|
Until now, many investors only use a single strategy to engage in investment transactions such as trend trading strategy and countertrend trading strategy. However, market changes rapidly, and investors often cannot judge whether the current market trend is trending or bumpy market. When the market trend is in the opposite direction of the investor's strategy, it is very likely cause heavy losses for the investors. In order to solve this problem, this study uses the k-means clustering method to divide futures prices into two groups: trend and bumpy market. According to the clustering results, it switches between trend trading strategy and countertrend trading strategy to verify whether the performance of switching strategies is better than that of single strategies. This study divides into two phases. At the first phases, the data used are the daily data of 25 overseas futures contracts, which span from July 2006 to December 2019. This study considers both return and risk, and the performance indicator used is the return-risk ratio. The empirical results show that the performance of switching strategies is better than that of single strategies. At the second phases, the Walk Forward Analysis is adopt, which is find the optimized parameters in in-sample and bring it into the out-of-sample to examines the performances. Using Mann Whitney U Test and Wilcoxon Signed-Rank Test statistical test, the empirical results show that the performance of switching strategies is better than that of single strategies.
|