Modelling and Forecasting in the Dry Bulk Shipping Market

This dissertation proposes strategies not only for modelling price behavior in the dry bulk market, but also for modelling relationships between economic and technical variables of dry bulk ships, by using modern time series approaches, Monte Carlo simulation and other economic techniques. The time series modelling techniques, described extensively in Appendix A, primarily consist of the Vector Error Correction model (VECM), the Vector Autoregressive model (VAR), the Autoregressive and Moving Average model (ARMA), the General Autoregressive Conditional Heteroskedasticity model (GARCH), and their extensions. What are new and interesting here of the modelling strategies and their applications lie in several aspects: • Modelling variables of dry bulk ships: the relationships between the economic performance and technical specifications of dry bulk vessels built over the period 1970 to 2008 are investigated in order to examine impacts of technical innovations during the past decades on the economic performance of dry bulk carriers. Furthermore, we construct a simulation-based optimization model to design a series of ships with different deadweight having the optimum main technical specifications and to obtain the optimum logistic plan for each vessel. • The extended time series models: the ARIMA, VAR and VEC models are extended by including exogenous variables to become the ARIMAX, VARX and VECM-X models in order to investigate impacts of exogenous economic variables on charter rates and ship prices. • Dynamic behavior of prices and volatilities on major routes under diverse market conditions: time series models are used to make the prediction of spot charter rates through an in-depth analysis of dynamic behavior between prices for Capesize, Panamax and Handymax vessels on specific major trading routes including the transatlantic, the fronthaul, the transpacific and the backhaul over the period 1990-2010 and over the period 2003-2010, respectively. Time series volatilities of charter rates, newbuilding prices and second-hand ship prices are examined over different periods as well. • Dynamic price behavior between different sub markets: relationships in terms of returns and volatilities at different routes are examined by making use of daily figures between Capesize and Panamax markets, and between Panamax and Handymax markets. The lead-lag relationships of prices between the spot freight and the forward markets, and interrelationships of prices among the freight, the second-hand, the newbuilding and the scrap markets are investigated as well in this research. • The price behavior, the price discovery, and the forecasting performance of FFA prices are examined extensivel