Economic calendars show a forecast, but not how much economists disagree. A surprise of 0.2 means something different if forecasts ranged from 0.1 to 0.3 versus from minus 0.5 to 0.8.
Some data providers show the range of forecasts. I started noting the high and low forecast for UK CPI and US payrolls.
What I found over a year:
When the actual was outside the forecast range, the five-minute move was usually at least twice the move for surprises inside the range.
Inside the range, even large "surprises" against the median often faded within 15 minutes.
It's a simple filter: is the number outside the range? If yes, expect a real move. If no, it's likely to be noise.
Do others use forecast ranges? Where do you get them?
A calendar lists three numbers for each release: previous, forecast, and actual. A common mistake is comparing the actual with the previous number and calling it a good or bad result. The market trades the surprise, which is actual versus forecast.
A reading of 3.0 percent when the last reading was 2.5 sounds strong. If the forecast was 3.3, it's a miss, and price drops. Many traders learned this the hard way.
Things get more complicated when forecasts themselves are uncertain. The consensus is an average of economists' estimates, and the range of estimates matters. A tight range means a miss of 0.1 is big. A wide range means even a 0.3 deviation might be within noise.
Whisper numbers add another layer. Informal expectations among traders may differ from the published consensus, particularly after strong moves in the lead-up. When the actual number meets consensus but beats the whisper, the market can still jump.
Positioning matters as well. A mild miss can trigger a large move if the market was positioned for a beat.
How is the surprise measured when preparing for a release, and is there a go-to source?
