| Abstract: | We suggest that it may be ‘too easy’ to attribute real exchange rate movements to deviations from the law of one price. We show that it is immaterial whether one uses seemingly traded goods, nontraded goods, or even a single, unimportant consumer good, say beer. The ease of attributing the variation to any such deviations is explained using a model with intermediate goods trade. In the model, the stage of production determines the traded/nontraded distinction. We find empirical substantiation for the model: law of one price deviations lose explanatory power and, defined appropriately in terms of intermediate goods, relative prices matter. |
