Yes, I know this topic has been discussed ad nauseum, but I believe I have found a core difference between the way Live Strategies and Simulated Strategies work. This has nothing to do with PIT, or any other impacts of updated technical data. It is to do with the timing of the ranking data used.
TL;DR: Live strategies use the Ranking data from the Monday morning update that is done around 3:00AM. Simulated strategies use Ranking data from the Saturday morning update done at 10:00AM. If there are changes in rankings between the Saturday AM and Monday AM, updates then there will be differences between the Live Strategy and Simulated Strategy.
To test and verify this, I used the “Easy to Trade” Universe and the Core:Sentiment Ranking. First, looking at the ranking Systems when run on Friday Saturday and Sunday of this current week:
Friday (1/02/26) (note: this uses rank data from the previous Monday, hence the 12/27/25 date stamp)
Saturday (1/3/26)
Monday (1/5/26)
Specifically, look at #12 in the last two. Saturday AM has RY @ #12, and Monday AM has MRBK @ #12. RY has moved down to #15.
Setting up a simple strategy to buy the top 12:
Live (ordered by ranking)
As expected, MRBK is ranked at the bottom (#12)
Simulated Strategy (started on Saturday Jan 3, and run through to Jan 5, so just one day):
As can be seen, MRBK has disappeared and RY has been bought. Same day, same strategy, different outcome.
I have checked this across multiple strategies, as I could not understand why there are differences for the same dates between live and simulated strategies. This explains to me what is going and is the root cause of sometimes significant differences between Live and Simulated performance.
Simple question: Is it possible to fix this? Just from a modeling perspective, there should be a way to align Live and Simulated strategies so that they both use the same Ranking data. The most obvious solution is to have an option for Simulated strategies to use the Monday AM ranking instead of Saturday. The other would be for Live strategies to have an option to use Saturday ranking. Either way it ensures that both live and simulated using the same ranking data for trades.
Note: I understand there are potential issues here. In effect there are two sets of data for Monday - the ranking data that exists prior to the market opening, and the market data that exists after it closes. No one said being a software developer was easy 
.
If this is by design, then I hope the above provides further clarity as to why Live and Simulated strategies differ. note: the smaller the number of stocks in a strategy (many of mine have just 5) the greater the potential impact is. One of my strategies last week and this week had two completely different stocks which has a significant impact on performance between live and simulated strategies. Simulated is always better - and yes, there has been lots of discussion around why - but I believe using different ranking data to order and place trades should not be one of them.
Apologies if this has been covered in other posts. Marco, I did see you posted a response about “simulated strategies as a way to test a concept” and mentioning the weekend update, but it was buried in a discussion about PIT, European stocks, etc. and I didn’t really see it explicitly stated the Live and Simulated strategies use ranking data from different days.
Thanks,
Simon