dc.description.abstract | In this analysis, three simple technical trading rules - filter rules, moving average rules and channel breakout rules - are tested in the foreign exchange market in the period from 1986-2016. The technical trading rules are tested on three currencies; British Pound, Japanese Yen and Swiss Franc relative to the US-dollar. The study applies three performance measures; mean excess return, Sharpe ratio and Jensen’s alpha to evaluate the predictability of technical trading rules. The performance is tested by using a standard t-test and a residual bootstrap by Lebaron (1998). To correct for data snooping, it is included additional weeks to the technical trading rules. The full-sample t-test suggest that several trading strategies generate significant profitable returns. However, some of the trading rules reveal data snooping bias and the profitability is due by chance rather than merit. In addition, the positive returns do not survive risk adjustments and reveal small signs of robustness. The residual bootstrap indicates that the technical trading rules show lack of predictive power to exploit the foreign exchange market. Overall, the study suggests that the frequent recommendations of technical analysis by practitioners is most likely explained by opportunistic choice of individual rules, which happened to perform well in the past. | nb_NO |