Predicting Future Spot Rates On The Basis Of Forward Rates

• Published in: Quarterly journal of business and economics Vol. 27; no. 3; p. 57
• Main Authors: Soenen, Luc A, van Winkel, E. G. F.
• Format: Journal Article
• Language: English
• Published: Lincoln Creighton University, College of Business 01-07-1988
• Subjects: American dollar; Economic theory; Forecasting; Foreign exchange rates; Foreign exchange translations; Forward exchange; Mathematical models; Rates; Spot; Statistical analysis; Time series
• ISSN: 1939-8123; 2327-8250

Time-series analyses of 10 currencies shows that time-series predictors outperform forward rates in predicting future spot rates for one-month contracts. For 3-month contracts, time-series models perform as poorly as forward rates. The currencies considered are: 1. Belgian franc, 2. Canadian dollar, 3. Danish krone, 4. German mark, 5. French franc, 6. Italian lire, 7. pound sterling, 8. Norwegian krone, 9. US dollar, and 10. Swedish krone. Calculations are based on end-of-the-week exchange rates against the Dutch guilder. The one-month time-series model works best with currencies of European Economic Community nations (except the lire), the Swedish krone, and the US dollar. The 3-month model yields good predictions for the mark and the French franc. The percentage yield for the Norwegian krone, pound, and Canadian and US dollars is substantial. For the other currencies, which are linked as part of the European Monetary System, the percentage yield is much smaller.