International Trade Finance and Learning Dynamics

We study how trade finance and long-term relationships between exporters and importers facilitate international trade by allowing exporters to learn about demand uncertainty and counterparty risk. Using detailed micro-level Chilean data, we document that new exporters are more likely to use cash-in-advance (CIA) arrangements and gradually switch to providing trade credit as they continue to export. These dynamics affect export growth and are more salient for firms with less exporting experience and selling to riskier destinations. We set up an international trade model in which firms make exporting and trade financing decisions subject to demand and counterparty risks and estimate it using microdata. We then use the model to quantify the relative importance of demand and counterparty risks and investigate how trade finance choices and learning affect the dynamics of exports. Our model implies that the response of aggregate exports and the number of exporters to shocks to aggregate interest rates can overshoot in the short run if long-term relationships and relationship-specific knowledge are destroyed. Building relationships takes time, making the response to these shocks sluggish and persistent. Crucially, these responses depend on the riskiness of trade destinations.