How to reconcile Autotrade data with Hypothetical Monthly Returns

I have read the topic “Why “hypothetical” results and performance?” but still have a question: are Hypothetical Monthly Returns built from the AutoTrade data? If yes, I understand that the Hypothetical Monthly Returns would have to be averaged from that; thus, while they are ‘hypothetical,’ they are nonetheless a “darned decent representation of reality.” Do I have this right?

Reality is on your (or my) account, everything else is hypothetical. Even average results can be off from that reality.

Simple example - two auto-traders, one always have -5$ PL per trade due to execution and another one always have +5$ PL per trade due to execution. Strategy will be reporting 0$ PL (average) per trade, but one auto-trader will be losing each trade and another one will be winning.

Yes, I get that. But can the “Hypothetical” returns be portrayed as kind of the best estimation of what actual returns would have been?

I would say that is a good description. The estimate is much better when there are autotrade subscribers, where the average fill prices are used instead of simulated fills.