Closed CharlesJ-ABu closed 7 months ago
Some extra thought: I personally understand this as lookahead bias as the calculation of indicators could vary based on different time range. I'm now working on a isolated algo method to calculate the indicators.
Checked the project and find that this lookahead bias is solved actually, I used out-dated strategy...
I find it strange that backtesting time range can affect the backtesting result... I tested the strategy within 20240101-20240105 and 20240101-20240107, and for a specific trade, the exit time and profit varied:
![image](https://github.com/mkajnar/HighProfitStrategy/assets/18326516/9b59b4c0-cc44-4e11-8a39-090c557286f5)
I'm wondering if this issue is related to the capacity of my server or it's originated from FREQTRADE, or we need to modify the strategy?