Gold Aurum Trend Engine
- 专家
- 版本: 13.5
- 激活: 8
Aurum Trend Engine — MT5 Expert Advisor for XAUUSD
Gold moves fast, trends hard, and punishes underprepared systems. Most off-the-shelf Expert Advisors fail on gold because they were built for forex pairs with fixed pip grids, martingale compounding, or indicators never designed for the volatility profile of XAUUSD.
Aurum Trend Engine was built for XAUUSD M1 with a layered multi-filter signal architecture that confirms trend alignment across timeframes and only fires when several independent conditions agree.
Backtest results
XAUUSD M1, 1 January 2025 to 11 August 2026, $1,000 initial deposit, 0.01 lot, 100% real ticks from Exness (139,149,511 ticks processed, history quality 100%).
| Net profit | $24,860 |
| Gross profit | $75,859 |
| Gross loss | -$50,999 |
| Profit factor | 1.49 |
| Recovery factor | 10.73 |
| Sharpe ratio | 2.17 |
| LR correlation | 0.94 |
| Max balance drawdown | $1,861 (9.51%) |
| Max equity drawdown | $2,317 (11.52%) |
| Total trades | 6,778 |
| Win rate | 37% |
| Average profit trade | $30.26 |
| Average loss trade | -$11.85 |
| Reward to risk | 2.56 : 1 |
| Average holding time | 6 hours 21 minutes |
| Profitable months | 16 of 20 |
Past backtest performance does not indicate future results. Live results will differ because of broker spread, slippage and latency.
How it works
Every layer must confirm before a trade is placed. This is why a 37% win rate is still profitable: when the system fires, the average winner is 2.56 times the average loser.
1. Higher timeframe trend direction
The EA evaluates H1 using the slope of a 50-period EMA over a rolling window. Buy trades are allowed only when the H1 bias is bullish, sell trades only when it is bearish. The timeframe is configurable through TrendTimeframe.
2. Trend strength corridor
A trend that is too weak is choppy noise; one that is too strong is overextended. The EA measures a normalized trend strength metric and allows entries only inside a corridor between MinTrendStrength (0.4) and MaxTrendStrength (1.6). This avoids both dead markets and exhausted breakouts.
3. M1 entry structure
Once higher timeframe conditions are confirmed, the EA drops to M1 for entry timing, using Bollinger Bands (20-period, 2.0 deviation) to locate the entry point relative to band structure.
4. LWMA alignment confirmation
The EA requires the fast LWMA (5-period) and slow LWMA (10-period) to be aligned in the intended trade direction. This confirms that short-term momentum is already moving the right way before entry.
5. EMA distance filter
Price must sit at least MinDistanceFromEMA (1.0 ATR) away from the 50-period EMA. This avoids entries in the low-conviction zone where price hugs the mean and whipsaws are common.
6. Strong trend boost
When an exceptionally strong trend is detected — TrendBarsRequired (5) consecutive directional bars within TrendLookback (20 bars) — the EA allows ExtraTrades (5) additional positions. This module contributes most of the steep growth phases in the equity curve. It can be disabled by setting ExtraTrades to 0.
7. Exit system
- Tiered take profit. Targets are distributed across 5 levels at 40-point spacing, so partial profit is locked in early while the position stays open for deeper moves.
- Structural stop loss. The stop follows market structure rather than a fixed pip value, widening in volatile sessions and tightening in quiet ones. Controlled by StopLossType.
- Stale trade exit. A trade open longer than StaleTradeBars (300) without reaching StaleTradeMinGain (0.5) is closed, so capital is not tied up in dead positions.
- Break-even. Optional. When enabled, the stop moves to entry once the trade reaches BreakEvenTrigger (3.0 ATR).
What the backtest charts show
Equity curve. LR correlation of 0.94 means 94% of the variance in the equity curve is explained by a straight upward line. There are no prolonged flat periods and no parabolic spikes. The steepest phase is early 2026, during the gold bull run.
MFE and MAE analysis. Correlation between profit and maximum favorable excursion is 0.88, meaning winners consistently reach deep favorable territory before the target is hit. Correlation between profit and maximum adverse excursion is 0.25, meaning losses stay shallow and contained. Correlation between MFE and MAE is 0.012, confirming winners and losers behave independently.
Time distribution. The EA is most active during London and New York sessions (08:00 to 16:00 UTC), with the strongest aggregate contribution during the 13:00 to 16:00 overlap. All weekdays show positive net contribution.
Monthly distribution. 16 of 20 months were profitable. The 4 negative months were shallow and recovered within one to two months. The most difficult stretch was the mid-2025 consolidation, when gold ranged for an extended period and trend filters reduced trade frequency.
