本論文研究目的在於不同投資組合下的各種投資績效與實際市場上的台灣50績效相比較，找出較適當的投資組合，在加入了機器人理財顧問後，作為提供金融業發展機器人顧問的選配模型建議與方法。實證中，將台灣50中前十大比例成分股的日價格資料分為樣本內與樣本外資料，求出各投資組合所需個股權重，並將結果代入樣本外資料，最後再進行投資組合績效分析。本文實證結果顯示： 1.利用M-V與CVaR模型代入樣本內資料求出各投資組合最適個股權重，結果發現M-V模型下的最適投資組合權重較集中，而CVaR的投資組合權重則較均勻分散。 2.將樣本外資料代入各投資組合的最適各股權重得出投資組合報酬，並與台灣50進行績效分析，得出績效較佳的模型為M-V模型。 3.將結果導入金融科技的機器人理財議題後，探討未來在金融科技的浪潮下，財富管理的投資組合建議，結果發現對於未來的機器人理財的模型配置，可以參考本研究的方法做更進一步的績效分析。 The purpose of this study is to compare the investment performance of different investment portfolios with the performance of Taiwan 50 Index in the market. After adding the Robo-advisor, find out the optimal investment portfolio to be recommendations and methods of choosing models. In the empirical, the daily price data of the top ten constituent stocks in Taiwan 50 Index are divided into the in-sample data and the out-sample data and inferring the weights of each portfolio. The results are substituted into the out-sample data and analyzing the portfolio performance. The empirical results show that: 1.Using M-V model and CVaR model into the in-sample data to find the optimal weight of each portfolio, the result indicates that the optimal weight of the M-V model is concentrated and the portfolio weight of CVaR is evenly distributed. 2.Using the out-sample data into the optimal portfolio of each investment model, and comparing the portfolio performance. The result shows that the performance of M-V model is better than the performance of Taiwan 50 Index. 3.After the result into the issue of FinTech, probing the recommendations of investment portfolio under FinTech. The result indicates that to choose the investment model for robo-advisor in the future, we can refer to the method of this study.
