Port Efficiency and the Financial Performance of Greek Public Ports Before and During the Economic Crisis

ABSTRACT

This paper investigates the impact of several financial ratios on the efficiency of ten Greek public ports using the ordinal logistic regression method. The selection of the financial ratios was carried out using the indices related to the three efficiency zones. The results indicate that there are financial ratios that positively affect and others that negatively affect port efficiency. In general, our main findings support the notion that the middle efficiency zone was the most influential of all during the economic crisis. This was confirmed by the ANOVA robustness test. The DEA approach suggests that the ports as a whole were inefficient during the three periods—pre-crisis, in-crisis, and across all sample data periods under investigation. Investment in the ports has been held back during the economic crisis; however, the troika (i.e., the International Monetary Fund, European Commission, and European Central Bank) helped the Greek government to accelerate their economic development during this time. Our results have significant implications for the government. It should, for instance, decide whether to privatise the ten ports partially or fully.

KEYWORDS:

• Financial performance
• port efficiency
• Greek ports
• ordinal logistic regression
• ANOVA
• DEA

Disclosure statement

No potential conflict of interest was reported by the authors.

ACRONYMS

PA: Port Authorities

PAA: Port Authority of Alexandroupolis

PAV: Port Authority of Volos

PAE: Port Authority of Elefsina

PAIG: Port Authority of Igoumenitsa

PAHE: Port Authority of Heraklion

PAK: Port Authority of Kavala

PAC: Port Authority of Corfu

PAL: Port Authority of Lavrion

PAP: Port Authority of Patras

PAR: Port Authority of Rafina

Notes

1. The advantage of our methodology is that it measures in a multi-stage DEA approach the efficiency scores of the ten Greek ports in comparison to the best one. Thus, we conclude that the two-stage DEA model can be equivalent to the multi-stage DEA approach which we have followed in this study.

2. We have performed a multi-stage DEA approach which is equivalent to a two-stage DEA approach. The second stage DEA approach can be based on different parametric or non-parametric regressions according to Simar and Wilson (2011) study.

3. We do not use the DEA to classify the efficiency zones, as we wanted to connect the efficiency zones (High, Medium and Lower) with the bankruptcy threat (Safe, Grey, Risky) based on Altman’s Z-score values, respectively.

4. The DEA multi-stage approach with a CRS model is well-known and we do not mention here the mathematical formula, however it is available upon request.

5. They are discussed in section 4.1.4.