The exchange rate exposure puzzle: The long and the short of it

•Exchange rate exposure tested across different time horizons (1 month to 5 years).•The estimation method used is appropriate when using overlapping data.•Literature: Evidence of transaction (short horizon) exposure is difficult to find.•Literature: Evidence of economic (long horizon) exposure is more easily found.•We do not find evidence of economic exposure. The exposure puzzle is more puzzling.

The exchange rate exposure puzzle has remained robust to empirical scrutiny however evidence suggests the puzzle abates when longer horizons are considered. This paper applies inference that is appropriate in a long horizon setting and finds this evidence is illusory.