optimal asset allocation in the presence of macroeconomic uncertainties and international sanctions against iran

the current study addresses an estimation of investor's optimal portfolio under conditions of uncertainty by using a combination of artificial neural network and markowitz models. for this purpose, such assets as stock prices, house prices, coin and bonds price are used with monthly data over the period 1378-1392. three variables including inflation uncertainty, oil uncertainty and free market dollar rate are used as state variables to investigate the impact of macroeconomic shocks on investor's decisions when choosing an optimal portfolio. autoregressive conditional heteroskedasticity (garch) is used to estimate state variables. following an estimation of the state variables, assets return and uncertainty were measured using multi layer perceptron (mlp) and radial basis function (rbf) neural networks. the results obtained from neural network model are used as input variables in estimation of markowitz's optimal portfolio. the results of analyzing mean variance show that housing is the dominant asset among uncertain assets over the period of real-estate boom holding the largest share of assets. over the recent periods, considered to be the period of housing downturn, investors no longer include housing in their portfolio turning instead to stocks and coins as prominent alternatives. generally, bonds have shown to be an asset with no uncertainty in all the periods making them a reliable alternative in the investor's optimal portfolio.