Consensus forecast

A survey result, not a market price. A data provider polls a panel of forecasters ahead of a scheduled release and publishes the median or the mean of their answers, which appears on an economic calendar beside the figure due. It is the reference point a release is read against, because a market has already positioned around what it expected: an outcome equal to the survey is capable of moving nothing at all, however dramatic the level it reports.

Because it is a survey, its composition matters and is rarely visible. Panels differ in size and in membership, contributors update at different times, and providers close their polls at different points before the release, so two calendars can print different consensus figures for the same event. Where the panel is small, one revised submission moves the published figure, and the dispersion of the answers, which describes how much disagreement sits behind the central estimate, is usually not published at all.

It is also not the same thing as what the market expects. A survey of forecasters and the positioning of participants can diverge, particularly where a related release earlier in the week has already shifted views, so a market can react to an outcome that matched the survey exactly. Practitioners who follow this closely tend to read the survey alongside prices that carry an expectation directly rather than treating it as the market's own view.