Performance Competition for ISCIFCM and DPEI Models Under Uncontrolled Circumstances

Abstract

The 2008 financial tsunami had a serious impact on the global financial industry. Thus, portfolio selection has become very important for individuals and companies to arrange their property and corporate financial management. High-return portfolios are usually accompanied by high risk; therefore, reducing risk and receiving remuneration are the main objectives. In short, constructing a satisfactory portfolio is very difficult. This paper proposes a new model to solve this problem. First, it uses the Investment Satisfied Capability Index and Fuzzy C-means Clustering (ISCIFCM) model developed by Chang and Chen (ICIC Express Lett 3(3):349–355, 2009) [4] and the DEA Portfolio Efficiency Index (DPEI) model proposed by Murthi et al. (Eur J Oper Res 98:408–418, 1997) [1] for stock selection in the securities market of Taiwan (Murthi et al. in Eur J Oper Res 98:408–418, 1997; Chang in Int J Organ Innov 2(3):225–249, 2010) [1, 3]. Then, Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) are applied to these stocks to find the optimal investment allocation of the portfolio by using the moving interval windows. Comparing the performance of the models, the results show that the stock portfolio returns of the ISCIFCM model in our research is superior to the DPEI model. Its performance is also better than Taiwan Weighted Stock Index (TWSI) and Polaris Global ETFs (Exchange Trade Funds) Stable Fund in any period. The findings confirm that using better strategies for investors could provide improved performance.