Open any economic calendar with the default filters and you get roughly forty rows a day: manufacturing surveys from countries you do not trade, second tier housing data, speeches by officials who have no vote. A trading desk does not work from that page. It works from a short list, usually built on Sunday, of the releases that can shift expectations for a central bank, plus the auctions and expiries that shift flow. Everything else is background.
The four numbers in every row
A calendar row carries the event name, the time, the previous figure, the consensus forecast and, once released, the actual. The move happens in the space between forecast and actual, because the forecast is already in the price. This is why a strong number can produce nothing at all. If a market has spent three weeks pricing a hot inflation print and gets one, the information content of the release is close to zero.
The line most retail traders ignore is the revision. When the previous month's figure is restated, the market has to reprice two months of data at once. A headline that beats consensus while the prior month is revised sharply lower can leave the trend of the series unchanged, and price will follow the trend rather than the headline. Payrolls in particular is a series where the revision routinely matters more than the print, which is covered in the guide to trading NFP.
Not all high impact rows are high impact for you
The red and orange flags on a calendar are the provider's classification, generally derived from historical price reaction. They are a filter, not a ranking. Build your own tier list instead, organised by the instruments you actually trade.
| Tier | Typical events | How a desk treats it |
|---|---|---|
| Tier 1 | Central bank decisions and press conferences, CPI, employment reports | Positions sized down or flattened in advance, calendar blocked out, no new entries in the surrounding window |
| Tier 2 | PMIs, retail sales, GDP releases, wage data, central bank minutes | Traded normally with wider stops, watched for confirmation of the existing narrative |
| Tier 3 | Second tier surveys, regional indices, most scheduled speeches | Ignored unless the market has already made that series the story of the week |
Tiering is not permanent. A series moves up a tier when it becomes the variable the central bank says it is watching, and drops back when attention shifts. In a period when a bank has told the market it is focused on services inflation, the services component of a CPI report can outweigh the headline entirely.
What happens in the release window
Statistical agencies publish on a fixed clock, to the second. Ahead of the print, market makers widen quotes and pull size from the book because nobody wants to be the passive side of the first tick. Algorithms parse the machine readable feed and act within milliseconds, so the first candle you see on a chart is already the result of thousands of decisions. Retail platforms are receiving a price, not competing for one.
Practically, that means the first move is not a trade you can enter by reading the number and clicking. Spreads widen, slippage increases on both market and stop orders, and the initial direction is frequently reversed within a few minutes as the detail beneath the headline gets read. Traders who work releases seriously tend to have a plan for the retracement rather than the spike.
Calendar times are published in a specific time zone and adjusted for daylight saving on different dates in different regions. Twice a year, a release you have watched for months lands an hour earlier or later than your habit expects. Check the zone setting on your calendar against your platform clock at every clock change.
Building the week from it
The routine that works is short. On Sunday, list the tier one events for the currencies you trade, with the exact time in your local zone. Note which of your open positions sit in the path of each event. Decide in advance what happens to those positions: closed, halved, or held with a stop that already accounts for a wider range. Write it down before the week starts, because the decision made at 14:29 on the day is not the same decision.
Add the structural items that are not economic data at all: index rebalances, option expiries, month end fixings and bank holidays that thin liquidity in one region while the others trade. A quiet calendar with a Japanese holiday in it is not the same market as a quiet calendar without one. The session structure matters here as much as the events themselves.
Which calendar and which feed
Free calendars from the retail platforms and from sites like ForexFactory carry the same underlying releases, and the differences are in filtering, revision display and how the consensus is sourced. Consensus figures come from surveys of economists and vary slightly between providers, which is why one calendar can call a print a beat and another can call it in line. If you trade releases at all, use one calendar consistently rather than switching, so that your notes on past reactions are comparable. The practical differences between the popular options are set out in the calendar tools comparison and in the walkthrough of the ForexFactory calendar.
Institutional desks pay for a low latency machine readable feed. Retail traders do not need one, and buying speed you cannot act on is wasted money. What retail traders do need is an accurate clock, a clear note of what is scheduled, and a rule for what they do when the number lands. Trading around scheduled data is high risk, and the widened spreads and gapped fills of a release window are exactly where an oversized position becomes unmanageable.
"Most people use the calendar to find something to trade. Use it to find the hours when you should not be in the market and the rest of the week gets easier."
— Alex Onta, Executive Director, SINGUARD
Key Takeaways
- Price responds to the distance between consensus and actual, so a strong number that was already expected often produces no move at all.
- Revisions to the previous figure can reverse the meaning of a headline, and most retail traders never look at that column.
- Impact flags are a provider classification, so build your own three tier list around the instruments you actually trade.
- Decide what happens to open positions before the week starts, and include holidays, expiries and month end alongside the data rows.
Frequently Asked Questions
What does the impact colour on an economic calendar actually mean?
It is the calendar provider's own classification, usually based on how much the release has moved markets in the past. It is a starting filter, not a ranking of importance for your instruments. A high impact tag on a currency you never trade is irrelevant, while a medium tag on the central bank you are positioned against may matter a great deal. Most traders end up building their own tier list.
Why does price sometimes move against a good number?
Markets trade the change in expectations for future policy, not the number itself. A strong print that was already fully priced in can produce no move, and a strong headline with weak internals or a downward revision to the previous month can send price the other way. The revision line and the detail beneath the headline often carry more information than the figure the calendar shows.
Should I close positions before a high impact release?
That is a risk decision each trader makes with their own rules, and there is no universal answer. What is measurable is that spreads widen, depth thins and slippage on market and stop orders increases in the seconds around a scheduled release. Traders who hold through releases usually size smaller and accept that a protective stop may fill well past its level. Leveraged trading carries a high risk of loss in any case.