Method is the ?rst of two methods proposed by Mantrala and Rao (2001) and has been reviewed in Section 2.We use a simpli?ed version, with ?xed prices and for a single period. Furthermore, instead of asking the experts to reach consensus on the minimum, maximum and modus of the demand and then taking the average of these three consensus ?gures as the forecast, we calculate the averages over the minima, maxima and modi from all experts (thereby weighing their forecasts equally).

This method is included for two reasons. First, inclusion of this method in our study allows for a comparison with the very simple expert method (Method 4), in order to determine whether the slight additional sophistication introduced in this method leads to better forecasts. Second, it has not been tested (by the authors who proposed it).