How to avoid Backtesting biases
Hello everyone, I need some advices to avoid Backtesting biases or misleading
some of old traders don't trust Expert advisors even if they see it's giant results of backtest because some actions in backtesting are not the same of real trading (even if it's every tick)
What about your opinions?
Hello everyone, I need some advices to avoid Backtesting biases or misleading
some of old traders don't trust Expert advisors even if they see it's giant results of backtest because some actions in backtesting are not the same of real trading (even if it's every tick)
What about your opinions?
If your tick data is completely reliable and certain date ranges are not bypassed in your expert during the testing phase, you will get real results.
If you have extensive enough back-testing data you can be sure to not be overfitting.
If you have test results on over 10k+ bars of your operational timeframe AND with 200 entries+ (one entry per 50 bars) I would say it proves not to be overfitted since you will be facing far too many setups/circumstances to fit to. Extend the test long enough to meet both criteria. 1500 bars (1500 days = 4 years) for the daily will do if that is your operational TF. Operational timeframe is the one you use on your indicators (not possible high or low reference timeframes).
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Hello everyone, I need some advices to avoid Backtesting biases or misleading
some of old traders don't trust Expert advisors even if they see it's giant results of backtest because some actions in backtesting are not the same of real trading (even if it's every tick)
What about your opinions?