Seasonal unit root tests for the monthly container transshipment of the port of Hamburg

This paper investigates the seasonal behaviour of monthly container transshipment data of the port of Hamburg. The test procedures of Franses and Beaulieu-Miron are used to examine the presence of multiple unit roots in the monthly seasonal frequencies. This is followed by the Canova-Hansen procedure for testing the stability of the seasonal patterns. Evidence suggests that these monthly transshipment data are non-stationary at frequency zero and have no seasonal unit roots. The analysis shows that the process is in the long run integrated of order one and that the seasonal variations can be modelled by dummy variables. Using the deterministic seasonality found here, further analysis and forecasting of container throughput of the port of Hamburg can be improved for market participants like containership lines.