ARIMA as an analytical procedure: relative efficiency and effectiveness of a limited number of observations
Data from 14 firms are used to investigate whether predictive ability, efficiency, or effectiveness of the autoregressive integrated moving average (ARIMA) model is increased with an increased number of observations in the database used for the model. The data are used to construct ARIMA models base...
Gespeichert in:
Veröffentlicht in: | RBER, review of business and economic research review of business and economic research, 1991-04, Vol.26 (2), p.1 |
---|---|
Hauptverfasser: | , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | Data from 14 firms are used to investigate whether predictive ability, efficiency, or effectiveness of the autoregressive integrated moving average (ARIMA) model is increased with an increased number of observations in the database used for the model. The data are used to construct ARIMA models based on 23, 35, and 47 monthly observations. The models are used to produce account balances for months 24, 36, and 48, respectively. Errors are then seeded into the months being predicted, and the relative efficiency and effectiveness of the models constructed are evaluated. The results provide preliminary evidence that a longer model estimation period may not significantly improve model efficiency or effectiveness. |
---|---|
ISSN: | 0362-7985 1058-3300 1873-5924 |