LAUSR

At the country level, macroeconomic volatility tends to correlate with trade openness although the direction of correlation is not stable across samples. Here I consider trade networks as sum of all pairwise trade linkages to emphasize that different linkages contribute differently to the transmission or mitigation of shocks, and show that across the network volatility is inversely related to centrality, a summary measure of strength of the linkages specific to a country. I study a multi-country, multi-sector trade model subject to idiosyncratic productivity and liquidity shocks, and characterize volatility as an explicit function of centrality, diversification and the Herfindahl of the trade network in equilibrium. With sufficient skewness in trade linkages across countries, similar shocks generate different levels of repercussions across the network. The conventional effect of diversification holds true that countries with better diversified portfolio fluctuate less compared. Centrality directly contributes to better aggregation of shocks. Combined effect of these two channels dominates the opposite effect that a more central country is also more exposed to shocks. The model calibrated to the E.U. generates and closely replicates the negative relationship between centrality and volatility. The theoretical model is then extended to capture stochasticity and sparsity in the trade networks.