Hi @Ramneet, thanks for getting back.
as backtesting is a tool to see the past performance of your current logic.
This makes me wonder, what else do we use as a metric of performance apart from P/L. Because even if we did take points gained / lost as metric instead of P/L, say "X strategy successfully made on an average +8 points in BN for every trade it took" - for that AlgoTest needs to show it as points, but it doesn't, it shows as points * lotsize at the time which is P/L. (refer image below)
Max profit, max loss, max DD, returns/maxDD, R:R everyhing here is denoted based on P/L, which is dependent on the lotsize. As a matter of fact, every single parameter mentioned in the stats that we get post backtest is highly coupled with the lotsize.
I hope my point here comes across clearly, static lotsizes are fine if we display outcomes in both points AND points * lotsize, but if we're solely choosing to show it as points * lotsizes, I believe it is important to have dynamic lot sizes to be accurate here - or we let users know that whatever backtest data comes through for banknifty after June 30, 2023 is accurate, and same is the case for nifty after June 25, 2021, anything before this might be inaccurate and not very dependable.
While I do understand the intricacies behind fiddling with the backtesting engine to put different math for different timelines, I believe it would be in AlgoTest's best interest to take this integration up to be the best backtesting tool out there. Like I mentioned in the very first comment in this thread, I frickin love you guys, but please make the variable lot size happen, that would be pretty awesome!
