Why I get different results on the same backtest?

New comment
 
I am using EA Builder and I get different results for backtests with the same settings. Why is that happening?
 
William Roeder #:

Simple code error - *newbie* - Expert Advisors and Automated Trading - MQL5 programming forum #4 (2023)

Can you help me (for money, obviously) to make EA Builder usable for me? I set up the EA and it ran perfect for 3 years, then went berserk. If I do now the same backtest I did 2 days ago it shows me sometimes different results.

That one issue. The second one is if I can make it work, for that I need someone who knows also trade, not for MQL5. 

 
Getting different results on the same backtest can occur due to several factors, including:

1. **Randomness**: Some trading strategies incorporate elements of randomness, such as stochastic processes or random sampling. Even with the same input parameters, random variations in the market or the execution of the strategy can lead to different outcomes.

2. **Data Quality**: The quality and accuracy of the data used for backtesting can significantly impact the results. Variations or inaccuracies in historical data, such as missing or erroneous data points, can lead to differences in backtest results.

3. **Execution Assumptions**: Backtesting often relies on certain assumptions about order execution, slippage, and transaction costs. Variations in these assumptions or changes in market conditions can lead to different outcomes when executing the same strategy.

4. **Parameter Sensitivity**: Many trading strategies depend on parameter values that are optimized or calibrated based on historical data. Small changes in these parameter values can lead to different trading signals and performance outcomes.

5. **Overfitting**: Overfitting occurs when a trading strategy is excessively tuned to historical data, capturing noise or random fluctuations rather than genuine market patterns. As a result, the strategy may perform well on historical data but fail to generalize to new market conditions.

6. **Market Dynamics**: Financial markets are complex and dynamic systems influenced by various factors such as economic indicators, geopolitical events, and investor sentiment. Even subtle changes in market dynamics can lead to different outcomes for the same trading strategy.

7. **Model Complexity**: Complex trading models or algorithms may exhibit non-linear behavior or sensitivity to small changes in input parameters. As a result, slight variations in market conditions or input data can lead to divergent outcomes.

To mitigate the impact of these factors and improve the robustness of backtesting results, practitioners often employ techniques such as sensitivity analysis, stress testing, and out-of-sample testing. Additionally, using high-quality data, realistic execution assumptions, and avoiding overfitting can help reduce the likelihood of getting significantly different results on the same backtest.
 
Buza20:
I am using EA Builder and I get different results for backtests with the same settings. Why is that happening?

The historical data from your broker is bad. It's the same for many brokers. You can use Tickstory to download accurate data from Ducaskopy. When you download and test EA's you will see the real story. 

New comment