The beta of the strategy for the out-of-sample data is 0.01, so it has the advantage of being market-neutral.
## Conclusions
The objective of this experiment is to see how one could use genetic algorithms to find trading strategies. A lot of improvement can be done, of course. For example, more complex strategies can be used, in order to obtain a richer final algorithm. Moreover, the fitness function could be improved as well, to include features you want the algorithm to have.