Performance Comparison between Switching Strategy and Single Strategy

• Main Authors: Ming-Qi Ma, 馬銘淇
• Other Authors: Jiang-He Xu
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/x29a78

碩士 === 國立虎尾科技大學 === 財務金融系碩士班 === 106 === Most investors adopt single strategy during trading. However, a period of loss would take place when the trend is opposite to the strategy no matter trend trading or oscillator trading was adopted. The movements in the market are considered to be formed by investors’ judgements which mostly depend on frequency of data or charts. In comparison with the performances among trend trading strategy, oscillator trading strategy and Trend-Oscillator Switching trading strategy, this study tries to examine whether the Trend-Oscillator Switching trading strategy could outperform the other two strategies. The study is split up into two phases. At the first phase, the hourly data of near month Taiwan Index futures from January 2009 to March 2017 are used.In order to avoid the price gap, which could cause false loss and false surplus when a position is switched, the prices are modified to solve out the condition. The empirical results show that the Trend-Oscillator Switching trading strategy has the best performance among the three strategies. At the second phase, the financial index futures and the electronic index futures are added to the study. The data spans from January 2009 to December 2016. Differently from the first phase, the Walk Forward Analysis is adopted, which splits data into in-sample data and out-of-sample data and examines the performances of the out-of-sample data by rolling over the entire period. After implementing the Wilcoxon rank-sum test and Wilcoxon signed rank test, it indicates that the Trend-Oscillator Switching trading strategy outperforms the oscillator trading strategy at the 1% significant level, while the Trend-Oscillator Switching trading strategy performs better than the trend trading strategy at the 5% significant level.