Backtesting Dynamic Optimization: Optimal vs. Optimized Returns

[73ze]

Hi Everyone,

Does anyone know if there is currently a way to plot P&L for the optimal vs. optimized positions with DO, please?

Since, within the DO code, optimal positions are obtained from system.optimisedPositions.portfolio_stage.get_position_contracts_for_relevant_date(), I looked at system.optimisedPositions.portfolio_stage.pandl_across_systems().percent.curve() but that is not it. One of Rob's blogs shows how to plot optimal vs. optimized Risk but I don't see how to do this for P&L.

[robcarver17]

system.accounts.portfolio() unrounded
system.accounts.optimised_portfolio() DO

[73ze]

Thank you so much for taking the time.

I had actually tried this first but, the second graph making no sense to me, assumed I was missing something.

This is what I get (with a capital of 1e6 GBP and no other change from a fresh PST git clone as far as I am aware):

system.accounts.optimised_portfolio().percent.curve().plot():

system.accounts.portfolio().percent.curve().plot():

[robcarver17]

This has been discussed before. You have an expensive instrument which is removed by DO. Please read the previous discussions before asking the same question again.

#570