Collaborative Research: Tradability of Goods, Exchange Rate Movements and Capital Flows

University Of Southern California, Los Angeles CA

Investigators

Abstract

In post-war data real exchange rates have been highly volatile and have deviated widely and systematically from their purchasing power parity values. Since the breakdown of the Bretton Woods system, real and nominal exchange rates are highly positively correlated. This fact is widely cited as (indirect) evidence of nominal price rigidities, such that real exchange rates are driven entirely by shocks to the nominal exchange rate, while relative national price levels are relatively smooth. Partly as a result of this evidence, the predominant theoretical frameworks for analyzing exchange rates are characterized by nominal rigidities. Such frameworks have been successful in replicating many features of international business cycle and exchange rate facts. The investigator and Caroline Betts have obtained preliminary results that are difficult to reconcile with the idea that real exchange rates are driven solely by shocks to the nominal exchange rate. They have shown that real exchange rates are also very highly correlated with the relative prices of nontraded to traded goods, however, suggesting that real factors are key determinants of real exchange rates. Furthermore, they show that a multicountry, multisector real business cycle model is capable of generating relative price behavior at both the sectoral and aggregate level that matches Mexico-US data. They develop a class of business cycle models that can address these puzzling empirical features of exchange rate behavior. The research would extend this work through the introduction of a role for national monies. A key goal is to reconcile the evidence of high correlations between real and nominal exchange rates with that of high correlations of real exchange rates and the relative price of nontraded to traded goods across countries. Another goal is to replicate other key features of international business cycle data. Much this research would be joint work with Caroline M. Betts of the University of Southern California. The research would extend the empirical analysis of Betts and Kehoe on the relationship between the Mexico-US real exchange rate and the relative prices of traded to nontraded goods. The analysis to date has used annual gross output deflators to measure sectoral relative prices, and we propose to investigate other sources and frequencies of price data. It would also extend the analysis to the other countries and to explore the empirical behavior of nominal exchange rates and nominal prices for this group of countries. The proposed, multicountry, multisector, business cycle models depart from traditional theories of real exchange rate determination in that the outputs of different sectors, rather than being either tradable or nontradable, have differing degrees of tradability. The degree of tradability is determined by the sort of trade costs emphasized by Obstfeld and Rogoff . The models incorporate considerable sectoral detail, and some allow roles for national monies. Real and monetary shocks drive international fluctuations. Preliminary results for a model without money are encouraging: One does not need nominal rigidities to account for the size or even most of the persistence of real exchange rate fluctuations. Country-specific demand shocks calibrated to match the variability and persistence of aggregate output are sufficient to capture international relative price movements by sector. Further, the model can account for the high volatility of the real exchange rate compared to the relative price of nontraded to traded goods across countries, despite the absence of monetary features in the model. Despite these successes, nonmonetary versions of the model are silent on the relationship between real and nominal exchange rates and the role of nominal shocks. Proposed research would introduce the simplest possible quantity equation model in which the real side of the economy is unaffected by money. In addition, we propose a cash-in-advance variant in which money can give rise to a distortion in the leisure-labor choice. Key features of the data that we would want any monetary model to replicate is the high correlation of real and nominal exchange rates and the behavior of relative prices shown in our empirical work.