How does monetary policy affect the exchange rate? Empirical evidence from G-7 countries

Theories on the real exchange rate behavior can be divided into four different categories: (i) models with fundamentals are a variation of simple monetary models, including models with flexible nominal prices and models with sluggish nominal price adjustment and overshooting; (ii) equilibrium models emphasizing real disturbances to tastes and technologies; (iii) portfolio-balance models, which assume that domestic and foreign securities are not perfect substitutes and emphasize different domestic and foreign assets, including current account balances; and (iv) speculative bubbles, which focus on unpredictable components of exchange rates.; My original motivation for carrying out empirical studies on the effects of monetary policy on exchange rate was to investigate the contradictory findings of different models. The first chapter presents the theoretical background and detailed empirical studies.; The next two chapters investigate what is the transmission channel of monetary policy.; The second chapter pays particular attention to the puzzle first suggested by Meese and Rogoff (1983), who argued that a random walk model beats all the competing models based on root mean squared error (RMSE), and mean absolute error (MAE) criteria. Since this study, many efforts have been made to overturn their results using different econometric techniques. However, less attention has been paid to the sampling error and misspecification problem, which is the problem of omitted variables. To compensate for the weaknesses of the current literature, the second chapter provides empirical evidence that the Meese and Rogoff puzzle is attributable to both sampling error and misspecification problem. Extending the data from the post-Bretton Woods era to 2001:12 overturns Meese and Rogoff's results, implying that a random walk model fails to outperform all the competing models for G-7 currencies based on RMSE and MAE criteria. In addition, testing a structural model after including the previously omitted variables shows that it not only beats a random walk, but also outperforms all the competing models, including interest rate parity, overshooting, and portfolio-balance models for such major currencies as the Deutsche mark, the Japanese yen, the Italian lira, and the pound sterling.; The third chapter addresses two key questions: what are the main determinants of exchange rates, and through what transmission channel does monetary policy affect the exchange rate? Relying on different assumptions, including fiscal policy variable and paying particular attention to stock market channel I show that monetary policy affects the exchange rate through the stock market channel. I show that the oil price does not matter in driving the exchange rate behavior among G-7 countries. However, the results suggest that stock returns, especially dividend yields, play a crucial role in determining exchange rate behavior. Though, adding government expenditures, as a fiscal policy variable to the VAR specification suggests that fiscal policy has a greater influence than monetary policy on exchange rates in G-7 countries. In fact, monetary policy affects exchange rate behavior through the pass-through effects on stock market, especially through dividend yields, rather than through standard monetary aggregates or the federal fund rate. In addition I find that real factors like industrial production play a more crucial role than monetary aggregates in affecting exchange rates. The impulse response functions to different shocks reveal that there is no evidence of exchange rate overshooting. (Abstract shortened by UMI.)...