The paper focused on measuring efficiency of investment strategies and portfolio optimization based on dynamic portfolio formation using the global minimum variance approach in a region of central European countries. The paper analyses DCC GARCH model, which was employed in order to obtain conditional correlation matrices. The analysis includes a comparison of global minimum variance (GMV) and newly proposed least correlated assets (LCA) portfolio formations based on individual shares and market indexes. Performance of constituted portfolios showed that dynamic form of portfolio optimization is an efficient tool in profit maximization and volatility minimization. The study shows that there is a potential for improvements of proposed methods. LCA portfolio formation showed that the number of parameters could be effectively lowered without a loss of profit.