I would like to ask: did anyone check if it is a good advice to throw correlated strategies after backtesting.
Let's say after backtesting we have 3 correlated strategies, A B and C. On backtesting they are similar, but in live B is the best of them. The problem is that B was created after A, therefore it was thrown away by the software, and we lost the best strategy.

I will be glad to hear your opinion.

Avraham Shenkar