In PROC ARIMA, if there is a discrepancy between the ID variable type and the INTERVAL= specification on the FORECAST statement, then the values of the ID variable may not be extrapolated for the LEAD= observations, and no error message is issued.

For example, if the ID variable is a SAS datetime variable with the time portion set to 00:00:00 for all observations, and the INTERVAL= option specifies a DATE interval, such as MONTH, then the ID variable values will not be extrapolated for the LEAD= observations.

To circumvent the problem, make sure that the INTERVAL= specification agrees with the type of the ID= variable. In the above example, a DATETIME interval, such as DTMONTH, must be used so that the ID variable values can be extrapolated. The ID variable types expected by PROC ARIMA are SAS date, time and datetime variables.