I ask grok ai regarding the above problem, it gives the follwing answer, is it correct?

In Zorro's [Test] mode (after WFO training is completed), the Net Profit statistics in the WFO Cycles section (Best / Worst / Avg / StdDev) are calculated based on each Out-of-Sample (OOS) test period (i.e., each WFO cycle’s test segment) being treated as an independent slice for performance evaluation. Regarding how floating profit/loss (unrealized P/L) from open trades is handled:
1. Does the Net Profit of the first test period include the floating profit of open trades?
Yes, it does include — but only the floating profit/loss of positions that are still open at the end of that period (unrealized P/L).

When Zorro computes the Net Profit for each OOS test cycle, it includes the current unrealized profit/loss of all open positions as of the end of that cycle.
This means: if there are any open trades at the end of the first OOS test period, their floating profit/loss (from entry price to the closing price of that cycle) will be added to this cycle’s Net Profit (even if those trades are only closed in later cycles).
Reason: Each OOS segment is treated as a “standalone statistical unit”. Therefore, the account equity at the end of the segment (including the current value of open trades) is used to calculate the “realized + unrealized” net profit for that segment.

2. If a trade carries over from the first test period to the second test period, in which period is its profit ultimately counted?
The profit is split across periods (part in the period where it was held, part in the period where it was closed).

From entry until the end of the first period: The unrealized P/L of this trade (price change from entry to the end price of the first cycle) is included in the first cycle’s Net Profit (as part of the unrealized portion).
From the start of the second period until exit: The price change during the second period (from the start price of the second cycle to the exit price) is counted in the second cycle’s Net Profit (as realized profit/loss when the trade is closed).
Summary: For a transitional trade that spans multiple cycles, its total realized profit will be split and attributed to different cycles’ Net Profit:
Previous cycle(s): only the unrealized portion up to the end of each cycle.
Final cycle (where it closes): the realized portion from the start of that cycle to exit.


This split means that Best / Worst / Avg / StdDev calculations can be influenced by transitional open positions, but each cycle’s Net Profit still reflects the “equity change during that window” (including the current value of open trades at the end).
Why is it handled this way?

Zorro wants the WFO Cycles statistics to reflect “how the strategy would have performed if you only looked at this time window” (similar to rolling-window equity change).
If unrealized P/L at the end of a cycle were not included, the statistics for many segments would be severely distorted — especially for strategies with longer holding periods (swing or position trading).
However, this approach also introduces “boundary effects”: the same trade’s P/L being split across cycles can make Avg / StdDev appear more volatile (particularly when StdDev is already high).