heal.abstract |
In this master thesis, a model of artificial neural networks was developed, aiming to the prediction of future time charter rates based on certain variables that define the shipping market. The dry bulk shipping market is characterized by great complexity and significant fluctuations, creating the need to find a better way to predict its trajectory. By observing history, one can easily understand that many events have caused significant changes in the structure, operation, and prevailing prices in both the market and the ships themselves, leading to its current form, which follows the stages of the shipping cycle. There are now many different markets in which ships operate, such as the Liner and Charter markets, various types of contracts tailored to the needs of voyages, cargoes, charterers, and shipowners, as well as many factors influencing shipowners' decisions. The most critical factors in price trends are the supply of shipping capacity by shipowners and the demand from charterers. For the model's development, the Capesize bulk carriers' market was examined, considering both endogenous and exogenous factors affecting the shipping market's trajectory. Due to the complexity of exogenous factors and the inability to find and incorporate their data into a model, only some endogenous factors were used. The endogenous factors on which the model is based are time charter rates, newbuilding prices, OPEX, the global fleet of this specific type of ship, the percentage of new ships entering the global fleet, the percentage and number of laid-up ships, and the percentage and number of ships destined for scrap. From these variables, the market's shipping capacity supply is ultimately calculated, and in combination with market demand and operating expenses, it leads to the final sought result: the daily time charter rate. Finally, the research results are presented and examined, and final conclusions and comments on the predictions are drawn |
en |