The Invincible v1.22 · MT5 XAUUSD… AI Review | QuantTester
The Invincible v1.22 · MT5 X: return 17,121.4%, DD 31.05%, PF 3.05, 357 trades. Read the review.
Strategy Safety Grade: B (60 / 100) · Moderate Quality
Trading risk analysis result
Strategy Type: Strict Lot Discipline. Classified as a fixed-lot, high-frequency paired-entry XAUUSD M5 system whose 3.05 profit factor and 87.4% win rate are real but structurally dependent on a 0.44 payoff ratio, where each loss is ~2.3x an average win and single losses reach -$95,000, leaving a Grade B safety profile with 31.1% max drawdown that is investable only with strict position-size governance.
Darwinex 集中度压力测试 (Top 5% Windfall Dependence Test)
剔除排名前 18 笔（5%）极端盈利单后：毛利占比 28.25%，剩余净利润 $992,144.66，剩余 PF 2.19。
Data transparency & Environment
- MT4/MT5: MT5
- Symbol: XAUUSD
- Timeframe: M5
- Net profit: $1,712,144.66
- Return: 17,121.4%
- Max drawdown: 31.05%
- Total Trades: 357
- Profit Factor (PF): 3.05
Critical Algorithmic Vulnerabilities
- [Rs — Risk Stability / Loss Scale Escalation] Loss magnitude escalates in absolute terms alongside lot growth: early losses run -$997.5 to -$1,472.5, mid-sequence -$6,412.5 to -$15,532.5, and late-sequence clusters hit a hard -$95,000 ceiling (2026-09-15 19:30, 2026-11-02 19:40, and two -$95,000 exits on 2026-11-03). Relative-risk sizing kept initial deposit risk near 10% per trade (0.2 lots on $10,000), but the absolute tail loss now equals ~5.5% of equity per event — replication stability stays contingent on capital-base discipline.
- [R+ — Return Consistency / Drawdown Depth] Equity peak drawdown reaches 31.05% (e.g., from ~$693,310 on 2025-07-03 to ~$477,600 intra-February-2026 near the -$25,745 loss cluster, and from a ~$508,000 peak April 2026 to ~$268,000 in April 2026) — underwater episodes coincide with consecutive multi-lot stop-outs, not with isolated bad trades, indicating correlated loss clustering.
- [Execution Feasibility / Stop Gap] The stop_loss_evidence pattern is unbounded_tail (max_over_median 14.81; only 11.1% of losses cluster at a consistent level; max loss -$95,000 vs median loss $6,412.5): a portion of losses terminate well beyond the typical level, so a fixed -$95,000 apparent threshold must be verified as a real hard stop rather than a backtest artefact of sparse ticks on 20-lot XAUUSD positions.
- [R+ — Trade Frequency Defect] Only 0.53 trades per day with average hold of 620.7 minutes (10.3 hours) and a 66.4% long bias means the system is exposed to multi-hour overnight/weekend gap risk on 20-lot XAUUSD positions; a single Sunday-open adverse gap can exceed the observed -$95,000 stop distance and invalidate the backtest's stop hierarchy.
Score Breakdown Details
- [Base Score] All valid statements start at 50 points neutral benchmark (Grade B-)
- +15 pts [Lot Discipline]: Fixed/proportional lot execution, no adverse counter-trend doubling
- -20 pts [Poor Payoff]: Win/loss payoff ratio only 0.44 (< 0.6), losses heavily erode gains
- +5 pts [Expectancy]: Overall Profit Factor reached PF 3.05 with solid positive mathematical expectancy
- +10 pts [Organic Profit]: Removing top 5% windfall deals leaves PF 2.19 >= 1.2; robust and non-luck-driven
- -10 pts [Elevated Drawdown]: Max historical drawdown reached 31.1% (20%-35%), downside pressure elevated
- +10 pts [Sample Horizon]: Sample duration spans 669 days (>= 180 days), covering diverse regimes
- Final Strategy Safety Score: 60 / 100 (Grade B).
Recommended Next Actions
- Verify with the broker/EAs that the -$95,000 loss events represent an executed hard stop loss rather than a delayed signal exit; if the latter, install a broker-side stop and re-measure the drawdown.
- Run a stress test with realistic XAUUSD spread and swap costs at 20-lot position sizing, then compare net profit and drawdown against the tested 50-point slippage assumption.
- Compute the maximum adverse excursion distribution and 95th-percentile loss duration for the -$95,000 cluster to distinguish structural stop by design from tail under-protection.
- Re-run with explicit entry-ticket pairing to confirm the two-slot basket and that all same-timestamp exits are intentional tiered exits rather than uncontrolled multi-position exposure.
- Model a periodic profit-extraction plan (e.g., reducing lot plateau after a 15-20% equity drawdown) since the current fixed 20-lot size makes absolute tail risk grow with balance.
Q&A About This Report (Quant Due Diligence)
Why does an 87.39% win rate still produce a 0.44 payoff ratio and only Grade B (60/100)?
The rating is driven primarily by realized payoff asymmetry: average wins of $8,170.01 are dwarfed by average losses of -$18,597.77, so each loss consumes roughly 2.3 wins. With -20 pts deducted for payoff and -10 pts for the 31.05% drawdown, the system relies on hit-rate persistence rather than loss containment. The positive expectancy (PF 3.05) and organic residual PF 2.19 after excluding the top 5% trades earn back +15 pts, producing the 60/100 Grade B.
Are the -$95,000 losses consistent with an intentional hard stop or with uncontrolled holding?
The computed stop_loss_evidence reports an unbounded_tail pattern (max_over_median 14.81; only 11.1% of losses cluster at a consistent level; median loss $6,412.5 vs max loss -$95,000). Three separate -$95,000 events (2026-09-15, and two on 2026-11-03) appear at the top of the loss distribution and coincide with 20.0-lot exposure, so the terminal level is plausible as a hard stop on the final lot size — but with only 45 losing trades the sample cannot confirm consistency, and ticket-paired verification is needed before treating it as a verified hard stop.
Is the 17,121% return a genuine compounding result or a backtest artefact?
The percentage is arithmetic from the report itself (net profit $1,712,144.66 on a $10,000 base, ending balance $1,722,144.66), achieved via disciplined but aggressive compounding: the lot plateau reaches 20.0 on the final balance. No tester warning (mismatched charts, modelling quality, control points) appears in evidence_flags, and history quality is 99%. The 31.05% drawdown is real but modest relative to return; the risk conclusion is about the payoff ratio and the -$95,000 tail loss size, not about the return value itself.
What is the biggest single failure mode for a live allocator copying this strategy?
Position sizing during a losing sequence. The late-stage losses cluster at -$95,000 per event (e.g., 2026-11-02 19:40, 2026-11-03 14:30, 2026-11-03 14:55) while the balance is approximately $1.78M-$1.88M. A short sequence of three or four such events — plausible if hit rate drifts below 87% or XAUUSD gaps against a 20-lot long — would equal roughly 5.5%-16% of equity. This is why the maximum observed drawdown of 31.05% is not a worst-case bound in live trading, and why the strategy-safety grade is B rather than higher.
How is QuantTester different from asking a general-purpose AI assistant myself?
QuantTester parses MT4/MT5 reports into structured metrics and trade records before AI review. Supported calculations can be checked against the original export, guaranteeing consistent audit facts.
Which broker is recommended for trading and running MT4/MT5 strategies?
For algorithmic trading, EA execution and reliable backtesting, we strongly recommend IC Markets (Raw Spread Account) with 0.0 pip spreads.