Hello,

I am educating myself how to use Zorro in the Robotwealth course and recently stumbled over a neat illustration of the Detrend = SUFFLE parameter.

In this course the authors conclude:
"If the StartDate and EndDate of your simulation don’t use all available price data, then the shuffled price curves will have different statistical properties to the data used in your backtest because they start at a different price level, but all finish at the same level."

They then go on to illustrate that you only get the same start and endpoint of your shuffled price curve, if you include all of the available data.

However, I have been unable to reproduce this behavior myself, no matter how much data I use there are alsways curves that start and/or end differently. So my question is whether the SHUFFLE functionality has been altered in later versions and if it is possible to shuffle the data while maintaining its statistical properties, like mean, sd, trend etc.?