This type of optimization is a good way to find the best parameters within the various in-sample timeframe using a reproduction mechanism that is very similar to what happens in nature.

Let us use this simple graphical representation:

As we can easily see from the example, the best two combinations from every optimization cycle (elements 1 and 2) are cross-mixed, generating a hybrid combination of parameters as a result of the two combinations up above (element 3 in our example) such combination will feature part of the parameters from element 1 and part from element 2. But it's not over!

The genetic optimization mechanism does not stop by simply adopting this hybrid combination of parameters, it makes a further step (or put into a stress state, so to speak) between element 3 and 4 the inserting of a random mutation in one or more parameters is foreseen, in order to simulate what really happens in nature and in biology in general.

Such casual mutation could reduce the system performance during the out-of-sample application, since it is a distorting and perturbing element.

On the other hand, it really strengthens the reliability of the test as well as the resulting equity line, especially if it can produce good and stable outcomes despite this random interference.  

Privacy Policy