Checking Futures Data Differences Before Trusting Backtests
Summary
A Chinese forum exchange discusses large differences between futures data from Xuntouyan and TQSDK and the resulting disagreement in strategy backtests. The original poster reports discovering that the unusually high returns from one Xuntouyan test came from incorrect data, while another participant says their installation works normally and suggests updating the vn.py connector and downloading the data again through its graphical interface.
The exchange offers a practical reminder to validate historical data before interpreting backtest performance. However, it does not provide the example data, screenshots, identified data error, or a comparison against live trading. It therefore cannot establish which provider is generally more accurate or that either dataset is closer to real execution. The advice is limited to checking the specific installation and dataset involved.
Key ideas
- Different futures data sources can produce substantially different backtest results.
- The original poster attributes an unusually strong result to incorrect historical data.
- Another participant recommends updating the connector and downloading the data again.
- The discussion provides no evidence establishing which source is more representative of live trading.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.