Three Essays on Exchange Rate Models and Their Applications

Chapter 1: This chapter attempts to identify the determinant of exchange rate pass-through into producer and destination prices using highly disaggregated firm-level data of importers and exporters of the Egyptian economy from 2009 to 2013. The main findings assert the hypothesis of complete exchange rate pass-through into destination prices at the lowest level of significance for the average exporting firm that is also importing. Furthermore, the firm with the highest import intensity has a lower percentage of pass-through into destination prices, indicating that the marginal cost channel plays a more significant role in determining the speed of the exchange rate pass-through compared to the markup channel. Moreover, differentiating between periods of depreciation and appreciation does not alter the main findings of the baseline model. Chapter 2: This chapter analyzes the cointegration relationship between exchange rate and macroeconomic fundamentals within a panel data framework. The main contribution is that it uses a broader sample of countries that adopted a flexible exchange rate regime throughout 1975–2016. Moreover, it examines the cointegration relationship between exchange rates and monetary fundamentals using Groen and Kleibergen (2003) method. The results indicate that there is at least one cointegration relationship between the monetary fundamentals and exchange rate. However, we rejected the null of a common cointegration relationship between those same variables among all the countries included in one panel. Moreover, the estimation of the coefficients of the economic activity and the monetary variables coincided with the prediction of the theory of monetary approach of exchange rate determination. Nevertheless, the results do not support that both the exchange rate and the money supply move with the same percentage in the long run. Chapter 3: This chapter estimates a GVAR exchange rate model augmented with a dominant unit. The estimated model uses a mix of fundamental variables suggested in the theoretical literature on monetary and asset approaches to determine exchange rates. In addition, the paper evaluates the out-of-sample forecasts from this model against benchmark models such as the random walk (RW) and the random walk with drift (RWD). The estimated model could not beat the RW or the RWD results when it forecast the nominal exchange rate for individual countries in the global sample or the panel sample. However, it did beat those benchmark models when it forecast the depreciation rate of the nominal exchange rate for both the individual sample and the panel sample. Furthermore, the results improved in favor of the suggested model when a time-varying weight matrix was used for the foreign variables instead of the fixed weight matrix