Description of Empirical Study and Results

In the following two sections we will focus on the task of predicting a rise (labeled “+1”) or fall (labeled “?1”) of daily EUR/GBP, EUR/JPY, and EUR/USD exchange rate returns. To predict that the level of the EUR/USD, for instance, is close to the level today, is trivial. On the contrary, to determine if the market will rise or fall is much more interesting for a currency trader whose primary focus is to buy the base currency if the exchange rate went down and to sell the base currency if the exchange rate went up.

All of the following experiments were performed on a PC equipped with an Intel Pentium M Processor 750 and running at 1,866MHz with 1,024MB of system memory under theWindows XP operating system. The simulations were programed in the R environment [201], an open source and high-level programing language that provides powerful tools for statistical analysis. The R packages 1071 [70] and kernlab [220] were adopted for the SVM model fitting. These packages use the SMO algorithm that is implemented by the LIBSVM tool [70].