Dynamics of exchange rate changes: Bayesian forecasting with dynamic linear models

The objective of this book is to empirically evaluate the parameters which drive exchange rate changes. The analysis will be performed with a dynamic linear model in a Bayesian framework. It will be demonstrated how to get from a static first order polynomial model to a dynamic regression model, incorporating the predictive parameters such as “purchasing power parity”, “interest rate differentials” and “volatility index”. Discussion will focus on how different currency pairs react on specific parameters. While analyzing data in the long-run as well as during the financial crisis from 2008-2009 it will become clear that specific parameters dominate specific business cycles. Putting all this information together the model will be compared to a simple trading strategy to show that using this model an investor can earn excess returns between 1.23% and 1.9% p.a.