本研究探討各項技術指標包含隨機指標(KD)、相對強弱指標(RSI)、威廉指標(WMS%R)交易策略之獲利性,接著將各項技術指標之買賣訊號結合成為聯合訊號交易策略並測試獲利性並且將交易頻率區分為5分鐘、15分鐘、60分鐘之個別測試獲利性,最後使用證據理論融合各項技術指標之買賣訊號交易策略並測試獲利性與聯合訊號交易策略進行比較。本文使用台灣加權指數期貨為樣本,取樣期間為2001年1月2日至2012年12月28日。 研究結果顯示聯合技術指標交易策略無論在日資料或日內資料都無法有效提高獲利能力,但確實能夠降低單一技術指標之間的矛盾訊號也能夠降低交易次數。證據理論融合各項技術指標交易策略能夠提高聯合技術指標交易策略的獲利能力與勝率並降低交易次數,其中結合隨機指標、相對強弱指標與威廉指標之買賣訊號交易策略組合在日資料交易頻率與日內交易頻率之60分鐘提升聯合訊號交易策略的報酬最佳。 關鍵字:隨機指標、相對強弱指標、威廉指標、聯合訊號交易策略、證據理論
The aim of this study is to examine the profitability of trading strategies by using technical indicators consisted of KD, RSI and William %R. Each of three technical indicators is tested for trading performance respectively, and then integrated with Combined Signal Approach (CSA). Both the single and CSA strategies are evaluated with various performance metrics. In addition, evidential reasoning approach is also applied to combine the outcomes of the single indicator and CSA for comparison. The experimental data is sampled at different frequencies including 5-minute, 15-minute, 60-minute and end-of-day data over the period of 02/01/2001 through 28/12/2012 from TAIEX Futures. Our empirical results show that CSA fail to increase the short-term profits for both end-of-day and intraday data. However, CSA is capable of reducing conflicting and contradictory nature of single indicator and reduce the number of trades. The evidential reasoning approach which fuses different indicators can improve the winning ratio and decrease the trading frequency. Among all trading strategies, the integration of KD, RSI and William %R by evidential reasoning approach and filtered with 60-minute and end-of-day data outperform the other trading strategies and achieve the best profitability. Keywords:Stochastic,Relative Strength Index,Williams Overbought,Combined Signal Approach,Dempster-shafer Evidence