This month, the dollar reached its lowest level for 14 months on a trade-weighted basis. This low value will raise the cost of imports and consumer prices in the US and should therefore reduce US imports. In parallel, the euro has reached very high levels against the US dollar and other currencies, which should harm European exporters. By how much?
Existing evidence suggests that the effect should be small. Even large movements in nominal and real exchange rates have little impact on aggregate variables such as import prices, consumer prices, and the volumes of imports and exports. The lack of sensitivity of prices to exchange rate movements has been well documented; prices of tradable goods respond incompletely to variations in exchange rates – substantially less than proportional to the exchange rate change. One possible explanation is that prices are rigid in the currency of the export market. However, Campa and Goldberg (2005) show that the incomplete pass-through of exchange rate changes into import prices is far from being a short-term phenomenon, as it remains after one year. This suggests that price rigidities cannot fully explain this phenomenon. Moreover, Gopinath and Rigobon (2008) have recently shown, using product-level data, that even conditioning on a price change, trade weighted exchange rate pass-through into US import prices is low, at 22%. There is also evidence that pass-through has declined with time.
If pass-through of exchange rate movements to import prices is small, it is not surprising that exchange rate movements have a small effect on trade volumes. The typical macro elasticities found by the empirical literature are indeed around unity or just above.
A firm-level explanation
What is behind this lack of price and quantity response to exchange rates? In a recent paper (Berman, Martin and Mayer, 2009), we try to answer this question from the angle of exporters. Most of the literature on pass-through has concentrated on import prices and consumer prices and another strand of papers have looked at the impact of exchange rate movements on aggregate imports or exports.
We go into much more detail by gathering information on firm-level exports by French firms to be confronted with predictions of the “new new” trade models, where differences across firms drive macro-economic effects. Analysing the exporters’ response to exchange rate changes allows us to put emphasis on the heterogeneity of the optimal response of exporters to exchange rate movements.
We show theoretically and empirically that high- and low-performance firms react very differently to exchange rate movements. Performance can be interpreted in terms of productivity or in terms of quality of the goods produced. Following a depreciation, high-performance firms optimally raise their mark-up rather than the volume they export, while low-performance firms choose the opposite strategy.
Another way to state this result is that high-performance firms absorb exchange rate movements in their mark-ups but low performance firms do not. The reason is that, due to local distribution costs, the demand elasticity perceived by high performance firms is lower than the one perceived by low performance ones.
Understanding aggregate pass-through
This heterogeneity is a novel finding, but is also interesting because it has important implications for the aggregate effect of exchange rate movements. Pricing-to-market is optimal for high-performance firms, precisely those firms that are more likely to be exporters. In our model, following the spirit of Melitz (2003), fixed export costs generate a selection mechanism through which only the best performers are able to export. Also, heterogeneity in productivity implies that a small portion of high-performance firms sell a very large share of aggregate exports. Hence, exporters, and even more so big exporters, are, by this selection effect, firms that optimally choose to absorb exchange rate movements in their mark-ups.
Exchange rate depreciation also triggers the entry of new firms into the export market, but these are less productive and smaller than existing ones, so the “extensive margin” effect is small at the aggregated level. We show that our model, with sufficient heterogeneity in productivity, can indeed reproduce the observed low level of the elasticity of the intensive and extensive margins of trade to exchange rate movements.
We test these predictions on a very rich French firm-level dataset. We collected information on firm-level, destination-specific export values and volumes from the French Customs and other information on firm performance at annual frequency on the period 1995-2005 so that we can exploit variation across both years and destinations. A big advantage of our dataset is that we have information that can proxy for the free-on-board price at the producer/destination level. We can therefore study the impact of exchange rate depreciations on the pricing strategy of exporters. Our paper is therefore complementary to existing studies on pricing-to-market and pass-through using information that proxy the pricing strategies of exporters through import prices (which contain transport costs) or consumer prices.
We find that firms with performance (measured by TFP, labour productivity, export size, number of destinations) above the median reacts to a 10% depreciation by increasing their (destination-specific) export price in euro by around 2%. In contrast, those firms below median performance do not change export prices in reaction to a change in exchange rate. Hence, only high performers partially price-to-market and partially absorb exchange rate movements in their mark-ups. On export volumes (again destination-specific), the reverse is true; the export volumes of the best performers do not react to exchange rate movements, whereas poor performers increase their export volumes by around 6% in response to a 10% depreciation. Distribution costs indeed play an important role. Following a depreciation, French exporters selling in sectors and countries with high distribution costs choose to increase their mark-up rather than their export volumes. Distribution costs in the destination market therefore change the pricing strategy of exporters towards more pricing-to-market. Finally, the selection effect emphasised above is also observed in the data. In sectors where exports are more concentrated amongst very productive firms – those absorbing exchange rate movements in their mark-ups – aggregate exports are least responsive to exchange rate changes.
Our research shows that exporting, because it requires high performance, goes hand-in-hand with absorbing exchange rate movements into prices. That exchange rates have little impact on import prices and trade volumes is therefore a natural consequence of this selection effect.
Berman Nicolas, Philippe Martin and Thierry Mayer (2009) "How do different exporters react to exchange rate changes? Theory, empirics and aggregate implications", CEPR Discussion Paper 7493.
Campa Jose and Linda Goldberg, 2005, “Exchange Rate Pass Through into Import Prices”, Review of Economics and Statistics, November 2005.
Gopinath, Gita, and Roberto Rigobon, 2008, "Sticky Borders," Quarterly Journal of Economics, 123(2), 531-575.
Melitz, Marc J. 2003. "The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity." Econometrica, 71(6): 1695---1725.