Estimating Demand Uncertainty Using Judgmental Forecasts

Measuring demand uncertainty is a key activity in supply chain planning. Of various methods ofestimating the standard deviation of demand, one that has been employed successfully in therecent literature uses dispersion among expertsacirc; forecasts. However, there has been limitedempirical validation of this methodology. In this paper we provide a general methodology forestimating the standard deviation of a random variable using dispersion among expertsacirc; forecasts.We test this methodology using three datasets, demand data at item level, sales data at firm levelfor retailers, and sales data at firm level for manufacturers. We show that the standard deviationof a random variable (demand and sales for our datasets) is positively correlated with dispersionamong expertsacirc; forecasts. Further we use longitudinal datasets with sales forecasts made 3-9months before earnings report date for retailers and manufacturers to show that the effects ofdispersion and scale on standard deviation of forecast error are consistent over time