The model type named Auxiliary is intended for series for which you lack recent reliable data, but at the same time do have access to recent data for another series that implicitly contains information about the series you want to forecast. For example, assume that you always have access to load outcome values for the total consumption within an area for the recent days. Since the data is up to date, the forecast models can quickly adapt to new data with accurate forecasts as a result. Also suppose that the consumption you need to forecast is only for a fraction of the customers in the area. But this data come from another system which is updated less regularly so your recent history data will always be a couple of months old. If we write the series we wish to forecast ( representing a fraction of the customers) as a function of the auxiliary series we know we can do accurate forecasts from (the total consumption), we have an alternative way of forecasting by applying the function defined on the forecasted values. The model type Auxiliary both points out the series which forecasted values we want to start off with, also the model settings determines how the program should proceed when it automatically estimates the function or relationship between the two series. Specify which series should be pointed out as auxiliary in the list (see below)