TD_MAMEAN forecasts a user-defined number of periods into the future, that is the number of periods beyond the last observed sample point in the series.
The TD_MAMEAN function uses available historical data to forecast activity h periods ahead. It is classified as a flat forecaster because given t historical observations, the forecast for any period beyond t+1 is equal to the forecast for t+1. Another use of this forecaster is to forecast one-period ahead. If historical data is available up to and including time t, then it can be used to forecast the value at time t+1. When the actual time t+1 value becomes available, it can be used to forecast the value at time t+2, and so on.