The error of a selected forecasting method divided by that for a benchmark forecast. The term is commonly used in judgmental forecasting. It is also used in quantitative forecasting. See Theil’s U and Relative Absolute Error.