Input parameters
Money management
| LotSize | 0.01 | Fixed lot per trade. Start with 0.01 per $500 balance. |
| MaxOpenTrades | 5 | Maximum concurrent positions, including boost trades. |
| MagicNumber | 84434000 | Change if running multiple instances on one account. |
Trade direction
| AllowBuyTrades | true | Enable long trades. |
| AllowSellTrades | true | Enable short trades. |
Entry filters
| UseTrendFilter | true | H1 trend direction gate. |
| TrendTimeframe | H1 | Timeframe for trend direction check. |
| UseTrendStrength | true | Trend strength corridor filter. |
| MinTrendStrength | 0.4 | Lower bound. Lower values allow weaker trends. |
| MaxTrendStrength | 1.6 | Upper bound. Higher values allow overextended trends. |
| MinDistanceFromEMA | 1.0 | Minimum distance from EMA in ATR units. |
| UseInsideBandFilter | false | Optional inside-band condition on H1. |
| InsideBandTimeframe | H1 | Timeframe for the inside-band check. |
| MaxDistanceToLevel | 0.0 | Maximum distance to structural level. 0 disables. |
Entry timing
| EntryFrequency | 2 | Confirmation bars required before entry. |
| MinBarsBetweenTrades | 1 | Minimum bars between entries. |
Take profit
| LevelSpacing | 40.0 | Points between consecutive TP levels. |
| TargetLevel | 5 | Number of TP levels. |
| PartialClosePercent | 0.0 | Percent closed at partial level. 0 disables. |
| PartialCloseLevel | 3 | TP level triggering partial close. |
Stop loss
| StopLossType | 4 | 1 fixed, 2 ATR, 3 swing, 4 dynamic structure. |
| StopLossTimeframe | M15 | Timeframe for structural stop calculation. |
| StaleTradeBars | 300 | Bars before a non-performing trade is closed. |
| StaleTradeMinGain | 0.5 | Minimum gain exempting a trade from stale exit. |
| UseBreakEven | false | Enable break-even stop. |
| BreakEvenTrigger | 3.0 | ATR profit at which break-even activates. |
| ProfitTakeStreak | 0 | Consecutive wins before a forced pause. 0 disables. |
| ProfitTakeTimeframe | H1 | Reference timeframe for the streak check. |
Strong trend boost
| ExtraTrades | 5 | Additional positions during strong trends. 0 disables. |
| TrendBarsRequired | 5 | Consecutive directional bars required. |
| TrendLookback | 20 | Lookback window for the streak count. |
| TrendBoostTimeframe | H1 | Timeframe for boost detection. |
Indicators (leave as default)
| BollingerPeriod | 20 |
| BollingerDeviation | 2.0 |
| FastLWMA | 5 |
| SlowLWMA | 10 |
| EMAPeriod | 50 |
| TrendSlopeBars | 20 |
| ATRPeriod | 14 |
Requirements
| Platform | MetaTrader 5 only |
| Symbol | XAUUSD |
| Timeframe | M1 |
| Minimum deposit | $300 at 0.01 lot, $1,000 recommended |
| Leverage | 1:100 minimum (1:500 used in backtest) |
| Account type | ECN or raw spread recommended |
| Spread | Under 20 points preferred |
| VPS | Recommended for 24/5 uptime |
Gold spreads widen during low-liquidity periods and major news releases. Setting a spread filter in MT5 is advisable.
Setup
- Download the EA from Navigator, Market, Purchased.
- Open an XAUUSD chart and set the timeframe to M1. This is the only chart required; the EA handles higher timeframe analysis internally.
- Enable Algo Trading in the toolbar and allow automated trading in chart properties.
- Drag the EA onto the chart.
- Set LotSize for your account. Leave the remaining parameters at default for the first run.
- Allow at least 100 trades before evaluating live performance.
Position sizing
| $300 to $500 | 0.01 |
| $1,000 | 0.01 to 0.02 |
| $2,500 | 0.03 to 0.05 |
| $5,000 | 0.05 to 0.10 |
| $10,000 | 0.10 to 0.20 |
Frequently asked questions
Does it work on other symbols?
No. Spread tolerance, ATR scaling and trend strength thresholds are calibrated for gold. Other symbols are not supported.
Does it use martingale or grid?
No. Each trade is independent with its own stop loss. MaxOpenTrades limits total exposure. The strong trend boost adds positions only when a trend is detected, never to recover a losing trade.
The win rate is 37%. Is that normal?
Yes, and it is intentional for a trend-following system. The average winner is $30.26 against an average loser of $11.85. A 37% win rate at 2.56:1 reward to risk produces a profit factor of 1.49.
What is the minimum account size?
$300 at 0.01 lot works. $1,000 gives a more comfortable buffer against the drawdown observed in the backtest.
Do I need a VPS?
For M1 trading, yes. A machine that sleeps or loses connection will miss trades and produce results that differ from the backtest.
Can I run several instances on one account?
Yes, with a different MagicNumber for each. MaxOpenTrades applies per instance, so watch total exposure across all of them.
Is it suitable for prop firm challenges?
The 9.51% balance drawdown may fit firms with a 10% limit, but daily drawdown rules and news restrictions vary. Test on demo first. Setting ExtraTrades to 0 gives the most conservative profile.
What happens during major news?
There is no built-in news filter. The trend strength and EMA distance filters naturally reduce activity during volatility spikes. You can disable the EA manually before scheduled high-impact releases.
Will there be updates?
Yes. All purchased copies receive free updates through the MQL5 Market system.
How it differs from common EA types
Martingale and grid systems show smooth equity curves until one sufficiently long adverse move wipes the account. This EA has neither. Every trade carries its own stop.
Scalpers target 2 to 5 pips and are highly sensitive to spread and slippage. This EA targets multi-level take profit with an average holding time above six hours, making it far less sensitive to execution quality.
Single-indicator EAs use one signal with fixed stops and no market context. This EA requires several independent filters across timeframes to agree before entry.
Risk disclaimer
Trading gold involves substantial risk of loss and is not suitable for all investors. Past backtest performance is not indicative of future results. Live trading results will differ from backtests because of broker spread, slippage, latency and market conditions that cannot be reproduced in simulation.
Only trade with capital you can afford to lose. Start on demo. Use conservative lot sizing. This EA is a trading tool, not financial advice. You are responsible for your own trading decisions.
