Skewness of local logarithmic exports

Abstract

The distributions of trade values and relationships among countries and product categories reflect how countries select their trade partners and design export portfolios. Here we consider the exporterimporter network and the exporter-product network with directed links weighted by the logarithm of the corresponding export values each year from 1962 to 2018, and study how the weights of the outgoing links from each country are distributed. Such local logarithmic export distributions by destinations and products are found to follow approximately the Gaussian distribution across exporters and time, implying random assignment of export values on logarithmic scale. However, a non-zero skewness is identified, changing from positive to negative as exporters have more partner importers and more product categories in their portfolios. Seeking the origin, we analyze how local exports depend on the out-degree of exporter and the in-degrees of destinations/products and formulate their quantitative and measurable relation incorporating randomness, which uncovers the fundamental nature of the export strategies of individual countries.