The lower partial moments risk measure in a novel fuzzy framework based on possibility density function. (July 2022)

• Record Type: Journal Article
• Authors: Deng, Xue; Chen, Jiaxing
• Is Part Of: Computers & industrial engineering. Volume 169(2022)
• Publication Date: 2022-07
• Subjects: Fuzzy set -- Portfolio selection -- Fuzzy lower partial moment -- Possibility density function -- Portfolio-driven method; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2022.108309
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: The lower partial moments are comprehensive and flexible downside risk measures. In previous studies, the calculation methods had high computational complexity. The proposed LPM can be calculated more simply and effectively, and its feasibility is confirmed by a numerical example. The proposed method in a fuzzy framework is an extension of the random framework. Abstract: As a comprehensive and flexible risk measure, the lower partial moments (LPM) can be converted into a wide range of the downside risk measures by modifying its target parameter and degree of moment. However, the distribution function of the return rate is hard to recognize, and the existing calculation methods have great calculation difficulties, so it has limited application. In previous studies, the historical data was mostly used to simulate the distribution function of return rate for approximate calculation of LPM risk, resulting in a huge computational complexity. In this paper, based on the possibility density function, the LPM risk measure is creatively defined in a novel fuzzy framework instead of a traditional random framework. By deriving the trapezoidal fuzzy number of return rate from the historical data, the LPM risk can be easily calculated according to its possibility density function, which can significantly improve the calculation process. Based on this idea, this paper derives the specific expression when the rate of return is a trapezoidal fuzzy number under the definition of fuzzy LPM. Furthermore, stock-driven and portfolio-driven methods are used to calculate the portfolio risk, and then two fuzzy mean-LPM models are constructed respectively. Finally, a numerical example based on global stock market index is used to illustrate the feasibility and validity of the proposed methods and models. Compared with the classic mean–variance model, our proposed model with adjustable target parameter and degree of moment is more flexible in reflecting investors' attitude towards risk.