Can Affine Term Structure Models Help Us Predict Exchange Rates?


  • I am very grateful to Ken West (the editor) and two anonymous referees for valuable, extensive comments that greatly improve the article. I also thank Nour Meddahi and Enrique Sentana for their useful comments and suggestions during the early stages of this project. This project started while I was a postdoctoral fellow at CIREQ and CIRANO; and their hospitality is greatly appreciated. Finally, I want to thank Greg Bauer, David Bolder, Qiang Dai, René Garcia, Alfonso Novales, Adrien Verdelhan, Jun Yang, and seminar participants at the European Central Bank, the CIRANO-CIREQ Conference on Financial Econometrics (Montréal, 2005), Finance Forum (Madrid, 2005), Bank of Canada Conference on Fixed Income Markets (Ottawa, 2006), International Symposium on Forecasting (New York, 2007), Northern Finance Association Meetings (Toronto, 2007), and Financial Management Association Meetings (Orlando, 2007) for their comments. However, I remain solely responsible for any remaining errors. The views expressed in this paper are those of the author and do not necessarily reflect those of the Bank of Canada.


This paper proposes an arbitrage-free model to extract the information that the term structure of forward premia contains for forecasting future spot exchange rates. Using monthly data on four U.S. dollar bilateral exchange rates, we find evidence that this model provides statistically better forecasts than those produced by a random walk for the British pound and Canadian dollar exchange rates. Negative results for the German mark/Euro and Swiss franc are explained by a rejection of the restrictions imposed by the term structure model.