Someone shed light to me on Walk Forward Optimization mixed numbers.

[olingerc]

Ok here is my scenario. I optimize my strategy using the optimizer tab for the past 6 months (strategy has anywhere from 4-6 variables). It spits out a value that is the most optimal for those 6 months lets say it is $14,000. Now taking these optimized settings and applying it to the walk forward (not imputing ranges for my variables just using walk forward not the "optimization walk forward" with my optimized settings based on my back test). My results show a increased value lets say $16,000. If I am testing the same historical data and none of the variables are being changed how is the walk forward yielding more? Is walk forward adding in data that isn't really there?

Now I know if I use the actual walk forward optimizer it will optimize a pre determined amount of data lets say 14 days and then apply those settings to a test period lets say 7 days. I understand that it will optimize my variables for the 14 days and apply those values to the following 7 day period and spit out a value.

But why when I change no settings and don't have it try to optimize anything it will show a increased value as the end result?

Now if I change the time periods from the 28 to 28 to 14 to 7 or 21 and 14 the end number does change. I guess my question is why isn't my end result with walk forward using the settings I used for optimization not exactly the same as a back test with the optimized settings?

[NinjaTrader_RyanM1]

Hello olingerc,

This is likely due to how session templates affect backtesting results:

When using session templates that are 24 hrs with times cross multiple days (e.g. FOREX session template), it is expected that doing things like backtests will include trades from seemingly the date before and the date after. This is because those dates are required to get the full trading session as defined by the session template.

When running walk forwards with the FOREX session template (or similar), each walk forward period may double count trades on "transition" dates. This is a limitation.

Example scenario:

* First period is 1/29 to 2/25. Second period is 2/26 to 3/25.
* First period counts trades from 1/28 to 2/26.
* Second period counts trades from 2/25 to 3/26.
* There is an overlap of two days of 2/25 and 2/26 being double counted between each walk forward period.

[olingerc]

Ok if there is overlap on certain days then that would explain the added money. It would explain that the smaller the optimization period and test period are the greater the added money.

[olingerc]

Even though in my strategy it will only initiate a trade between 9:55am and 4:05pm?

[NinjaTrader_RyanM1]

Not sure what you may be seeing. Can you check if you notice an issue using SampleMACrossOver. If you notice anything, please send a couple screenshots of your optimizer settings - The settings made before running a test.

The screenshots that we need are the walk forward test, and the standard optimizer test. We will then run the same tests here and let you know what occurs.