Article ID: | iaor1989572 |
Country: | United Kingdom |
Volume: | 8 |
Start Page Number: | 381 |
End Page Number: | 398 |
Publication Date: | Dec 1989 |
Journal: | International Journal of Forecasting |
Authors: | De Gooijer Jan G., Klein Andre |
Keywords: | forecasting: applications, transportation: water |
This paper uses multivariate time series models to specify the maritime steel traffic flow in the port of Antwerp. The time series considered are the total outgoing and total incoming maritime steel traffic and the total steel production in the EEC. The obtained time series models provide useful insight into the general behaviour of the maritime steel traffic flow during the period 1971-82. In particular, they provide a quantitative interpretation of important changes which took place in the European steel industry during that period. The multivariate time series models produce forecasts which are a substantial improvement over those obtained by univariate time series models. This is especially the case for the series of total incoming maritime steel traffic in the port of Antwerp, when differencing and transformation of the original data are applied.