And detrending balance curves just means doing the same with the balance curve. For detrending, the curves are simply tilted until their start and end points have the same level.

1) Produce balance curves of all strategy variants, using detrended trade results and no trade costs. Note down the profit P of the best strategy.

2) Detrend all balance curves by subtracting the mean return per bar (not to be confused with detrending the trade results!). This way you get a series of curves with the same characteristics of the tested systems, but zero profit.

What exactly do “detrended trade results” and “mean return per bar” mean?

Does “detrended trade results” mean subtracting the final value of the balance from each bar in the equity curve? So for example, if the strategy variant has $1,000 of profit at the end, you subtract $1,000 from every bar in the equity curve?

Does “subtracting the mean return per bar” mean summation of all bar returns divided by n bars (average return per bar) and subtracting that value from every bar?

Thank you.

Sorry if this is confusing but i really need a way to make the bootstrap test and this seems like a great way. I really want this to work!
Regards Sebastian

White’s Reality Check will probably not work here unless you’ve kept a record of all variants that you tested. If it’s portfolio rotation, first do a test with detrended price curves, then compare with a series of tests with detrended and shuffled price curves. This way you’ll get a z-score of the system. If the z-score is low enough, like 5%, you have a winner.

Can you recommend a reality check for this kind of a system?

Thanks,
Andrew