How to Read an Economic Calendar: CPI, Payrolls and Rate Decisions

An economic calendar tells you when uncertainty is scheduled. It does not tell you which way a price will go, and the way these pages are usually presented — a forecast, a previous value, an impact rating in red — quietly suggests otherwise.

Used properly it is one of the most useful things on a trading screen. Used as a prediction tool it is one of the most expensive.

The releases that actually matter

For currencies and gold, a small handful do most of the work.

Central bank rate decisions. The Federal Reserve above all, because the dollar is one side of most major pairs; then the ECB, the Bank of England and the Bank of Japan for their own currencies. The decision itself is often less important than the press conference and the projections that follow it, because those are what move expectations about the next decision.

Inflation. US CPI is the single release most likely to reprice Fed expectations, and with them the dollar, gold and every dollar pair.

Employment. The US jobs report — payrolls, unemployment and wage growth in one release — moves markets hard, and the initial move is frequently reversed within minutes as the detail is digested.

Everything else is context. A retail sales number matters when it changes the interest-rate story and is ignored when it does not.

Our economic calendar lists these from each institution's own published schedule, with the source linked beside every entry.

Why a good number can send a price down

This is the part that makes calendars confusing, and understanding it is most of the value.

What moves a price is not the number. It is the gap between the number and what was already expected.

The expectation is in the price before the release. If everyone anticipates strong inflation, positions are already set for strong inflation. A strong figure that is merely as strong as expected changes nothing — the move happened over the preceding week. A strong figure that is less strong than the market had positioned for produces a fall, and to somebody watching only the headline it looks irrational.

This is why "trade the news" is much harder than it sounds. You are not forecasting the number; you are forecasting the difference between the number and a consensus you cannot fully observe, and then the market's interpretation of that difference.

What happens in the seconds around a release

Three things, all of them working against a trader who has not planned for them.

Spreads widen. Liquidity providers pull back before a known event because they do not want to be the ones filling orders into a jump. A pair that costs 0.8 pips at 12:29 can cost several times that at 12:30:01. That widening is not a fault; it is the market pricing its own uncertainty.

Fills move. With fewer quotes available, a market order can fill some distance from the price you clicked. This is slippage, and it is at its largest exactly here.

Stops get reached that the mid-price never approached. Your stop triggers on the bid or the ask, not on the middle. A spread that widens by four pips can take out a stop sitting three pips away while the market itself never traded there.

None of this is unusual behaviour. It is normal, predictable, and the reason the defensive use of a calendar is the reliable one.

Using it defensively

The calendar's most valuable function is telling you when not to have a position you cannot afford to have.

• Before you open a trade, check what is scheduled for that instrument in the hours ahead. A three-day swing position opened the day before a rate decision is a bet on the decision whether or not you intended it to be.
• Reduce size, or stand aside, around the big ones. Many experienced traders do exactly this, which is the opposite of how news trading is usually sold.
• Do not place a stop within ordinary spread-widening distance of price before a release.
• Know which releases touch your pair. A Bank of Japan decision is a USD/JPY event and largely irrelevant to GBP/USD. Our calendar tags each entry with the instruments it moves.

What the columns mean

Time — usually given in a fixed reference zone. Ours are in UTC. Most US data lands at 12:30 UTC: half past eight in New York, half past one in London.

Previous — last period's figure, sometimes revised. Revisions matter more than people expect: a strong number that comes with a large downward revision to the last one is not the strong number it appears to be.

Consensus or forecast — a survey of economists, and the thing the market is measured against. We deliberately do not publish one: it is somebody else's proprietary survey, and printing an expected number beside a technical signal edges towards a prediction rather than a schedule.

Impact rating — an editorial judgement about how much a release usually moves things. Reasonable as a filter, and worth remembering that it describes the past.

A note on the central banks

Rate decisions have a rhythm worth knowing.

The Federal Reserve announces at 18:00 UTC with a press conference half an hour later; four of the eight annual meetings include updated projections, and those are the ones that move the dollar most, because the path matters more than the step.

The ECB decides at 12:15 UTC with a press conference at 12:45 — the conference usually moves the euro more than the decision.

The Bank of England announces at 11:00 UTC, with the vote split published alongside; a decision that was expected can still move sterling if the split was not.

The Bank of Japan has no fixed announcement time. It publishes when the meeting ends, normally late morning in Tokyo, and the delay itself is read as a signal — the yen can move before the text appears.

Should you trade the news at all?

It is the hardest condition to trade and the most expensive to get wrong, which is precisely the opposite of how it tends to be marketed to beginners.

If you do, the sizing has to reflect that spreads will widen and fills will move — which in practice means a smaller position than usual, not a larger one. Our position size calculator works the size from the stop, and a wider stop for news conditions produces a smaller position automatically.

Nothing on this page is a recommendation. The calendar is a schedule of when things are likely to move, not an indication of which way.