The Trade Imbalance Network and Currency Returns

Centre for Endowment Asset Management (CEAM)
Centre for Endowment Asset Management (CEAM)

Overview

This project examines how global trade relationships between countries help to explain why some currencies earn higher returns than others. Traditional models of currency premia often look at each country in isolation, or at simple measures such as interest rate differentials. This paper takes a different approach. It views the world economy as a network of trade imbalances: some countries consistently run trade surpluses, others run deficits, and these imbalances are linked through complex global value chains. 

Analysing a graph on screen.

Authors

Ai Jun Hou

Stockholm University

Xiaoxia Ye

Nottingham University Business School

Lucio Sarno

Cambridge Judge Business School, University of Cambridge

About the research

Financial intermediaries, such as global banks and asset managers, must absorb these imbalances on their balance sheets. When a country is highly central in the global trade imbalance network, intermediaries are more exposed to that country’s risks and have less capacity to take on additional positions. The research argues that this limited risk bearing capacity feeds directly into the risk premia investors require to hold different currencies. 

What the researchers do: 

  • They extend the theoretical framework of Gabaix and Maggiori (2015) to a multi-country setting with imperfect financial markets. 
  • Using bilateral trade data for up to 41 countries (1995–2021), they build a global trade imbalance network and compute a new measure called the CentralityBased Characteristic (CBC) for each currency. CBC summarises how central a country is in absorbing worldwide trade imbalances, adjusted for the variance–covariance structure of currency returns. 
  • They then sort currencies into portfolios based on CBC and test whether CBC predicts future currency excess returns, controlling for well known factors such as carry, trade imbalances, and global risk measures. 

Findings

  • Currencies of countries with high CBC (more central, bearing more of the world’s trade imbalances) earn significantly higher average excess returns and Sharpe ratios than currencies with low CBC. 
  • The return spread from a simple long–short strategy that buys highCBC currencies and sells lowCBC currencies is large, robust outofsample, and cannot be explained away by standard currency factors. 
  • Decomposing the mechanism, the authors show that most of the cross-sectional variation in currency premia is driven by a “neighbourhood” effect: how exposed a country is to the imbalances of its trading partners, not just its own net trade position. 
  • CBC is closely related to independent proxies for intermediaries’ riskbearing capacity (such as FX volatility indices and funding spreads), supporting the interpretation that financial constraints and trade networks jointly shape currency risk premia. 

The project provides a new, network based view of currency markets. It shows that to understand and forecast currency returns, it is not enough to look at individual countries’ trade balances or interest rates. Instead, we must consider how countries are embedded in the global trade imbalance network and how this interacts with the limited balance sheet capacity of international financiers. 

Top