DP3779 Trade, Finance, Specialization and Synchronization
I investigate the determinants of business cycles synchronization across regions. I use both international and intranational data to evaluate the linkages between trade in goods, trade in financial assets, specialization and business cycles synchronization in the context of a system of simultaneous equations. In all specifications, the results are as follows. (i) Simultaneity is important, as both trade and financial openness have a direct and an indirect effect on cycles of synchronization. (ii) A variety of alternative measures of financial integration suggest that economic regions with strong financial links are significantly more synchronized, even though they are also more specialized. (iii) Specialization patterns have a sizeable effect on business cycles, above and beyond their reflection of intra-industry trade and of openness to goods and assets trade. (iv) The simultaneous approach makes it possible to disentangle the impacts of both inter- and intra-industry trades. The estimated role of trade is in line with existing models once intra-industry trade is controlled for. Furthermore, trade-induced specialization has virtually no effect on cycles synchronization. The results obtain in a variety of datasets, measurement strategies and specifications. They relate to a recent strand of International Business Cycles models with incomplete markets and transport costs, and on the empirical side, point to an important omission in the list of criteria defining an Optimal Currency Area, namely specialization patterns.