Akali's robustness on XAUUSD: a cautious reading of the synthetic test

The overall reading of the report is of acceptable robustness , with a score of 70.1 out of 100. This places the result in a reasonable range, although with noticeable sensitivity when the market deviates from the original historical trend . In other words, some of the behavior is preserved, but not completely uniformly.

Test setup

In this analysis, the EA Akali was evaluated on the XAUUSD symbol, using 100 synthetic worlds , all of which were considered valid for the final result.

Temporality: M1

Analysis range: 2020.03.01-2026.03.01

It's important to clarify the methodological approach: this test doesn't determine whether the Expert Advisor (EA) is good or bad , it doesn't predict future returns , and it doesn't provide a definitive assessment of the system . Its objective is to measure the extent to which the behavior observed in the original historical data holds true when the market follows different but plausible trajectories .

Overall result

The verdict aligns with a balanced interpretation: acceptable robustness, but not solid . There is sufficient support to rule out that the original result was purely accidental, but uniform stability is not observed outside of the original historical data .

Two key metrics help to understand this:

• Viability: 100% → the system survives in all synthetic scenarios
• Consistency: 59/100 → reasonably stable behavior, but far from robust

This defines a system that is resilient , but does not always maintain the same quality of results .

Four key points that really matter

• Dependence on the historical path.
  The Bias block scores 77/100 and is labeled as high path dependence . This implies that the original historical data is relatively favorable compared to the synthetic scenarios . It doesn't invalidate the backtest, but it does indicate that some of the performance depends on the specific path of the past market .
• The execution is more stable than the result.
  On an operational level, the system demonstrates robustness. The number of trades, at 1387 , is representative . Trade stability is 6% , indicating low dispersion, and the collapse rate is 0.0% . This suggests that the EA continues to operate similarly even in alternative markets .
• The risk is clearly more sensitive.
  The original drawdown is low ( 4.8% ), but the median synthetic drawdown rises to 8.8% , and the Stress DD reaches 19.0% . Furthermore, the DD Stability is 167% , reflecting a wide dispersion. This indicates that the risk in the original backtest is more optimistic than that observed in alternative scenarios .
• The quality of trade products is not stable.
  The Profit Factor of 3.98 appears to be representative , but both Expected Payoff and Recovery Factor are labeled as fragile . The system continues to operate, but the quality of the return per trade deteriorates when the market changes .

How does the overall profile fit?

The Profile block summarizes the situation as a partial mismatch . There is a moderate alignment between the original and synthetic profiles, but not a complete one.

The biggest difference lies in the risk: the original drawdown is 4.8% , while the typical synthetic value is 8.8% . This reinforces the idea that the original backtest is plausible, but somewhat optimistic compared to other market performances.

Conclusion

The most reasonable interpretation of Akali in this analysis is that it exhibits acceptable robustness with caution . The system demonstrates operational continuity, complete survival in synthetic scenarios, and a sufficient basis to avoid considering the original result as random.

However, it also exhibits dependence on the historical path , greater exposure to risk outside of the historical path, and fragility in key performance metrics .

In summary: some of the behavior remains outside the original historical pattern, but with a still visible dependence on the past .

And it's important to keep in mind the scope of the test: it doesn't validate the system globally or confirm future profitability . Its function is solely to evaluate consistency under plausible market variations.