Predictability of currency returns: The evidence from over-the-counter options market

Abstract

Can we make an accurate foreign exchange rate forecast if we know what is the market’s best “guess” on future direction of that rate? The author attempts to answer this question by investigating the predictive ability of risk-reversals - market traded derivative contracts that measure the expected skewness of exchange rate distribution. In order to find evidence in favor or against the use of risk reversals based forecasts the paper presents analysis of recent research in selected subject. The author applies econometric methods based on previous empirical works, to quantify the relationship between EURUSD exchange rates and 1 week ahead market expectations embedded in current prices risk reversals for the period of 01/2006-04/2010. The results obtained in the sample period show that variation in weekly changes of risk reversals can explain up to 55% in variation of the same period currency returns pointing to significant positive relationship. But the outcome of out of sample predictability test, in selected specification, could not beat benchmark Random Walk model with RMSE ratio of 1.06. Despite the low predictability, the evidence on risk reversals documented in this research paper contributes to existing literature by using weekly sampling and direct market quotes of risk reversals as explanatory variables to avoid “error in estimation” problem. Moreover the period before and after the events of fall 2008 is analyzed as well as robustness checked within several sampling methods. The debate of using options market implied expectations for accuracy of forecasts is still open and it seems that there is more uncovered issues left for research.