The "forecast" or "consensus" column in an economic calendar is usually the median of estimates collected from economists by a data provider. Knowing that helps interpret surprises.
Useful details:
Forecast range: the spread between the highest and lowest estimates. A result outside the range tends to produce stronger reactions.
Provider differences: calendars may show their own forecast rather than the survey median.
Late updates: consensus can shift in the days before the release as new data arrives.
Whisper numbers: the market's unofficial expectation sometimes differs from consensus, especially for high-profile releases.
A practical habit: for each major release, write down consensus, previous value and the forecast range. After the release, note the reaction size. Over time, this shows how sensitive each pair is to surprises of a given size.
Which calendar do you use, and do you check the forecast range?
Forecast range is the column I wish more calendars highlighted. When the high and low estimates sit 0.3 points apart on CPI, a print outside that band has moved EURUSD harder for me than a same sized miss against a tight consensus.
Provider differences bite occasionally. One calendar showed a forecast a tenth firmer than the survey median I use elsewhere, and I almost sized up into a print that was only surprising on that screen.
Whisper numbers matter most on FOMC week and big payrolls. I do not chase them, but if the whisper is clearly softer than consensus I expect a muted reaction to a mild miss.
My habit is writing consensus, previous, and range before the release, then noting reaction size in R on the pair I care about. After a few dozen rows the sensitivity pattern becomes obvious.
Do you weight the range more than the median when the dispersion is wide? I still trade the surprise against consensus alone most of the time.
