An empirical investigation on the time-series behavior of the U.S,-China trade deficit

Abstract

This article is concerned with the time-series behavior of the U.S. bilateral trade deficit with China. It performs a unit-root analysis and a cointegration test on U.S. exports to and imports from China. Our results show that the stationarity hypothesis is uniformly rejected for a variety of trade measures over a number of model specifications, which is robust even if endogenously determined structural breaks, both single and multiple, are allowed. Furthermore, it is found that imports and exports are cointegrated with a coefficient substantially larger than one. Our findings put a step toward a better understanding of the issue.