The Effect of COVID-19 on minor dry bulk shipping: A Bayesian time series and a neural networks approach. Issue 4 (2nd October 2021)

• Record Type: Journal Article
• Authors: Molaris, V. A.; Triantafyllopoulos, K.; Papadakis, G.; Economou, P.; Bersimis, S.
• Is Part Of: Communication in statistics. Volume 7:Issue 4(2021)
• Page Start: 624
• Page End: 638
• Publication Date: 2021-10-02
• Subjects: Time series analysis -- neural networks -- dry bulk shipping; Mathematical statistics -- Data processing -- Periodicals; 519.505
• DOI: 10.1080/23737484.2021.1979434
• Languages: English
• ISSNs: 2373-7484

Abstract

ABSTRACT: Supply chain is a crucial part of the world economy and everyday life. During the outbreak of COVID-19 in 2020, governments across the globe have tried to maintain the supply chain operations in most of the essential goods and services. Dry bulk shipping, a sector characterized by trading versatility and a multitude of cargoes carried, has an important role to play towards that aim. Port restrictions, placed to limit the spread of COVID-19, have created the impression that the market of dry bulk shipping has, to some degree, been affected and become less predictable, as many other transporting services. In this article, this belief is investigated using both state space modeling and neural networks. In particular, using these methods, one-month ahead predictions (during 2020) of the overall freight cost are obtained for three types of vessels (representing the majority of the minor bulks industry). A good forecast performance is observed and this validates the models and the estimation methods proposed. Our results suggest that there is no significant effect of COVID-19 in these types of vessels and their operations. This can provide useful information to shipping managers and policy makers.