درخت حوزه‌های تخصصی

The research on the Markowitz model and optimization of its portfolio using a variety of evaluation indicators and meta-beta-algorithms has always been the focus of attention of accounting and finance researchers. The results of studies carried out by various types of optimization methods are different in the Markowitz modified models. The purpose of this study is to measure the optimal portfolio and its corresponding return, with respect to the portfolio in the traditional Markowitz model, as well as to compare the position of the refining and petrochemical companies versus stock market outperformers, through integrating the operational criteria and the new indicators of liquidity using the genetic algorithm in the Markowitz model. Therefore, financial data related to the research variables for 35 cases of TSE-listed refinery and petrochemical companies from 2012 to 2016 fiscal years were extracted from Rahavard Novin database software and simulated by the genetic algorithm. The results show that returns on the stock portfolio optimized using the genetic algorithm and without considering the liquidity limitations and filters have a significant and positive difference with the return on the stock portfolios optimized with regard to the liquidity limitations and filters. Furthermore, the application of liquidity limitations and filters in the formation of optimal stock portfolios leads to a conservative increase in the choice of stocks (portfolio formation), which leads to a reduction in the risk and return of investment in such portfolios.

This study aims at getting a better performance for optimal stock portfolios by modeling stocks prices dynamics through a continuous paths Levy process. To this end, the share prices are simulated using a multi-dimensional geometric Brownian motion model. Then, we use the results to form the optimal portfolio by maximizing the Sharpe ratio and comparing the findings with the outputs of the conventional model. To examine the robustness of the results, we have evaluated its performance for different investment horizons and various volumes of price information over a long period (approximately twenty years) in the Tehran Stock Exchange (TSE). Findings indicate that within the trading dates spanning the interval 24-Mar-2001 to 19-Sep-2020, the return of the portfolios obtained from applying this simulation scheme for maximization of Sharpe ratio is (244% on average) higher and their risk (standard deviation) are lower (1227% on average) than those realized by the conventional methods. Additionally, a comparison of the simulation approach with a performance of the actual market portfolios indicates that the Sharpe ratios of the simulation method are higher (0.055% on average) than those resulting from the total market performances. The results of the stochastic dominance test show that our proposed strategy has a first-order stochastic dominance (FSD) over the conventional one and market portfolios, that means at each level of cumulative distribution, the Sharpe ratio of our method is higher, and as FSD test makes no assumptions about the curvature of investors' utility functions, these results do not depend on the degree of risk aversion of investors, and as long as investors prefer a higher Sharpe ratio, they would be better off if they follow our proposed strategy.