Fuzzy Mean-CVaR Portfolio Selection Based on Credibility Theory

This paper develops a fuzzy portfolio selection problem that minimizes conditional value-at-risk (CVaR) and estimates CVaR by fuzzy credibility theory and also calculates expected return by fuzzy credibility mean. Using fuzzy techniques makes the model more precise and accurate due to uncertainty of financial data. The use of CVaR helps investors make better decisions because it indicates the size of loss. This study considers some constraints for model including liquidity, cardinality, minimum and maximum investment proportion. The liquidity constraint is measured by turnover of each asset as a trapezoidal fuzzy number. The liquidity constraint converts to a linear constraint by using fuzzy credibility theory. Using CVaR as a risk measurement and efficient constraints makes the model appropriate and adequate for portfolio selection. Finally, a numerical example is provided by 10 stocks chosen from Tehran Stock Exchange Market in 2015 and it shows the effectiveness and applicability of the proposed model.