How to Read an Economic Calendar

The most precisely scheduled data in existence

Almost nothing in a release tracker happens exactly when it was supposed to. Software slips, hardware slips, films move by quarters. Official economic statistics are the exception: a national statistics agency publishes its release calendar up to a year ahead, down to the minute, and hits it. The date is not an estimate.

That precision exists because these figures move markets, and a release that leaked early or arrived late would advantage whoever knew first. Agencies therefore treat the publication timestamp as part of the product. For anyone building a calendar, this means economic entries can be trusted as forward-dated facts in a way that a game release date cannot.

This guide explains what the fields on a calendar entry mean. It is descriptive only — nothing here is investment advice, and none of it should be used to make trading decisions.

The four numbers

A typical calendar entry carries up to four values, and each answers a different question.

Previous is the figure reported last time the same indicator was published. It is the baseline, and it is the number most likely to have been quietly revised since it was first printed. Forecast is the consensus expectation, compiled from economists surveyed before the release — it is not a prediction by the agency and the agency has no involvement in it. Actual is the figure the agency publishes at the release time. Revision, where present, is a correction to a previously published figure.

The single most important thing to understand is that the interesting quantity is not the actual value but the distance between the actual and the forecast. A figure can be historically strong and still register as a disappointment because expectations were higher. This is why an entry showing only the actual number, with no forecast alongside it, has thrown away most of its information.

• Previous — last period's reported figure, subject to later revision.
• Forecast — consensus expectation from surveyed economists, not an official projection.
• Actual — the figure published by the agency at the scheduled time.
• Revision — a correction to an earlier published figure, itself a separate dated event.

Why revisions deserve their own entry

Economic statistics are estimates built from incomplete survey data, and they get better as more data arrives. A first print is published quickly with partial coverage; subsequent estimates incorporate more responses. Revisions are therefore normal, expected and built into the process rather than being errors.

They can also be large. An initial estimate revised substantially in a later release changes the picture retroactively, and if the record simply overwrites the old figure, the fact that the first print said something different disappears. That matters, because decisions were made on the first print.

The correct treatment is to record the revision as its own dated event linked to the original rather than editing the original in place. The history of what was believed and when is part of the data.

What impact ratings actually measure

Calendar entries carry an impact rating, usually presented as high, medium or low, or as a colour. It is easy to read this as importance. It is closer to expected volatility — a statement about how much markets have historically moved when this indicator surprises, not a judgement about economic significance.

Central bank rate decisions and headline inflation and employment figures sit at the top because they feed directly into policy expectations. A narrow sectoral index sits lower, not because it does not matter to that sector, but because it rarely moves broad markets.

Two things follow. A low-impact release can be highly relevant if it is the specific indicator you care about, and a high-impact release that lands exactly on forecast can produce almost no reaction at all. The rating describes potential, and potential is only realised when the actual diverges from the forecast.

Time zones are where calendars go wrong

Every serious problem with economic calendars is a time zone problem. Releases are scheduled in the agency's local time, which means the observed time in any other zone shifts when either zone changes for daylight saving — and those transitions do not happen on the same dates worldwide.

For several weeks each year, a release that normally lands at half past one in the afternoon in one zone lands at half past two. Calendars that store a local wall-clock time and apply a fixed offset get this wrong twice a year, reliably.

The correct handling is to store the release as an absolute instant with its timezone and render it in the reader's local time. If you are consuming a calendar rather than building one, it is worth spot-checking a known release around a transition date before trusting it.

Speeches, statements and scheduled uncertainty

Not every calendar entry has numbers. Central bank speeches, policy statements and press conferences appear on the same calendar with a date, a time and often a high impact rating, but no forecast and no actual — because the content is language rather than a figure.

These entries are scheduled uncertainty. The time is known precisely and the content is unknown entirely, which is a different shape from a statistical release where the content is bounded and the surprise is a matter of degree.

For tracking purposes they are still worth recording as events, because the value is entirely in knowing they are coming. There is nothing to check afterwards in the way there is with a data release, which is exactly why they belong on a forward-looking calendar rather than in a news feed.

A closing caution

Understanding the mechanics of a calendar entry is not the same as being able to use it. The relationship between economic data and markets is not stable, is not linear, and is heavily conditioned on what else is happening. Reading the fields correctly tells you what was published; it tells you nothing about what to do.

Everything in this category is recorded as a dated pointer to an official publication, with the agency's figures reproduced as published and attributed to the source. It is a calendar and an archive. It is not analysis, and it is not advice.