What short-selling data cannot tell you

Official short-position data is unusually trustworthy: it is compulsory, regulator-collected, and published on a fixed schedule. That reliability makes it tempting to ask it questions it cannot answer. This page is the honest inventory of what these numbers leave out, and it is longer than the list of what they contain.

It cannot tell you why a position exists

This is the largest blind spot, and no disclosure regime addresses it. A reported short position carries no reason, and the reasons are not equivalent:

• A directional bet that the share price will fall.
• A hedge against a long position elsewhere, in a convertible bond, in an index, in a related company in the same sector.
• Arbitrage: merger arbitrage around an announced acquisition routinely produces large short positions in the acquirer that express no view about it at all.
• Market-making and liquidity provision, where inventory moves short as a by-product of filling other people's orders.

All of these print as the same number. A company with a large disclosed short position may be the subject of intense scepticism, or may simply be one leg of a trade about something else entirely. The data cannot distinguish them, and neither can we.

It is a floor, not a census

In the UK and Europe, only positions above a disclosure threshold are counted at all. A hundred funds each holding 0.19% of a UK company would contribute nothing to the published aggregate. The true level of shorting is therefore always at least the published figure and frequently more, by an amount nobody outside the regulator can measure. Australia's much lower reporting floor and the US's firm-level FINRA reporting make their data closer to a census, which is precisely why those figures should not be compared directly with the UK's; see who must disclose a short position.

It cannot tell you who holds the position

In the US and Australia always, and in the UK since July 2026, published data is aggregate and anonymous. An aggregate of 3% might be a single fund with a concentrated position, or six funds at 0.5% each, materially different situations that print an identical number. Europe is the exception: its registers still name holders above 0.5%, which is why the position holder pages exist for European markets and cannot exist for the others.

It cannot tell you when a position was opened, or at what price

Disclosure captures a position's size on a date, not its history. A position at 2% today might have been opened last week or held for two years. There is no entry price, no profit or loss, and no indication of how long the holder intends to keep it. A position that has been flat for months and one opened this morning look the same.

It cannot tell you what happens next

High short interest is sometimes described as a predictor: of a fall, or of a squeeze. It is neither. Elevated short interest has preceded sharp declines and sharp rallies, and most often has preceded nothing in particular. The mechanics of why a high figure does not imply a squeeze are set out in short squeeze basics. We publish no signals, no forecasts and no ratings, and we do not tell you what any figure implies for a share price, because the data does not support such a claim.

It cannot be read across companies without care

The denominator is issued share capital, which varies enormously in liquidity and free float. A 3% short position in a large, heavily traded company and a 3% position in a small illiquid one are different in every respect except the number. Comparing a company against its own history is the comparison the data genuinely supports; ranking unlike companies against each other is the comparison it supports least, even though a league table is the most natural way to present it.

It says nothing about the company's prospects

A disclosed short position is a fact about who holds what, not a judgement about the business. Short sellers are wrong regularly. Treating a high figure as evidence of wrongdoing, distress or impending failure reads a verdict into a number that contains none.

What it does tell you

Having said all that, the narrow thing this data does well, it does very well. It is compulsory rather than voluntary, collected by a regulator rather than self-reported to a vendor, published on a fixed schedule, and revised in public when it turns out to be wrong. Within one market, over time, it gives you a reliable, official answer to a genuinely useful question: how much disclosed short exposure exists in this company, and is that more or less than before?

That is the question this site is built to answer, and it is the only one we claim to answer. How we handle the data end to end is documented in the methodology.

This page is educational. HowShorted is a data reference, not an adviser: nothing on this site is investment advice.