Position date vs publication date

Every short-selling figure on this site carries two different dates, and reading one as the other is the easiest way to misunderstand the data:

• The position date, the day the position was actually held.
• The publication date, the day the regulator released the file saying so.

No regime publishes short positions in real time. By the time a figure is public it describes a state of the world that is at minimum a day old, and in Australia's case most of a week old. We store the two dates as separate fields and never conflate them; the figure shown on a company page is labelled with its position date, because that is the date the number is actually about.

How long the delay is

The lag is a deliberate feature of these regimes, not a technical limitation. It exists because a position holder is given time to report after crossing a threshold, and the regulator then needs time to collate and verify what it received before publishing.

• United Kingdom: holders must notify the FCA by 15:30 the working day after the position is held, and publication follows around midday. The FCA describes the published file as reflecting positions two working days earlier (T+2), though in practice the file we ingest has consistently carried more recent positions than that. We date each file by what it actually contains rather than by assumption.
• Australia: ASIC publishes on the fourth business day after the trade date (T+4). A file appearing on Friday describes the previous Monday. This is the longest lag of the daily markets, and it is the price of the lowest reporting threshold.
• United States: FINRA member firms report their positions as of two settlement dates a month (the 15th and the last business day), and FINRA publishes the consolidated figures about nine business days later. That is the longest lag of any market we cover, and it is set by regulation.
• Europe: national registers are updated as positions cross the 0.5% public threshold, typically the following business day. Because European disclosure is event-driven rather than a daily snapshot, an individual position's date reflects when that position last changed, not when the register was last refreshed.

Why a figure can sit unchanged for weeks

This surprises people more than the delay itself. Open a company page and you may find a position date from several weeks ago, sitting in a file published today. The number is not stale by mistake; it is carried forward at source.

Under threshold-based regimes, a holder reports when they cross a level, not every day they continue to hold. A fund that reaches 0.62% and then sits still has nothing further to report: no 0.1% step has been crossed in either direction. The regulator keeps publishing 0.62% with its original position date until something changes. So an old position date does not mean the data is broken; it means that position has not moved enough to require a new report.

This is also why Australia's daily-reporting model produces a position date that advances every single day while the UK's does not: ASIC requires a report each day regardless of whether anything changed.

Why published figures sometimes change afterwards

A published number is not necessarily final. Regulators amend and re-publish previously released figures: a late notification arrives, an error is corrected, or a position held back for verification is released into the aggregate. A figure you read last week may not be the figure in this week's file for the same position date.

We handle this by re-reading a trailing window of recent days on every run and reconciling what we already hold against what the regulator now says. A revised figure replaces the old one, and the revision is recorded (the old value, the new value and when we detected it), so a company page can show a revised marker rather than silently changing history. The methodology sets out the mechanics.

What this means in practice

• Do not read a published figure as today's positioning. It describes a date printed next to it, which is always in the past.
• Do not read an unchanged figure as an unchanged position. It means no reporting threshold was crossed, which is a much weaker claim.
• Do not compare a UK date to an Australian date. A file published the same day in both markets describes positions from different points in the past.
• Expect the recent past to be provisional. The most recent few days are the ones most likely to be revised.

Every figure on this site is traceable to the source file and publication date it came from, which is what makes these distinctions checkable rather than something you have to take on trust. What the data still cannot tell you, whatever its dates, is covered in what short data cannot tell you.

This page is educational. Nothing on this site is investment advice.