Comparison between Crisp and Fuzzy Stock-Screening Models

Academic researchers and practitioners have proposed various stock-screening models that always contain more than one stock selecting rule and corresponding parameters. However, the criteria in traditional screening models employ crisp norms, which are unreasonable in reality. This paper proposes the fuzzy stock-screening model to select stocks in the portfolio. The screening rules consist of those regarding the price-earnings ratio, the earnings growth rate, market capitalization, return on equity, and the price-book ratio. Empirical studies with datum from Taiwan’s stock market compare the performance of the proposed stock-screening models with the conventional one. Empirical results show that the portfolio selected by the proposed model outperforms the portfolio by the conventional models in terms of investors’ expectations.