by day based on activity observed in the securities lending market — where shares are borrowed before being sold short. Each estimate comes with upper and lower confidence bands: they narrow when the official data is fresh, and widen as time passes since the last official print.
In threshold markets (EU, UK, Japan), the estimate combines the disclosed positions with an estimate of the undisclosed, below-threshold portion, again informed by securities lending activity — turning the "floor" into an estimate of the full picture.
In all other markets — those with neither type of official reporting — the estimate is derived from securities lending activity alone.

When a new official figure is released, the estimate is recalibrated against it — so you may notice the estimate adjust on the day after an official publication. That's the model course-correcting against ground truth, and it's a feature, not a glitch.

One caveat worth knowing: shares are borrowed for reasons other than short selling, and ORTEX corrects for the largest of these before estimating — see Not all borrowing is short selling in the metrics guide.

Common questions

Which number should I trust — the official figure or the ORTEX estimate?
They answer different questions. The official figure is the authoritative historical record for its snapshot date. The estimate is the best available answer to "and where is it now?" — with confidence bands that are honest about the uncertainty. Use the official print for the past, the estimate for the present.

Why did the flagged shorts for my European stock suddenly drop?
Most often, one or more holders reduced their position below the 0.5% public threshold. They exit the register at that point, even though they may remain short just below it. This is the single most common misreading of European short data.

Is short interest self-reported by hedge funds?
No — a common misconception. Official figures come from regulated reporting by brokers and market participants under each jurisdiction's rules, and ORTEX's daily signal comes from securities lending market data. None of it is voluntarily supplied by short sellers themselves.

Can short interest really exceed 100% of a company's shares?
Yes. If a borrowed share is sold short, and the buyer lends it out again, the same underlying share supports two short positions. Repeated enough times, total short interest can legitimately exceed 100% of shares outstanding — no naked shorting required.

Related guides

Short Interest Metrics Explained — what each ORTEX short interest metric means and how to read it

📘
Coverage and updates
ORTEX publishes short interest estimates for all markets worldwide. Estimates update every trading day, with live intraday updates on top; official figures update on each regulator's schedule shown above.

ORTEX data guides are for information and education only and are not investment advice. Data is provided "as is" and may contain errors or omissions; see our terms of service for details.