We observe that, according to what may be expected, the stocks selected for our optimal tracking portfolio coincide with the larger stocks of the AEX-index.
In addition we note that the minimum weight in the tracking portfolio(that of AHOLD shares) is around 4,7% and the maximum weight (that of Unilever stocks) somewhat less than 14,5%.
Since the AEX-index consists of 25 stocks and is replicated by using just 10 stocks, we have also been able to check whether the GA tracker found is indeed the optimal
portfolio with respect to the training data.
We tried all possible combinations of 10 stocks out of 25 and calculated,based on the training data, the corresponding minimum expected tracking error as given by (2).
The best solution found appeared to coincide with the solution found by hybrid GA.
So, an important first conclusion of our experiments is that the hybrid GA is able to consistently find the optimal solution.
We observe that, according to what may be expected, the stocks selected for our optimal tracking portfolio coincide with the larger stocks of the AEX-index.
In addition we note that the minimum weight in the tracking portfolio(that of AHOLD shares) is around 4,7% and the maximum weight (that of Unilever stocks) somewhat less than 14,5%.
Since the AEX-index consists of 25 stocks and is replicated by using just 10 stocks, we have also been able to check whether the GA tracker found is indeed the optimal
portfolio with respect to the training data.
We tried all possible combinations of 10 stocks out of 25 and calculated,based on the training data, the corresponding minimum expected tracking error as given by (2).
The best solution found appeared to coincide with the solution found by hybrid GA.
So, an important first conclusion of our experiments is that the hybrid GA is able to consistently find the optimal solution.
การแปล กรุณารอสักครู่..
