Optimization - what's your approach?

[meanreversion]

Whilst backtesting a system, it's quick to come up with the "optimal" set of parameters which make the strategy profitable. Do you then use these parameters to apply the model to the market?

Furthermore, let's say the strategy is tested on two markets.. the optimal parameters will most likely be different .. should you then use different sets of parameters or attempt to come up with one set which is "best fit"?

 

[Trader333]

Optimization - what's your approach?

I don't optimise, I used to do years ago but not any more as it made no real difference.


Paul

 

[meanreversion]

You arrived at a fixed set of parameters and stuck with them?

 

[Trader333]

You arrived at a fixed set of parameters and stuck with them?

I just found that what worked in the past rarely worked in realtime when I had used optimisation. This is why almost all the MT4 EAs that are supposed to make huge profits fail. I use pivot points and PA now more than anything else.


Paul

 

[zupcon]

My advice is dont waste your time optimising strategies. I would however advise looking at the sensitivity of a given system to its parameters