Dynamic Customized Portfolio Selection－Multi-objective Stochastic Programming

• Main Authors: LU, KENG-FU, 盧畊甫
• Other Authors: HUANG, JIH-JENG
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/68hse9

碩士 === 東吳大學 === 資訊管理學系 === 105 === In an era of inflation and salary freeze, many people invest in stocks to increase their incomes. But stocks are very risky commodities. How can we reduce the risk of investment? How to make long-term steady profit? It has always been a question for investors to think about. The purpose of this study is to establish a dynamic customized portfolio strategy that can effectively keep profit and reduce risk. This strategy will be based on investor preferences, recommend suitable portfolio for the investor, and according to market changes, dynamically adjust the portfolio of investment targets and weight structure, bring the portfolio to maintain low risk and high yield. This study using January 2013 to December 2016 of the Taiwan 100 stocks as investment targets, users fill in customized portfolio questionnaire, calculated weights by the analytic network process (long-term returns, short-term returns, yield rate, long-term risk, short-term risk, beta value), and then calculate the total score of Taiwan 100 with the highest scores top 10 as the user of the most suitable portfolio, then use multi-objective stochastic programming to calculate the proportion of each stock investment, then use constant proportions portfolio insurance strategy(CPPI) to adjust the portfolio during the Back-Testing, every six months to dynamically adjust the portfolio, during the period, 8 times. Finally, the performance results are compared with the Taiwan 100.The empirical results show that each questionnaire has at least 4 times, the Treynor index exceeds the Taiwan 100, and the Jensen index is the same; in addition, the average of Treynor index and the average of Jensen index of each questionnaire both exceeded the Taiwan 100, this study uses the CPPI strategy to dynamically adjust the portfolio, and has better investment performance in the long-term investment than the Taiwan 100.