Economic networks in and out of equilibrium

TL;DR

This paper investigates whether real economic networks are in equilibrium or not by comparing their structure to maximum-entropy models, revealing the International Trade Network is near equilibrium while the Dutch Interbank Network shows signs of non-stationary dynamics and crisis precursors.

Contribution

It introduces a method to assess economic network equilibrium status using maximum-entropy ensembles and applies it to real-world trade and banking networks.

Findings

The International Trade Network is nearly in quasi-equilibrium.

The Dutch Interbank Network is out-of-equilibrium with structural changes.

Early-warning signals of the 2008 interbank crisis were identified.

Abstract

Economic and financial networks play a crucial role in various important processes, including economic integration, globalization, and financial crises. Of particular interest is understanding whether the temporal evolution of a real economic network is in a (quasi-)stationary equilibrium, i.e. characterized by smooth structural changes rather than abrupt transitions. Smooth changes in quasi-equilibrium networks can be generally controlled for, and largely predicted, via an appropriate rescaling of structural quantities, while this is generally not possible for abrupt transitions in non-stationary networks. Here we study whether real economic networks are in or out of equilibrium by checking their consistency with quasi-equilibrium maximum-entropy ensembles of graphs. As illustrative examples, we consider the International Trade Network (ITN) and the Dutch Interbank Network (DIN). We show that, despite the globalization process, the ITN is an almost perfect example of quasi-equilibrium network, while the DIN is clearly an out-of-equilibrium network undergoing major structural changes and displaying non-stationary dynamics. Among the out-of-equilibrium properties of the DIN, we find striking early-warning signals of the interbank crisis of 2008.