I think I found a bug in your walk forward back testing,
Anytime an optimization iteration will find 0 trades, it will test the 'out sample' data with a total random inputs from the optimizer, this is highly unpredictable and unreliable.
Is this supposed to be the way this test work?
Why not choose as default let say Minimum values of the optimizer if no trades been found in the 'in sample' data.
Hope you can fix this.
Thanks in advance for any input on the subject
Comment