ABSTRACT
Most global trade statistics in the public domain refer to official customs data, which are not generally available on a micro (individual cargo) level. With the increasing availability and completeness of ship positioning data from the global Automated Identification System (AIS), it is possible to derive more timely and detailed trade statistics for homogeneous commodity groups. The objective of this article is twofold: (1) to compare the accuracy of AIS-derived trade statistics to official customs data in the crude oil market and (2) to add a breakdown of trade by vessel size over time. We find that while AIS-derived data for seaborne crude exports show good alignment with official export numbers in aggregate, there are substantial temporal and geographical differences across countries and time due to the use of pipelines and transshipment in parts of the supply chain. We highlight the challenges in properly structuring and aggregating micro-level cargo data. Our findings are important for the proper derivation of shipping demand from trade data.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1. The only exception with readily available micro-level data on individual shipments is the United States, where information contained in the Bill of Lading is provided to third-party ‘publishers’ that disseminate this into the public domain.
2. Consider the extreme example of Russian crude exports to Asia which may originate in the Baltic or Far East ports, a difference of thousands of nautical miles.
3. Blonigen and Wilson (Citation2013) provide a useful review of the evolution of trade modelling.