Quant Sniper EURUSD
- Experts
-
Freddy Rafael Magallanes Ruiz
Welcome to my official MQL5 profile!
I am a Quantitative Researcher and Algorithmic Trading Systems Engineer dedicated to the intersection of market microstructure, statistical arbitrage, and Deep Reinforcement Learning (DRL). - Version: 16.0
- Activations: 5
Quant Sniper AI — Institutional Deep RL (PPO V16 LSTM)
Quant Sniper AI — Institutional Deep Reinforcement Learning Expert Advisor (LSTM)
Quant Sniper AI is an institutional-grade algorithmic trading system designed specifically for EUR/USD on the M15 timeframe. Unlike typical commercial Expert Advisors that rely on outdated indicators or hazardous money management, Quant Sniper AI is powered by an autonomous Deep Reinforcement Learning (DRL) agent based on the PPO (Proximal Policy Optimization) algorithm coupled with a 64-unit Recurrent LSTM (Long Short-Term Memory) neural network.
With native sequential temporal memory, the model continuously decodes market microstructure, institutional order flow, and volatility cycles in real time.
🛡️ Strict Capital Preservation (100% Anti-Martingale & Anti-Grid)
Over 90% of retail EAs eventually blow trading accounts due to unbounded risk exposure. Quant Sniper AI is engineered with rigorous risk discipline:
- ❌ Zero Martingale: We never double down or increase lot sizes after a losing trade.
- ❌ Zero Grid: No accumulation of losing positions against market momentum.
- ❌ Zero Uncontrolled Hedging.
- ✅ Dynamic Stop Loss on every single position, calculated automatically based on real Average True Range (ATR) market volatility.
- ✅ 100% Prop Firm Ready: Fully compliant with daily loss limits and maximum drawdown rules of leading proprietary trading firms (FTMO, Apex, Topstep, FundedNext, etc.).
⚡ Key Technological Advantages
- Native ONNX Inference Inside MT5 Kernel:
The full neural network is compiled directly into the .ex5 executable. It requires no external DLLs, no third-party web servers, and no external Python bridges. Decision inference executes directly in memory in under 1 millisecond. - Macro Quant Gate (Triple-Layer Validation):
The AI cannot enter trades on raw impulse. Every signal must pass a quantitative macro filter:- Trend Alignment: Structural Exponential Moving Average ($EMA_{20} / EMA_{50}$) confirmation.
- Volume & Volatility: ADX filter ($\ge 22.0$) to avoid dead flat consolidations.
- Momentum: MACD histogram confirmation in the direction of the trade.
- High-Conviction Sniper Mode:
The EA initiates trades only when the policy outputs a confidence level $\ge 35%$ (well above the 25% random baseline of a 4-action space). When conditions are noisy or ambiguous, the model remains in HOLD, keeping your capital safe. - 70% Trailing Stop + Native PPO Close:
Once price reaches 70% of the distance toward Take Profit, a dynamic trailing stop locks in profits while providing 0.8 ATR of breathing room. If the model detects momentum exhaustion before hitting TP, it fires an early Close Signal (Action 3) to secure profits. - Rollover Filter & Anti-Gap Protection:
Avoids high-spread rollover transition hours and closes open positions every Friday night before market close, shielding your account from weekend price gaps.
⚙️ Detailed Input Parameters
1. Institutional Risk Engine
- (Default: 1.0): Percentage of account equity risked per trade. Position lot size is dynamically calculated based on the exact Stop Loss distance.
- (Default: 1.0): ATR volatility multiplier for Stop Loss distance.
- (Default: 1.5): ATR volatility multiplier for Take Profit target (favorable 1:1.5 Risk-to-Reward ratio).
2. Macro Quant Gate
- (Default: 22.0): Minimum ADX trend strength threshold required to authorize market entry.
- (Default: 0.35): Minimum neural network confidence threshold (0.35 = 35% mathematical conviction).
3. Trailing Stop & Native Execution
- (Default: true): Enables dynamic trailing stop management.
- (Default: 0.70): Percentage distance toward Take Profit required to trigger trailing (70%).
- (Default: 0.80): ATR trailing buffer distance to prevent premature stopouts.
- (Default: true): Allows the AI model to exit early when momentum wanes.
- (Default: true): Liquidates open positions on Friday at 21:00 server time to eliminate weekend gap risk.
4. Session Timing Filter
- InpUseSessionFilter (Default: true): Restricts execution to institutional volume windows.
- InpStartHourSession1 / InpEndHourSession1 (Default: 03 to 08): European / London volume window.
- InpStartHourSession2 / InpEndHourSession2 (Default: 14 to 18): US / New York volume window.
📋 Requirements & Setup Recommendations
- Currency Pair: EUR/USD
- Timeframe: M15
- Minimum Balance: $100 USD (or any standard prop firm challenge account).
- Account Type: Raw Spread / ECN with low spread conditions.
- Leverage: 1:30 to 1:500.
- Execution: A low-latency Virtual Private Server (VPS) is recommended for uninterrupted 24/5 execution.
