查詢結果分析
來源資料
頁籤選單縮合
| 題 名 | 多重技術指標之動態投資組合績效評估策略--以臺灣50為例=The Application Analysis of Combined Technical Indicators on Investment Performance: Evidence from the Taiwan 50 ETF |
|---|---|
| 作 者 | 蔣齊諺; 楊至中; | 書刊名 | 大數據應用與管理 |
| 卷 期 | 6 2025.12[民114.12] |
| 頁 次 | 頁1-10 |
| 分類號 | 563.5 |
| 關鍵詞 | 多指標組合策略; 技術分析; 臺灣50; 消融實驗; Multi-indicator combination strategy; Technical analysis; Taiwan 50; Ablation experiment; |
| 語 文 | 中文(Chinese) |
| 中文摘要 | 本研究針對技術指標的多重組合策略在台灣 50 ETF(0050)投資中的效能進行分析。透過 2015 年 至 2025 年的日資料進行回測,我們首先評估了五項 常用技術指標(相對強弱指數 RSI、移動平均收斂發 散指標 MACD、指數移動平均線 EMA、隨機震盪指 標 KD、布林通道)的單一策略表現,接著以多數決 機制整合各指標的最佳參數策略構建多指標組合 (Full Model),並設計嚴格(5/5)、標準(3/5) 與寬鬆(2/5)三種門檻檢驗決策嚴謹度對績效的影 響。實驗結果顯示,多指標策略相較單一指標能有 效濾除雜訊,提高交易訊號的精準度和平均獲利水 準:例如在標準門檻下僅產生 7 筆交易但平均單筆 報酬高達 11.74%,勝率 57.14%,證實多重確認提升 了預測效力。寬鬆門檻下交易次數雖增至 36 筆且累 積總報酬率達 95.03%,但勝率下滑至 44.44%,單筆 獲利降至 2.64%,顯示降低門檻能捕捉更多行情卻 伴隨較高雜訊干擾。透過消融實驗(逐一移除指 標),本研究量化了各指標的邊際貢獻:結果發現 EMA 趨勢指標對整體績效提升最為關鍵,移除後總 報酬率驟降近九成;布林通道次之,去除後獲利亦 明顯下滑;MACD 則提供了重要的趨勢確認資訊, 而 KD 訊號與其他指標重疊度較高,移除後策略穩健 性反而略有提升。上述實證證明,多指標整合策略 能在一定程度上優化交易績效並降低錯誤訊號,同 時以消融分析評估各指標的獨特價值,提供了技術 分析應用於 ETF 交易的實務貢獻。 |
| 英文摘要 | This study investigates the effectiveness of multi- technical-indicator combination strategies in investing in the Taiwan 50 ETF (0050). Using daily data from 2015 to 2025, we first evaluate the performance of five common technical indicators—Relative Strength Index (RSI), Moving Average Convergence Divergence (MACD), Exponential Moving Average (EMA), Stochastic Oscillator (KD), and Bollinger Bands— under single-indicator strategies. We then integrate their optimal parameter settings through a majority-vote mechanism to construct a comprehensive multi-indicator model (Full Model). Three decision thresholds are designed—strict (5/5), standard (3/5), and loose (2/5)—to examine how varying levels of consensus affect trading performance. The results show that the multi-indicator strategy effectively filters out noise and enhances signal accuracy and profitability compared to single-indicator approaches. Under the standard threshold, only seven trades were generated, but the average return per trade reached 11.74% with a win rate of 57.14%, confirming the predictive value of multi- confirmation. The loose threshold produced 36 trades and a cumulative return of 95.03%, but with a lower win rate (44.44%) and weaker average returns, indicating that relaxing the threshold captures more opportunities but a l s o i n t r o d u c e s h i g h e r n o i s e i n t e r f e r e n c e . Ablation experiments, in which indicators were removed one by one, quantified the marginal contribution of each component. The EMA proved to be the most critical trend indicator, as its removal caused nearly a 90% drop in total returns, followed by Bollinger Bands, while MACD offered essential trend confirmation. In contrast, the KD indicator showed high redundancy with others, slightly improving stability when excluded. These empirical findings demonstrate that integrating multiple indicators can enhance trading performance and reduce false signals. The ablation analysis further clarifies each indicator’s unique role, providing practical insights for a p p l y i n g t e c h n i c a l a n a l y s i s i n E T F t r a d i n g . |
本系統中英文摘要資訊取自各篇刊載內容。