AntiOverfit Audit: Goldwave EA MT5 on XAUUSD
Goldwave EA MT5 was evaluated on XAUUSD in an AntiOverfit Audit generated on April 15, 2026. The reported robustness score was 91.7/100, which maps to the public grade A — High Robustness. The test used 100 synthetic worlds, with all 100 remaining valid for analysis. On this result alone, the main public takeaway is straightforward: the strategy’s observed behavior remains broadly stable when the original historical path is replaced with multiple plausible alternatives from the same market.
That is an important distinction. An AntiOverfit Audit is not a full review of the EA, not a verdict on commercial quality, and not a forecast of future profitability. Its purpose is narrower and more technical: it examines how dependent the backtest outcome is on the exact historical path that happened to occur, and how well the same behavior holds up when that path is varied in realistic ways. In other words, the question is not whether the original run looked good, but whether it still looks structurally coherent once the market path is changed.
Audit Setup
The audited product was Goldwave EA MT5 on XAUUSD. The report was generated from the file Goldwave EA MT5.xml. The analysis used 100 synthetic worlds, and all 100 were accepted as valid. The original run showed a maximum drawdown of 4.4%, which falls into the mild drawdown context range in the report.
The headline score of 91.7 places this result firmly in the highest public bracket. That grade matters because it summarizes the overall robustness picture into a single public label, but the more useful reading comes from how the score was built. In this case, the audit combines a very strong stability profile with a clear but manageable note of path dependence in some original-run relationships.
What the AntiOverfit Audit Measures
The AntiOverfit Audit compares the EA’s behavior on the base historical run against many plausible synthetic market paths derived from the same market. This is not just a reshuffling of trade outcomes or a cosmetic stress test. The underlying goal is to see whether the system’s observed structure remains recognizable when the order and shape of market movement are varied within a plausible range. That makes the audit a robustness test under path variation.
This matters because many backtests look convincing on the exact path history delivered, yet degrade meaningfully when the route to those outcomes changes. A robust result is one where the original run does not look isolated or overly special relative to the synthetic distribution. A less robust result is one where the historical run appears unusually favored by the specific path that occurred. The audit therefore focuses on representativeness, dispersion, downside control, survival across worlds, and the alignment between the original profile and the synthetic profile.
Overall Reading of the Result
For Goldwave EA MT5 on XAUUSD, the broad reading is positive. The synthetic distribution appears controlled, survival is complete, downside behavior remains well contained, and the activity footprint is highly stable across alternative paths. Those are strong signs that the original test is not collapsing when the market path is altered.
At the same time, the report does not describe a perfectly path-neutral system. The BIAS score of 68/100 is explicitly labeled as elevated path dependence. That means the original run still appears somewhat favored relative to the synthetic distribution, even though the broader structure remains intact. This is a meaningful nuance. The right interpretation is not that the strategy fails the audit, but that its robustness is high overall while some headline advantages of the original run should be read with caution.
Key Interpretive Points
-
High overall robustness is supported by multiple independent signals. The final score of 91.7 is not being carried by one isolated metric. Consistency scored 96/100, Downside scored 99.1/100, and Viability reached 100.0%. That combination suggests the system’s behavior remains stable across plausible alternative paths, bad-tail outcomes stay controlled, and none of the synthetic worlds dropped out of acceptability. When those three areas are all this strong at once, the audit is pointing to structural resilience rather than a lucky single-path impression.
-
The main reservation is path dependence, not instability. The report’s own natural interpretation states that the original backtest still looks stronger than the synthetic distribution. That is consistent with the BIAS reading of elevated path dependence. In practical terms, the original historical run may still flatter the system somewhat relative to what is typical across alternative paths. That does not cancel the strong result, but it does change how the original backtest should be read: as broadly representative, yet somewhat optimistic in places rather than perfectly central.
-
The downside profile is one of the strongest parts of the audit. The original drawdown was only 4.4%, and the report classifies that as mild in contextual terms. More importantly, robustness is supported by the synthetic drawdown behavior as well. Stress DD was 5.0%, with the original drawdown staying below that adverse-tail reference, and DD Stability was 52%, indicating moderate rather than extreme dispersion in downside outcomes. The message here is that risk does move across synthetic worlds, but not in a chaotic or uncontrolled way.
-
Trading activity is exceptionally stable across synthetic paths. The original trade count was 801, and the activity distribution appears unusually tight. Trade Stability was 0%, with P10 and P90 both at 801 trades. Collapse Rate was 0.0%, using a collapse threshold below 401 trades. This is a very constructive sign because it suggests that the EA is not dependent on the exact historical path merely to remain active. The execution footprint survives the path changes almost perfectly, which strengthens the case that the strategy logic is engaging the market consistently rather than opportunistically.
-
Some performance metrics remain sensitive even within a strong overall audit. Profit Factor at 5.49 and Recovery Factor at 28.83 are both described as path-sensitive. In both cases, the original result remains plausible but not central, sitting outside the inner band while staying within the broader outer range. That is a manageable form of sensitivity, not a breakdown. The more notable exception is Expected Payoff at 183.23, labeled fragile, where the original value sits outside P10-P90 on the unfavorable side. Interestingly, that means the historical run underperformed the typical synthetic outcome on this measure, which cuts against the simplistic idea that every original metric was overly flattering. It also shows why robustness analysis is more informative than relying on one headline number.
-
The profile alignment is strong even though one gap stands out. The report labels the overall profile as ALIGNED. That means the original shape across the main robustness axes remains close to the typical synthetic profile, with no major structural deviations. The largest reported gap appears in Recovery Factor, where the original value of 28.83 sits below a typical synthetic value of 34.63. That does not undermine the audit result, but it is a useful reminder that even highly robust systems can show uneven sensitivity across individual dimensions.
Professional Conclusion
On the evidence provided, Goldwave EA MT5 on XAUUSD achieved a strong AntiOverfit Audit outcome and should be published under the public label A — High Robustness. The core reason is not simply that the original backtest looked strong, but that its behavior remained broadly stable across 100 plausible synthetic market paths. The synthetic distribution stayed controlled, downside remained well managed, survival was complete, and trading activity was strikingly consistent across alternative worlds.
The main caution is equally clear and should remain part of the public reading. The original run still appears somewhat favored relative to the synthetic distribution, with the report explicitly flagging elevated path dependence in the bias component. That means the backtest should not be treated as a perfectly central summary of all plausible path outcomes. Even so, the overall structure holds up well, and the audit points to a result that is robust in form rather than narrowly dependent on one lucky historical route.
External robustness audit. Not a promise of future results.
