The exchange rate, asymmetric shocks and asymmetric distributions ECB Working Paper, No. 1801
Provided in Cooperation with: European Central Bank (ECB)Suggested Citation: Demian, Calin-Vlad; di Mauro, Filippo (2015) : The exchange rate, asymmetric shocks and asymmetric distributions, ECB Working Paper, No. 1801, ISBN 978-92-899-1614
Non-technical SummaryThe link between exchange rates and international trade has been the subject of a broad body of literature without a clear consensus on the size of the effect emerging. Overall, there is a disconnect between estimates using aggregate data and the ones using firm level information: macro estimates tend to be insignificant or very close to zero while micro based research tends to find significant and economically meaningful results. This paper attempts to bridge the gap between these two strands of the literature by using sector level productivity statistics derived from firm level data as additional explanatory variables for exports performance.The main value added of the paper is empirical and derives from the use of a novel data set of productivity statistics for 22 sectors in 10 European Union (EU) countries over 11 years, constructed from firm level data, as part of the CompNet project at the ECB (Lopez-Garcia, di Mauro et al. 2015).Four main results emerge from our estimations:1. The inclusion of the productivity distribution in the export equation drastically affects the average elasticity estimate by reducing unobserved bias. In our benchmark specification, it more than doubles the estimated elasticity from 34% to 77%. Having derived sector specific elasticities allows us also to compute country and EU-wide sector specific elasticities that are reliably identified.2. In line with previous literature, the exchange rate elasticity of exports is lower in sectors where the dispersion of firm productivity is higher.3. Exports appear to react mostly to appreciations rather than depreciations. More specifically, in our preferred specification, the exchange rate elasticity is about 80% in the case of appreciation and statistically not significant in the case of depreciation. Moreover, the size of the coefficient in case of appreciation is much larger than the baseline estimate, indicating that most of the empirically estimated elasticity results from firms' response to appreciation. The negative relationship between a sector's productivity dispersion and export elasticity still holds, but it is significant only in the case of appreciation.4. Exchange rate movements matters more when they are relatively sizeable. To show this, we split the sample in relation to the size of the exchange rate change -i.e. outer 20% and 10-90% range of the distribution, respectively. Results show that in the case of small movements -which in our sample are those between 9% depreciation and 12% appreciation -the exchange rate elasticity is smaller than for extreme movements.