The exchange rate exposure puzzle has remained robust to empirical scrutiny despite varied and ever more sophisticated empirical approaches. The difficulty in finding empirical evidence of a link between exchange rate changes and firm returns is more evident at shorter horizons with the prevailing view in the literature being that the puzzle abates when longer horizons are considered.The findings of this paper suggests the long horizon evidence is illusory. Specifically, the application of inference that has good finite sample properties in a long horizon regression setting overturns evidence of long horizon exposure using conventional estimation methods.

ABSTRACT

No abstract