Essays on exchange rates in developing countries

This dissertation advances the understanding of exchange rates in developing countries. In the first two chapters, I study how real exchange rates move in the long run, while in the third chapter, I examine the forecasting abilities of exchange rates of commodity exporters from developing countries.;The conventional wisdom is to rely on the Balassa-Samuelson hypothesis (BSH) to explain how real exchange rates move in growing economies. However, despite the growth takeoffs since the early 1980s, the real exchange rate in developing countries recorded limited appreciation or none at all over the last 30 years, in contrast to the BSH prediction. Chapter 1 contributes new empirical evidence to show that advanced countries follow conventional wisdom but developing countries move in the opposite direction, which I call the "anti-Balassa" effect. In fact, I find that the relationship is negative or non-existent in cross-country data. At the same time, the evidence from panel data reveals that developing countries follow a non-linear relationship. Further, I find that the level of development can systematically impact the co-movements between the real exchange rate and income. ii.;To explain why the BSH is not satisfied in low-income developing countries, a simple model is developed in Chapter 2, one that dampens the BSH by the presence of "labor surplus" (a large supply of low-wage workers in the traditional sector). The reallocation of labor surplus to the productive tradable sector leads to rapid growth but contains the appreciation of the real exchange rate. A testable implication of the model is that the relationship between real appreciation and growth depends on the size of the labor surplus. I present empirical evidence consistent with the model. I find that the BSH is suppressed in countries that have a sizeable pool of labor surplus from which to draw.;The theoretical and empirical results in Chapters 1 and 2 suggest that the structure of the economy could be central in determining the relationship between the real exchange rate and income; thus, policymakers should pay close attention to this factor when evaluating the real exchange rate.;Finally, in Chapter 3, I examine the nominal exchange rates of commodity exporters from developing countries in Latin America, Asia, and Africa. The new results contribute further evidence that the predictive relationship between commodity currencies and future commodity prices in Chen, Rogoff and Rossi (2010) is indeed robust. Considering in-sample and out-of-sample tests, I find that commodity currencies from developing countries have forecasting power for global commodity prices. This is especially true for Latin America and Asia.