Exchange rate volatility, sectoral trade, and the aggregation bias

This paper proposes a sectoral theoretical model in an imperfect competition framework, with country-specific and industry-specific original variables, notably factor productivity, scale economics, or product differentiation. It is then empirically estimated in a panel data model, at a sectoral and geographical disaggregation level, to test the impact of exchange rate volatility on G-7 countries' exports. Economies of scale are estimated from a nonlinear translog production system. Two exchange rate volatility measurements have been used: the moving sample standard deviation and the GARCH approach. The main finding shows that the impact of exchange rate volatility on exports varies considerably, depending on the industry covered and the export destination markets. As a consequence, there is both a sectoral and geographical aggregation bias when estimating the effects of exchange rate variations.