investigating portfolio performance with higher moment considering entropy and rolling window in banking, insurance, and leasing industries

According to modern portfolio theory, diversification should cover the risk. this theory is based on the normality of asset return. experimental findings indicate that the assets return non-normality. higher moments are sed to upgrade traditional models with the primary presumption of a normal distribution in recent years. this study uses a higher moment and the entropy for diversification and selects a portfolio given a non-normality assumption. it is essential to use up-to-date information to increase the model's efficiency, and accordingly, we used the rolling window for new price information. for the financial information method, we use the total index return in the last five working days and weigh the shares of the banking, insurance, and leasing industries on the next working day and evaluate this for three years. python, math, and numpy libraries were used to analyze the data. the comparison between models based on the portfolio evaluation indices indicates that, given using entropy for diversification, a much higher moment model can provide better portfolio selection results in most cases. the results showed that the mean-variance-skewness-entropy model, according to the performance evaluation criteria of asr, madr, ssr, omega, and jensen, and the modified treynor show better performance than the other models and only in the sr evaluation model, which is somewhat traditional, it has shown poorer performance than other models. therefore, the hypothesis of using entropy as a criterion to improve portfolio performance can be confirmed. comparing the models based on portfolio evaluation indices indicates that the use of entropy for diversification does not significantly reduce the optimal values of other objective functions. as observed, higher efficiency was obtained when using entropy and higher moments than in other models.