A Re-Examination of the Exchange Rate Disconnect Puzzle: Evidence from Japanese Firm Level Data

Abstract

The empirical literature examining aggregate data has generally found small or insignificant effects of rate fluctuations on export volumes. This lack of association between real quantities, such as export volumes and the rate is the so-called exchange rate disconnect puzzle. Using firm level data, however, the relationship between export volumes and rates turns to significantly negative. This paper attempts to reconcile these aggregate and firm level findings, using firm level data from Japan. We estimate a simple microeconomic model of exports to show that an appreciation of the rate reduces export volumes at the firm level. After consistent aggregation, the relationship still remains significant at aggregate levels. However, we show that the omission of some key productivity variables, or ignoring the distributions of heterogeneous firm level characteristics biases the elasticity of exports to rates toward zero.