Adjustment of forecasts with model consistent expectations

This paper provides a practical procedure to support the judgemental adjustment of statistical forecasts, for the class of problems where judgement is necessary to compensate for omitted variables in the model. The analysis suggests that many casual adjustments in practice may be prone to a double-counting bias, and that the use of model-consistent expectations, for the omitted variables, should provide a basis for judgemental adjustments free of such bias. The procedure is applied to a real case study, from the petroleum industry, based upon capital cost forecasting for major offshore facilities.