Edurecio – A while back I spent a lot of time trying to create a self optimising strategy that checked what had worked best recently and then used those settings until the self optimisation said that something else worked better recently. It produced nice equity curves in back testing but completely failed in forward testing. In every strategy there is some sort of variable that is fixed – whether it is an actual value or just using closing price instead of median price. So something is optimised and fixed in every strategy before we even try to check what worked best recently for other variables and hope that it works for a bit longer with a new value for them. I gave up on the concept of self optimisation and it proved to me that regularly re-optimising does not work either.
Just my humble opinion.
Good Morning.
Vonasi, I agree with you.
But you talk about reoptimizing. I do not speak of optimizing, but of looking for average values that make the system work without pretending to be optimized.
I presume that an optimization of a time series will not work in the future, since then, I am looking for an average obtained with the accumulated optimization of each month. As always, there will be months with better results, months with worse results and even some negative months.
Negative months: for example last month, which was so flat for this system with market exits based on weeklypivots, that would not work under any logical forecast.
I leave a series of data obtained until June.
What I would put in July the average data (blue) obtained until June in the system. I would see how many gains that data gives me (10) but at the end of the month, I would calculate what the optimized data would have been (3) to have an average to use again in August. And so month after month, observing a trend in the means towards a midpoint of gains in the system.
I am not looking for an optimization, I am looking for an average of valid values.
It will not be a perfect system because I think there is not, but it seems to me an approximation that should be valued.
