Evaluation of investment portfolio models in mutual fund in global financial markets (with emphasis on multi-objective meta-heuristic algorithm)

The complexity of financial instruments and markets makes it difficult for investors to decide on the type of asset, so that investors are always faced with the problem of optimizing the set of assets in their decisions. Therefore, choosing the right investment portfolio in order to maximize profits is one of the main concerns of investors. With this statement, the purpose of this article is to compare the explanation and performance of the optimization problem and the predictive power of ARMA-historical simulation models and ARFIMA-Monte Carlo (one of the most established models in the world) in fund portfolio optimization ‌ is a joint venture. The statistical and sample population included data from selected funds traded on the stock exchange of selected member countries of the Federation of Asian and European Stock Exchanges (FEAS) between 2013 and 2019. The results showed that the ARIMA-value at risk (historical simulation) model has a higher efficiency boundary compared to the ARIMA-value at risk (Monte Carlo simulation). It also incorporates the Pareto frontier drawn by the PESA-II algorithm for the other model. To determine the significance of this difference, the performance of the Mann-Whitney test has been investigated. The results indicate that the ARIMA-value-at-risk-optimal portfolio Sharp criterion is better than the ARIMA-value-at-risk (historical simulation) criterion.