Global Firms in Large Devaluations.

I investigate the consequences of firms' joint import and export decisions in the context of large devaluations. I provide empirical evidence that large devaluations are characterized by an increase in the aggregate share of imported inputs in total input spending and by reallocation of resources toward import-intensive firms, contrary to what standard quantitative trade models predict. These facts are explained by the expansion of exporters, which are intense importers. I develop a model where firms globally decide their import and export strategies and discipline it to match salient features of the Mexican micro data. After a devaluation, the model reproduces the pattern of low aggregate substitution and firm reallocation observed in the data. Compared with a benchmark without global firms, the model predicts higher growth of total exports and imports and a smaller reduction in the trade deficit. [ABSTRACT FROM AUTHOR]