Econometric forecasting for container handling: ARMA model application

Since investments in maritime sector are very costly and irreversible; forecasting, critical in planning the future, has a particular importance in this field. As long-term forecasting is more important in investments, short term forecasting is more important in operations and distribution of port resources. The purpose of this study is to forecast monthly the sum of Turkey's import and export container, for the period from the ninth month of 2018 to the last month of 2019, and also to contribute to the decision-making of port managers in port operations. Short-term monthly estimations were made in order to follow the changes in the year. In accordance with this purpose, in this study, Box - Jenkins method, which includes dynamic structures, which is simple and fast in calculation, which is much better than other methods in short - term forecasts in the literature, has been selected. One of the most important results of the article is that the Akaike and Schwarz information criteria used in the estimation of the model, which is one of the steps of the Box - Jenkins method, are not always accurate, it can be misleading. To ensure that port managers have knowledge about monthly handling, in this study, for the last three months of 2018 and the period covering 2019, it was made monthly forecasting for the sum of import and export containers in Turkey. It is thought that it will be also easier for the users to calculate their shares with monthly projections and manage the port operations through these calculations. In this study, direct contribution of deterministic trend and seasonality variables to the model and direct effects of these variables in projections are the contribution of the study to science with the idea that it can inspire other studies.