An econometric model of EUR/USD exchange rate: what is the role of macroeconomic and financial factors?

Abstract

Exchange rates have always been a field of interest for economists, fund managers, investors, specialized traders, companies, as well as merchants who would like to buy or sell commodities and other products outside their country, hedge their positions or even speculate from their fluctuations. The EUR/USD is one of the most important exchange rates; if not the most important globally, owing to the fact that it is the most actively traded currency pair and the huge number of funds and the investors that exist in the Eurozone and in North America. Furthermore, we shall not forget that the price of oil is determined in dollars so a great number of transactions daily take place between these two economic zones involving significant values. The purpose of this dissertation is to study, based on the theory that determines exchange rates, the effects of macroeconomic and financial factors in the EUR/USD exchange rate. First of all, after a brief literature review that reviews the work that has been already done from researchers concerning our exchange rate, the econometric methodology that we will use is referred. Afterwards, we will move to the empirical analysis where we will study meticulously the relationship between macroeconomic and financial factors and the EUR/USD exchange rate and we will form both the long-term, as well as, the short-term model with independent variables chosen after many tests, mainly with the trial and error method and dependent variable first the logarithm of EUR/USD rate and then the first difference of the latter for our short-term model. Moreover, we will try to forecast with the help of our econometric model this exchange rate after the 2008 financial crisis that led to a great global recession in order to realize the predictability and reliability of our model, its capability to beat the random walk in the short-term level and if there is a possibility for this specific model to be used for future forecasts. Finally, we will conclude our project evaluating our model and analyzing how it could be used for predictions with greater reliability.