EURODECISION handles forecasting issues using a quantitative approach. These quantitative methods break down into two sub-families: endogenous and exogenous methods.
Endogenous forecasting methods, also called extrapolative, are used when the study of the records of the series gives sufficient elements on its structure to construct a powerful forecasting method. These methods reproduce the structure of the past to forecast the future. Examples of these methods are: moving average smoothing, exponential smoothing, Holt method and Holt-Winters method.
Exogenous forecasting methods, also said to be explanatory, are used when:
- the analysis shows that the influence of exogenous factors explains the fluctuations in the series
- and it is easy to evaluate the values of these factors (which will be used to estimate the future values of the series to be forecast).
These methods select explanatory factors, learn their influence on the records and deduce the future values of the series.
Linear regression is an exogenous forecasting method.