Econometric estimation of second-hand shipping markets using panel data analysis

Panel data analysis is becoming increasingly popular in shipping markets since it enables the employment of a wider source of variation which allows a more efficient estimation of a model's parameters. This study applies an econometric analysis on a balanced panel data set of tankers' second-hand prices for five different vessel types as cross-section identifiers. Empirical analysis investigates the existence of second-hand prices' differentiation according to the vessel size using monthly observations for over a forty years time period. The key question concerns relationships among second-hand prices, spot rates and newbuilding prices and their dependence on whether vessel sizes experience low or higher rates of interdependence. Analysis focuses on the aspect of heterogeneity among variables, which is due to the effects of unobserved variables. The models estimate fixed and random effects and examine both cross section and time effects. Unit root and cointegration tests are performed in order to check for stationarity and for the existence of any long-run equilibrium relationships among variables. Also, Hausman test is adopted to test the existence of correlated random effects. Empirical results lead to conclusions and implications regarding the use of spot rates and newbuilding prices as intermediate means for the prediction of second-hand prices.