The treatment of missing values varies with the forecasting model. For the smoothing models, missing values after the start of the series are replaced with one-step-ahead predicted values, and the predicted values are applied to the smoothing equations. See Chapter 15, Forecasting Process Details, for greater detail about how missing values are treated in the smoothing models. For MODEL=IDM, specified missing values are assumed to be periods of no demand.

The treatment of missing values can also be specified by the user with the SETMISSING= option, which changes the missing values prior to modeling.

Even though all of the observed data are nonmissing, using the ACCUMULATE= option can create missing values in the accumulated series.