Transfer function models for forecasting maritime passenger traffic in Greece under an economic crisis environment

ABSTRACT

Maritime transportation constitutes a dynamic Greek economic sector. Greece accounted for 25% of the Maritime Passenger Traffic (MPT) in the European Union in 2007. However, due to the economic recession since 2009, maritime passenger transportation has been facing increasing fuel prices along with declining demand. In this paper, we develop transfer function models, including macroeconomic indicators, for forecasting the coastal and the Adriatic MPT in Greek ports. The majority of the MPT models include the Gross Domestic Product (GDP) as an explanatory variable, reflecting the financial impact on maritime flows. Although the employment and the economic crisis dummy are also used as explanatory variables, the unemployment and the oil price do not affect MPT. The coastal traffic models are generally more consistent when the seasonality trend is eliminated, while the international traffic models perform better when the inherent seasonality is considered. This first effort contributes toward filling the gap of MPT forecasting models for supporting policy-making on maritime transportation.

KEYWORDS:

• Time series
• transfer function models
• maritime passenger traffic
• economic crisis
• Greece

Acknowledgments

The authors are grateful to the Hellenic Statistical Authority (ELSTAT), Greece, for providing all necessary data for this research work. This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors. On behalf of both authors, the corresponding author states that there is no conflict of interest.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1. The explanatory (or independent) variables are the inputs manipulated/changed to test their effect on the dependent variable. The dependent variable (output) is observed to measure the effect of the independent variables on them.

2. The backshift operator operates on an element of a time series to produce a previous element. For example, for a given time series $\mathrm{X}=\left\{{\mathrm{X}}_1,{\mathrm{X}}_2,\dots \right\}$ and a certain n order, then ${\mathrm{B}}^{\mathrm{n}}\cdot {\mathrm{X}}_{\mathrm{t}}={\mathrm{X}}_{\mathrm{t}-\mathrm{n}}$, which refers to the value of the time series n time steps before the time t.

3. It should be also mentioned that both the seasonal and deseasonalized models of MS traffic show the lowest fitting capacity, probably due to the fact that they refer to a rather low demand line.