Back to Community
Grade system for backtesting key figures

Hi everyone,

I want to backtest my trading algorithm. The algorithm trades on a low timeframe with 400 “unique” trades. With unique I mean that it can place 1 or more of those 400 trades given different market conditions, and not that it is just placing 400 trades. Before I go and test the algorithm, I want to backtest the algorithm with the key figures like sharp ratio, drawdown periods, annual return, etc.

I want to group the 400 unique trades in “groups” and I already have a way to do so. I made 40 groups of 10 unique trades per group.
So, when I am going to backtest the trades I want to backtest each “unique” trade, each group and the entire trading algorithm (so the 400 unique trades combined). But that means that I am going to have 400 tearsheets when I am backtesting. If I want to manually check these sheets it would take me a lot of time…

So, I figured that I could give each tearsheet a grade based on the key figures in the sheet. I basically want to give every key figure in the tearsheet a score so that the tear sheet gets a grade and with that I can easily program a model which automatically picks the best “unique” and “group” trades. (Because better trades & groups would give higher overall grades)

But I have no clue how to give the key figures a score. Can anyone help me with this? Or does someone have a better way to test the “unique” and “group” trades?

So say I have annual return, sharpe ratio, drawdown period and gross leverage as key figures. Would each key figure have an equal importance of 25%? And how could I for example grade the sharpe ratio, I know the minimum is 0, but what is the maximum, is it 2,3,4,etc?

I hope you guys can help me out! Thank you!