Date of Submission

Spring 2011

Academic Program

Economics and Finance


Dimitri Papadimitriou

Abstract/Artist's Statement

After the end of the Bretton Woods agreements, exchange rates forecasting has become a fairly complex and intriguing problem for economists around the world. This paper addresses the issue of exchange rates theory and focuses on the USD/EURO forecasting. We will analyze and contrast an established model, the MEMMOD developed by the German Central Bank, with a self-made VAR model. The VAR model will be developed based on an initial set of variables, and limited to the best match. Once the model is built, both the MEMMOD and VAR will be used to exhibit in-sample forecasting for eight periods (Between 2004-2006). The values used in this simulation are quarterly data, and the exchange rate forecast is also made on a quarterly basis (one quarter is equivalent to one period). The results correspond to the literature, namely that exchange rates are hard to predict and models, though capturing a lot of information about the movement of exchange rates, cannot accurately forecast future exchange rates.

Distribution Options

Access restricted to On-Campus only

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.