Quantum Bitcoin V3.2 EA… AI Review | QuantTester
Quantum Bitcoin V3.2 EA · Me: return 26,403.9%, DD 55.01%, PF 3.44, 449 trades. Read the review.
Strategy Safety Grade: C (35 / 100) · High Risk Alert
Trading risk analysis result
Classified as Hybrid (Tiered Basket + Trend Following): BTCUSD H1 backtest produced a 3.44 profit factor and 26,403.9% net return over 641 days, but this is achieved exclusively through synchronized multi-tier basket entries with 12% lot step-ups during adverse floating drawdowns and zero hard stop-losses, and the realised loss tail (max loss $72,943.58 = 11.97x median loss) plus a 55.01% maximum equity drawdown demonstrate severe asymmetric downside, leaving the strategy inadmissible for live capital at institutional risk limits.
Darwinex 集中度压力测试 (Top 5% Windfall Dependence Test)
剔除排名前 23 笔（5%）极端盈利单后：毛利占比 38.74%，剩余净利润 $1,198,927.28，剩余 PF 2.11。
Data transparency & Environment
- MT4/MT5: MetaTrader 5
- Symbol: BTCUSD
- Timeframe: H1 (2025.01.03 - 202
- Net profit: $2,640,390.26
- Return: 26,403.9%
- Max drawdown: 55.01%
- Total Trades: 449
- Profit Factor (PF): 3.44
Critical Algorithmic Vulnerabilities
- [Risk Stability Breakdown / Rs] Loss Escalation and Stop Drift: Early-sequence losses cluster near -$17,309.73/-$20,591.62, while later baskets produce -$49,275.23 (ticket 470), -$72,943.58 (ticket 491) and a simultaneous 7-order floating basket (max_sim_loss $258,927.36); loss magnitude expands roughly 3-4x across the sample, undermining replication stability.
- [Momentum Decay / D-Score] Basket-Proportional Drawdown Concentration: The equity curve compounds from $10,000 to $2,650,390.26, but 55.01% peak-to-trough erosion occurs without any hard risk ceiling; balance fell from $1,747,650.78 on 2026-01-14 into a deep underwater episode driven by six simultaneous sell tickets (491-496, -$283,801.84 aggregate), producing a broken quote-shape profile for allocators.
- [Execution Feasibility / Capacity] Untested Friction on a 0.01–0.03 Lot BTCUSD Ladder: Position size is tiny relative to a multi-million-dollar balance, so USD returns are not scalable without increasing lot size, which would proportionally amplify the same unprotected floating drawdown; real spread widening and swap on week-long holds (tickets 491-497 held 2025-11-12 to 2026-01-14) are unverified.
Score Breakdown Details
- [Base Score] All valid statements start at 50 points neutral benchmark (Grade B-)
- +5 pts [Order Basket]: Controlled multi-entry execution in same minute; layered basket architecture
- -25 pts [No Hard SL]: No disciplined hard stop-loss; holding unhedged drawdown to severe depths
- +5 pts [Expectancy]: Overall Profit Factor reached PF 3.44 with solid positive mathematical expectancy
- +10 pts [Organic Profit]: Removing top 5% windfall deals leaves PF 2.11 >= 1.2; robust and non-luck-driven
- -20 pts [Extreme Drawdown]: Max historical drawdown reached 55.0% (> 35%), weak defensive execution
- +10 pts [Sample Horizon]: Sample duration spans 641 days (>= 180 days), covering diverse regimes
- Final Strategy Safety Score: 35 / 100 (Grade C).
Recommended Next Actions
- Model the worst simultaneous basket (tickets 491-497) as a stress case: recompute floating equity at 1x, 1.5x, and 2x adverse extension to determine margin depletion thresholds under plausible leverage.
- Request the EA's basket-exit and tier-deployment logic documentation to verify whether tier spacing is price-based (grid-like) or time-based, and whether any negative-close cutoff exists.
- Run a walk-forward or out-of-sample segment (e.g. hold-out 20% of the 641-day window) to confirm PF 3.44 is not period-specific.
- Re-run with realistic commission and BTCUSD swap applied to multi-day holds to quantify net expectancy erosion.
- Collect a minimum 90-day demo/live forward record before considering capital allocation, monitoring maximum floating basket exposure as the primary control.
Q&A About This Report (Quant Due Diligence)
The strategy shows a 3.44 profit factor and 26,403.9% return — why is it only Grade C?
The grade reflects the EA's own internal downside structure, not return magnitude. It carries no hard stop-loss (stop_loss_evidence pattern = unbounded_tail, max_over_median 11.97) and holds 7-order adverse baskets until reversal, producing a 55.01% maximum equity drawdown on a $10,000 base. The single largest loss of -$72,943.58 (ticket 491) is ~3x the typical loss scale, showing that returns are earned by surviving deep unprotected floating exposure rather than by bounded risk.
Is the profit concentrated in a few outlier trades?
Partially, but not fatally. The top 5% of trades (23 of 449) account for 38.74% of gross profit; excluding them leaves a still-positive net profit of $1,198,927.28 at PF 2.11. The top 1% (5 trades) contribute 15.67%. A residual PF above 2.0 means the edge is not purely windfall-driven, yet the largest single win of +$247,224.06 came from a 0.03-lot tier deep inside an adverse sell basket — the expected fat-tail payoff of the tiered design.
What exactly is the position-management pattern, and is it martingale?
The rows show staggered same-direction entries with 0.01 → 0.02 → 0.03 lot steps (e.g. tickets 491-497 opened across 2025-11-12 to 2025-11-18, all closed together on 2026-01-14). This mirrors the configured InpLotsMultiplier of 1.2, far below exponential doubling. It is characterised here as a Hybrid (Tiered Basket + Trend Following) structure — tiered accumulation against adverse moves with recovery exits — a severe risk pattern but not a confirmed geometric martingale under the four-condition test.
How does the strategy behave when a basket does not recover?
The dataset does not contain a terminal unresolved basket: every deep drawdown sequence (e.g. tickets 491-497 totalling -$283,801.84 across six losing tiers, and tickets 470-473 at -$118,296.95) eventually closed with a large recovery tier in profit. The realised loss tail reached -$72,943.58, and peak simultaneous negative exposure was $258,927.36 across 7 orders. How the system performs under a genuinely one-directional move that never retraces remains unverified and is the decisive live-capital unknown.
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.