This paper decomposes the real exchange rate (RER) into its components, extending the approach by Engel (JPE, 1999) and Betts and Kehoe (JME, 2006). Engel finds that the relative prices of non-traded goods account for almost none of the RER variation. Betts and Kehoe show that, using a different price index the relative prices of non-tradable goods do indeed matter. They then argue that non-tradable goods need to be introduced into theoretical models to account for RER co-movements.
The decomposition in this paper yields two additional terms: the first one is the relative distribution cost in the two countries, the second one is the relative profit margin in the two countries. I show that the relative profit margin accounts for a large proportion of the RER movement.
The paper then goes on to show that this observation, that the RER is positively correlated with relative profit margins, is only consistent with theoretical models employing local currency pricing (LCP), rather than producer currency pricing (PCP). I simulate a two versions of the model introduced by Corsetti at al (Handbook of Monetary Economics, 2010), one using LCP and the other using PCP. I show that only the LCP model implies a positive correlation between the RER and relative profit margins. This paper therefore adds empirical evidence to the open economy literature and to the pass-through literature.