Automatic SARIMA modeling and forecast accuracy

Abstract

The crucial phase of the practical application of the Box-Jenkins methodology is the specification and verification of the suitable model. When it is made “manually” it is difficult and subjective. A new automatic SARIMA modeling method has been proposed. It is a procedure based on the classic model selection process; i.e., model specification, parameter estimation and diagnostic control. By use of a simulation study, the proposed automatic SARIMA procedure is compared with the authors´ own implementations of automatic SARIMA procedures based on the AIC and BIC and the in-sample forecasts evaluation from the point of view of model quality, as well as the accuracy of the forecasts obtained.

Keywords:

• SARIMA
• Model building
• Forecasting
• Automatic procedure
• Simulation

Additional information

Funding

This paper was written with the support of the Czech Science Foundation project “DYME – Dynamic Models in Economics” No. P402/12/G097.