MODELING THE DRY BULK SHIPPING MARKET USING MACROECONOMIC FACTORS IN ADDITION TO SHIPPING MARKET PARAMETERS VIA ARTIFICIAL NEURAL NETWORKS

Dimitrios Lyridis, Nikolaos Manos and Panayotis Zacharioudakis

Articles, 2014, vol. 41, issue 2

Abstract: This paper attempts to describe the shipping market of bulk carriers introducing for the first time macroeconomic factors in addition to other microeconomic and shipping market parameters. This is done by building explanatory models using Artificial Neural Networks (ANNs). In effect, the model tries to forecast the Baltic Exchange Dry Index (BDI) using real data for a twenty-year period for a wide range of macroeconomic factors (nineteen) and maritime indexes (four), for which data are available on a daily, weekly, monthly or 3-monthly basis. This depends mainly on the agency or bureau that is responsible for the compilation of every distinct macroeconomic figure. However, in case the figure under examination is tradable in a specific market, the data resolution may also depend on the characteristics of this market. The effectiveness of the model depends also very much on whether or not neural networks are able to predict accurately the evolution of BDI. For this reason, a period around and after the financial crisis of 2008 was selected. However, the data series spans 20 years from 1991 to 2011 and in this way captures market movements in the more extreme up and downturns. During this period the BDI presented a steep decline of more than 90% and dropped from the level of 9,912 units (May 2008) to 558 units (Dec 2008). Therefore, this paper investigates whether it was possible to predict in June 2008, what would have happened 2,4,6, 8 and 10 months forward. The results reached, after the implementation of the specific methodology described in this paper are satisfactory: models built using ANNs and taking minto account macroeconomic factors in addition to other parameters (shipping and microeconomic) are able to predict the evolution of the Baltic Dry Index much more accurately than existing methods. The performance of the model is very satisfactory for the prediction period, malthough there are really no robust models for accurate predictions for longer periods.

Date: 2014
References: Add references at CitEc
Citations: View citations in EconPapers (1)

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:jte:journl:2014:2:41:4

Access Statistics for this article

More articles in Articles from International Journal of Transport Economics
Bibliographic data for series maintained by Alessio Tei ().