Monday, August 10th
Roman Merga (Rochester), Real Exchange Rate Uncertainty Matters: Trade, Shipping Lags and Default
I study how real exchange rate uncertainty affects international trade. At the aggregate level, I show that there is a negative relation between real exchange rate uncertainty and interna- tional trade. One standard deviation increase in real exchange rate uncertainty is associated with a drop in total trade over GDP of 5%. Then, using Colombian firm-level data, I document 3 firm- level facts consistent with the existence of precautionary margin in international trade. In particular, when real exchange rate uncertainty increases: 1) exporters reduce their export intensity; 2) they are more likely to stop exporting and 3) less likely to start exporting to new markets. Additionally, I find that this behaviour is amplify for those exporter paying higher interest rates and/or facing higher shipping. These results are opposite to the predictions from standard sunk cost models of in- ternational trade. As a result, I show that these types of models will under-estimate the effects that real exchange uncertainty has on international trade. To overcome this issue, I incorporate firm-level debt default and international shipping lags into a standard sunk cost model of international trade. In the model, an increase in the real exchange rate uncertainty raises the probability for exporters to end up in a financially vulnerable situation. To hedge against this risk, exporters respond by increasing mark-ups or quitting the export market. These precautionary practices, consistent with the empirical findings, generate a drop in aggregate exports through both the extensive and the intensive margin of trade. Once the model is calibrated to match Colombian data, it predicts that a one standard deviation increase in the real exchange rate uncertainty generates a drop in total exports between 5% and 10%.