Does anyone know what position sizing method should be used during system design, optimization and verification? Does the method change at some stage in the strategy development cycle?
For example, should strategies always be developed and tested, optimized and WFA with fixed shares as position size and then position sizing should be added only after success just before real-time trading?
Or is there a step after which position sizing is changed/introduced such as after the design is complete but before the optimization?
I am looking for a detailed scientific explanation as to what the correct procedure should be from a logical and math and statistical perspective. I am not looking for non-substantiated opinions.