Description

This paper extends earlier research by adding SWIFT data on documentary collections to the short-term forecast of international trade. While SWIFT documentary collections accounted for just over one percent of world trade financing in 2020, they have strong explanatory power to forecast world trade and national trade in selected economies. The informational content from documentary collections helps improve the forecast of world trade, while a horse race with machine learning algorithms shows significant non-linearities between trade and its determinants during the Covid-19 pandemic.

More publications in this series: Working Papers

More publications by: Narek Ghazaryan ; Alexei Goumilevski ; Joannes Mongardini ; Aneta Radzikowski