How often should an EA be re-optimized ?

[[Deleted]]

Hi,

How often should an EA be re-optimized ?

Say I'm aiming H4 or H1 charts... How often should I re-optimize it ?

Is there a common sense or rule for that ?

Thanks.

[[Deleted]]

What there are, are many schools of thought. Religions almost.

Some famous fx folks suggest that optimization is a waste of time.

My opinion: it depends very much upon your particular strategy and each parameter may well have a different optimization lifecycle.

CB

[Patrick]

You should utilize some walk forward testing to see how often to optimize your EA. Run an optimization where you get at least 50 trades (at least 50 because statistics tells us that a pool of at least 50 samples is less likely to be subject to luck or extraordinary circumstances). And then apply those results to some out of sample data (the immediate X months after the optimization) to see how the optimized results perform. Then check to see when (how many months) the optimized results start to wear off. Do this several times over several data samples to figure out the best optimization schedule for your system.

BB