Study on the Introducing of Portfolio Theory and Value-at-Risk Model into Robo-Advisor - Take the Taiwan Stock Market as an Example

• Main Authors: Syuan-Ya Huang, 黃暄雅
• Other Authors: Wo-Chiang Lee
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/v9gnwy

碩士 === 淡江大學 === 財務金融學系碩士班 === 106 === The purpose of this study is to select the stocks from the sample to establish the investment portfolio, test the performance, import the feasible method into the robot advisor, and provide the method reference for establishing the investment portfolio into robot advisor management in different samples in the future. In the empirical study, the data is cutting into training sample and test sample by using the moving window method. After selecting stocks through the Markoviz''s mean-variance model in the training samples ,the study calculating the risk values by VaR and the portfolio performance by ratio in the test samples.Import the above methods into the robot advisor, providing a reference for the investment model of the robot advisor management model. The empirical results show that: 1. Using the M-V model to find the optimal portfolio, the results show that under the sample data of this study, the optimal portfolio weights concentrate on a certain number of companies. 2. VaR method is used for risk control. The VaR penetration rate in the test sample is within the control range. 3. Introduce the investment portfolio selection method of this thesis into the robot advisor. After the simple investment selection interface and investor risk tolerance survey are completed, insert the M-V model, so that the financial management of the robot advisor is established based on statistics and performance data.