Temporal aggregation effects on the construction of portfolios of stocks or mutual funds through optimization techniques. Some empirical and Monte Carlo results

In this paper we test the effects of temporal aggregation (disaggregation) on the efficiency of portfolio construction using the mean variance optimization approach. Using Monte Carlo techniques and empirical data from the Athens Stocks Exchange we confirm that the use of temporally aggregated data effects very seriously the efficiency of the constructed portfolio. Especially as the degree of temporal aggregation increases the application of optimization techniques could lead to different results regarding the percentage of stocks participation, the weights and finally the total portfolio performance.