A detail that catches people out: data releases often include revisions to the previous figure, and the market sometimes reacts more to the revision than to the headline.
Examples I've seen:
US jobs: headline beat, but the previous two months revised down significantly. The dollar spiked and then reversed within minutes.
UK GDP: monthly figure in line, previous month revised up. GBP moved more than the headline would suggest.
Retail sales: a strong headline was offset by a revision in the control group.
How I handle it:
Before the release, I know the previous figure and consensus.
At the release, I read the revision too, not just the headline.
No scalp in the first minute. The first move often follows the headline, then the market digests the details.
If the revision contradicts the headline, I expect two-way price action and stay out or trade small.
The first minute after a release with a big revision is often the most expensive one: wide spreads and a reversal.
Which releases have the most important revisions in your experience? US payrolls is the obvious one, but I wonder about others.
Chinese data lands during the Asian morning, and the Australian dollar reacts first. Iron ore, trade, and growth figures all feed through to AUD and NZD.
The reaction time is quick. Algorithms read the headline and trade within milliseconds. A retail trader watching the chart sees a candle forming before the number has even been read.
The first move can be large and wrong. Headline numbers are often followed by revisions, details, or components that change the interpretation. The second move, which comes a minute or two later, is frequently in the opposite direction.
The sensible stance for a scalper is to watch rather than trade the first spike. The retracement and the follow-through give a better read, and the spreads settle in a minute or so.
Trading it means accepting that the first move is noise, and the second one is information.
After Chinese data, which is more tradable: the first move or the second?
