Page 1 of 1

Walk Forward Optimization - Does it Work?

Posted: Sep 13 2024
by Jeff R
Hello Everyone,

I am a futures trader and have created a number of profitable strategies that show excellent results when back tested. However, the walk forward optimization results for these strategies are almost never profitable. That is why I am still sim trading, and not yet committing real money to my trades. I don't believe that I am over-optimizing my strategies, as they are generally very simple and have just two or three optimization parameters (e.g., look back period and size of stop loss). All of the equity curves look good in back testing; in fact, some are almost linear, with a very smooth upward slope. So, I am baffled as to why the WFO does not produce better results. It seems to me that if a strategy produces an almost linear equity curve when back tested that it would at least produce a profitable result for the out-of-sample data in WFO. Any ideas on why my WFO results are so poor? Have any of you had the same experience with WFO?

Thank you in advance for any insights you can share!

Jeff R

Re: Walk Forward Optimization - Does it Work?

Posted: Sep 15 2024
by faraz
Reduce the roboustness criteria to be more realistic closer to strategy numbers.

Screenshot 2024-09-15 131130.png
(40.17 KiB) Not downloaded yet






Trading Excellence ! ! !
1stChoiceStrategy.com