投資組合模式之建構──以寶來ECFA指數成份股為例

Construction of the Investment Portfolio Model – An Example of Polaris ECFA Index Constituents in Taiwan

近年來隨著台灣與大陸互動頻繁，加上ECFA的簽訂，使得與ECFA議題相關的產業都有不錯的表現。本文提出一個投資策略，首先，使用資料包絡分析法選出寶來ECFA指數成份股中相對效率較高的投資標的，再利用組合編碼遺傳演算法做資金的配置，最後以技術指標分析買賣的時機，並與大盤的平均報酬率作比較。研究內容以寶來投信所提供的寶來ECFA指數成份股，結合台灣經濟新報提供的2008年至2010年的股價資料及財務報表作實證分析。實證結果為：1. 本方法能有效控制報酬率的波動，且長期投資時有優於大盤的報酬表現。2. 本方法有優於同質性文獻的報酬表現，驗證此投資組合方法有其學術價值，且能套用在其餘類型股票上。

In recent years, Taiwan and China have been trading more frequently. Stock prices of Taiwan have done well since signing the Economic Cooperation Framework Agreement (ECFA) with China. This research will propose an investment strategy. First, Data Envelopment Analysis (DEA) is used to choose higher relative efficiency of investment targets. Second, Combination Encoding Genetic Algorithm (CEGA) should be used to allocate funds. Finally, technical analysis is used to find the trading opportunity. Then to compute the return rate and compare with the market's average return rate. This paper uses the ECFA index constituents which are provided by Polaris International Securities Investment Trust. Empirical analysis data use stock price and financial statements of Taiwan Economic Journal (TEJ) from 2008 to 2010. The empirical results are as follows: 1. The investment strategy can control fluctuation in return and achieve better performance than the market’s average in long-term investments. 2. This method is compared with the literature, and also provides a better return rate. It verifies that the method has its academic value, and can apply to other stock types.