應用遺傳演算法於投資組合建構

Application of Genetic Algorithms in Constructing Investment Portfolio

陳俊宏(Jiun-Hung Chen) ； 王素華(Su-Hua Wang)

《創新研發學刊》 9卷2期 (2013/12) Pp. 19-31

摘要

投資組合指的是數種金融資產所形成的集合，包含應該選擇哪些資產建構投資組合，及組合中的每一種資產應該分配多少資金，藉以提高收益並降低風險。本研究提出一個架構，整合不同種類資產的風險等級，並透過投資人對風險的偏好為投資人定義風險屬性，從眾多的資產中篩選出符合投資人風險屬性的資產建構投資組合。並且可供投資人自行選擇投資組合內是由多少個資產所組成，而投資組合的搜尋方法則是利用人工智慧之遺傳演算法，找出風險屬性符合投資人期許且夏普值最高的投資組合。最後研究中將依所提出之架構建置一個軟體系統，實際輸入資產的數據與投資人的資料進行系統的運作，證明本研究之可行性，並從實例中找出系統未來可供改善之處。

關鍵字

投資組合 ； 遺傳演算法 ； 夏普值

並列摘要

The definition of an investment portfolio is a collection of financial assets. A good portfolio would consider which types of assets to invest and the ratio of each asset within the portfolio in order to increase profitability while lowering risk. This study had created a specific method to organize and categorize assets into different risk levels, then, based on an investor's risk preference, chooses which assets to invest in a portfolio based on risk levels. The investors would be able to choose any combination of types and number of assets to invest in their portfolio. The method used to create the most desirable combination of assets with the highest sharpe ratio uses artificial intelligence based on genetic algorithm. Lastly, a computer software was implement based on the method mentioned above. To prove the viability of the method in this study, real live asset statistics and investor information will be used during the process. Hopefully, it will be able to find future improvements in this study.

並列關鍵字

Portfolio ； Gene algorithms ； Sharpe ratio