I study the relative importance of domestic frictions and border price insensitivity for the response of domestic consumer prices (CPI) to exchange rate fluctuations. Using firm and transaction-level data from Chile, I estimate that the presence of domestic frictions --- distribution costs, variable markups and nominal rigidities --- reduce the responsiveness of domestic CPI to exchange rate fluctuations by 60% relatively to an economy that abstract away from it. These frictions are quantitatively more important than the insensitivity of border prices. The presence of domestic frictions also matters for the channels of CPI sensitivity: contrary to prior work, most of the sensitivity arises from the change in the price of imported consumption goods. This channel is more important than the costs arising from imported inputs in the production of domestic goods. The reason is that domestic frictions dampen the price sensitivity of domestically produced goods relatively more. Furthermore, the sensitivity varies across products because of the heterogeneity in domestic frictions, import exposure, and consumption shares. The heterogeneity matters for the overall (in)sensitivity as domestic products with higher import exposure face larger frictions and have lower consumption shares. Ignoring the heterogeneity identifies the wrong products from which most of the sensitivity arises, with implications for monetary policy targeting in open economy and redistribution dynamics.

Lack of detailed data on the characteristics and quality of imported goods poses a challenge for measuring consumption gains from rising imports. To tackle this problem, we propose a methodology to infer unobserved quality change using only data on prices and market shares in a differentiated product market. The method identifies a demand system in which product substitutability varies across products based on quantity and quality. We validate the method using data from the US auto market where information on product characteristics and price instruments are available. Without using these additional sources of information, our strategy estimates price elasticities and quality changes in line with the predictions of the standard estimations of BLP demand. We apply this strategy to the US customs data (1989-2006), and find that quality improvements have lowered the price of US imports relative to the CPI by 17%. For comparison, unit values have fallen by around 11% relative to the CPI and increasing variety has contributed an additional 4%. Using a demand sys- tem that ignores the heterogeneity in product substitutability leads to a substantial overestimation of the extent of quality improvements.

The huge trading volumes in the foreign exchange rate markets are highly con- centrated among few financial players. Differently from standard models of exchange rate determination, we reject the assumption of perfectly competitive financial markets and assume traders to be imperfect competitors. We develop an international portfolio choice model with noise shocks and traders’ heterogeneity in market power. Large non-competitive traders internalize the impact of their portfolio decisions on the determination of prices. We show that the presence of strategic investors leads to the amplification (dampening) of the impact of non-fundamentals (fundamental) trade on the exchange rate, reducing its informativeness. The implications of the models are that the presence of strategic investors: i) increases exchange rate volatility; ii) reduces the role of fundamentals in explaining exchange rate movements (exchange rate discon- nect); and iii) increases foreign asset excess return and makes it more predictable. Our theoretical predictions are empirically confirmed using trading volume concentration data from the NY Fed FXC Reports for 18 currencies from 2005 to 2019. Welfare analysis suggests that the consolidation in the financial sector in the last three decades increased investors’ welfare by 30%.