Dynamic Portfolio Optimization with Transaction Cost

The optimal selection of portfolio is one the most important decisions in managing investment funds. Several approaches have been proposed to determine what is the best trading strategy. Most of these common approaches are based on single period optimization, however, as investment is a long-term concept, such short term profit maximization cannot fully exploit the opportunities that an investor might get if he/she looks into a longer term. To this end, in this paper, we intend to extend the single-period optimization into a multi-period optimization and also, to make it more realistic, we will incorporate transaction cost in our model. To investigate the validity of the proposed scheme, we have analyzed several examples using which we have presented the steps of our approach and also statistically compared the performance of the single and multi- period optimization using Mann-Whitney U test. Based on the results of this paper, we can conclude that multi-period and single-period optimization might have similar performance if we look at short time span of the system. However the superiority of multi-period optimization becomes more evident as we extend the time span of the system which gives multi-period scheme more freedom to suggest better portfolio selections.