Portfolio selection by Using Multi Attribute Decision Making based on grey relational analysis and linear programming

One of the most important challenges in the stock market for investors is selecting Portfolio. This study by considering the financial ratios as evaluating indicators tries to determine appropriate model for investment decisions in stocks. In this study, the combination of respectively models, linear regression, multi attribute decision making and linear programming were used to forecast the future of financial ratio trends, ranking companies and asset allocation. In the first step of study, after selecting 17 financial ratios and indicators as variables of the hybrid model, their values from the first quarter of 2007 to the first quarter of 2015 was calculated for companies included in the sample. Then by using the moving average with exogenous inputs and Auto Regressive Moving Average with exogenous input, these variables were predicted for the studied period (second quarter 2015). In the next step, we used Shannon entropy to determine the weight of indexes and the grey relational analysis for ranking companies. Finally, by using a linear programming model, a model was developed to select the optimal portfolio. According to this model, a portfolio of stocks formed and by using the Sharp ratio, its performance was compared with the overall index and the top 50 index. The results showed that the hybrid model has had a better performance than the overall index and the top 50 index, in the period of the study.