I am building a strategy where I check the trading signal on the daily timeframe, but I want to execute the order at 3:20 PM, before the market closes at 3:30 PM.
The challenge is:
The daily candle is not fully completed at 3:20 PM, so the final daily indicator value is not yet available.
If I use daily data for backtesting, it introduces look-ahead bias because the daily close is only known at 3:30 PM.
If I instead use intraday timeframe data (like 1-minute) to simulate the execution price at 3:20 PM, the indicator signals on intraday charts may differ from the daily indicator.
So my question is:
👉 What is the correct way to design both the backtest and the live trading logic in this case?
Specifically:
How should I calculate the daily signal so that it reflects the information actually available at 3:20 PM?
Which data should be used for execution price in backtesting?
How do I avoid mismatch between daily signals and intraday data?
Hi @7350982949 ,
We cannot fetch data only upto 3:20pm during backtesting using day candle. Instead you can consider 1-min or 5-min timeframe candles from 9:15 to 3:20pm and resample the data till 3:20pm and use it.