walk forward

[renkoben]

I've done "walk forward" tests on Two of my entries for a strategy (in development)
optimization of "take profit at" comes in about 11 ticks for each. This does limit profits but...

Here I'm trying to ballpark how successful EACH entry is, and to see if it fails in the WF. "What do you think is going to happen?"

optimization (before WF) of these entries result in about 68% profitable and profit factor of 1.40 or so.

THEN THE WALK FORWARD.........results?

% profitable drops to 54% and Profit factor drops to .84 (Each entry loses money.)

the back test periods or "out of sample" are 45 days long, (I have 7 of them) "In sample" is 90 days

AT THIS STAGE. What should I do? Who can point me in the right direction?

Thank You:

[renkoben]

Post Script

I have 4 parameters (I think)

[NinjaTrader_Bertrand]

renkoben, so to understand better, the optimized segments are profitable but the actual walk forward testing parts are not as good as expected? This normally shows a strategy that does not generalize well, i.e. which is overfitted by your optimization to the exact seen data only.