Price Rigidities and Relative PPP

We measure the proportion of real exchange rate movements accounted for by cross-country movements in relative reset prices (prices that changed since the previous period) using CPI microdata for five countries. Relative reset prices account for almost the totality of the real exchange rate movements. This is at odds with the predictions of most workhorse sticky price models used to generate volatile and persistent real exchange rates. In these models relative reset prices are sluggish because relative wages are either sluggish or mean revert quickly. We show that models where movements in relative wages are persistent and track the nominal exchange rate do replicate the empirical properties of both the real exchange rate and of relative reset prices.