Multi-Asset Portfolio Management Including Fixed Income Securities by Value at Risk based Models in Iran Market

This paper aims to provide the best approach to optimize the multi-asset portfolio of five asset classes including cryptocurrencies, foreign currencies, gold, stock and investment funds in three groups of fixed income investment funds, stock investment funds and Mutual investment funds. Purposely, the mean-value at risk and mean-conditional value at risk models have been developed and solved using the artificial bee colony algorithm. The performance of the value at risk based models is compared with the mean-variance, mean-semi variance and mean–absolute deviation models. Also, the profitability of the models is evaluated in the presence of three real world constrains, i.e. quantity constrains, class constrains and the both. Our investigation time period is from August 2015 to December 2020. The in-sample and out-of-sample results showed that the conditional value at risk model outperforms the other models, whether in the presence of constraints or not. Also, the artificial bee colony algorithm was superior to the imperialist competitive and particle swarm optimization algorithms in multi asset portfolio management, based on Sharpe, conditional Sharpe and return on risk ratios.