Goal of this paper
Inspired by the outstanding results of the genetic algorithm,we decided to use this approach for tracking the Dutch AEX index.
We will assume that the number of stocks to be selected is fixed and that the composition of the chosen tracking portfolio is not changed which implies that we will
not analyze optimal timing strategies and, related to that,not take transaction costs into account. Instead, we will focus, in two ways, on the minimization of the tracking
error by determining (a) the optimal subset of index stocks in combination with (b) the optimal weights of these stocks in the tracking portfolio.