An innovative approach for constructing a shipping index based on dynamic weighted complex networks

• Published in: Physica A 2021-09, Vol.578, p.126101, Article 126101
• Main Authors: Cai, Wenxue, Liang, Fenfen, Wan, Yanchun, Zhong, Huiling, Gu, Yimiao
• Format: Article
• Language: English
• Subjects: Complex network; Dynamic weight; Index construction; Shipping index
• ISSN: 0378-4371; 1873-2119
• DOI: 10.1016/j.physa.2021.126101

The shipping index is a barometer of the shipping market and plays an important role in reflecting the fluctuations of the shipping market. The traditional shipping index compilation method uses fixed weights for calculation, however, which cannot accurately reflect market fluctuations. Complex network theory is often used to study changes in the internal structure of complex systems, and existing studies have also proven the feasibility of analyzing shipping market networks using this method. This paper proposes approaching the shipping market as a complex system, and constructing a multi-stage shipping market network with routes as nodes and inter-route freight rate fluctuations as edges. We construct a weighted routing model based on network characteristics and propose a dynamic shipping index calculation method based on a new weight distribution. For validity verification, we use the Pearl River Shipping Index as an example. The results show that, about 80% of the periods, the original index and the New Pearl River Shipping Index generated by our calculation rise and fall in a consistent manner. Furthermore, when the market fluctuates, the newly constructed shipping index based on the complex network can better identify the changes in the shipping market: the trend changes are more obvious, and more effectively reflect the occurrences of local events in the shipping market. Finally, based on the complex network, the route weights of different periods can be constructed. A changed route weight can affect the rest of the routes through the linkage effect. The wider the scope of influence, the higher the weight of the route, and the more obvious the impact will be on the final shipping index due to freight rate fluctuations on this route. •A novel method for constructing a shipping index based on the node degree of a complex network is proposed. It builds a complex network to reflect internal changes in the shipping market and transforms the changes into route weights and reflects them in the shipping index.•A case study of Pearl River Shipping Index is applied to verify this novel model.•This novel model can help shipping exchanges optimize index compilation and provide more accurate shipping indices. The new index can more sensitively reflect the fluctuations of the shipping market, which can help shipping practitioner grasp the future trend of the shipping market.•To the best of our knowledge, this is the first work to apply complex network theory to c