EA AurumNeuro Vanguard - expert for MetaTrader 5

AurumNeuro Vanguard is an intelligent Expert Advisor (EA) specifically designed for XAUUSD / Gold trading, combining a hybrid Neural Risk Architecture with the Unified Market Dynamics Engine (UMDE).

Rather than relying solely on conventional technical signals, AurumNeuro Vanguard combines causal price analysis, an online-learning neural network, and ATR-based adaptive risk management to evaluate market conditions and generate disciplined entry and exit decisions.

Key Features

UMDE Core Engine

Detects market direction by analyzing price velocity, entropy, and Causal Price Dynamics (CPD) to filter out lower-quality market signals and focus on stronger trading conditions.

//+------------------------------------------------------------------+
//|                                                                  |
//+------------------------------------------------------------------+
class CUMDE_Kernel
{
private:
   MarketState    m_pool[STATE_POOL_SIZE];
   MarketState    m_l1_buffer[L1_SIZE];
   ushort         m_free_ptr;
   int            m_head;
   double         m_cpd, m_hamiltonian, m_cpd_smooth, m_scale;
   bool           m_shadow_mode;
public:
   CUMDE_Kernel() : m_free_ptr(0), m_head(0), m_cpd(0.0),
      m_shadow_mode(true), m_cpd_smooth(0.0) {}
   void InitScale()
   {
      point = SymbolInfoDouble(_Symbol, SYMBOL_POINT);
      m_scale = (point < 0.01) ? 1000.0 : 100.0;
   }
   void UpdateState(double price, double vol, double spread)
   {
      m_head = (m_head + 1) & (L1_SIZE - 1);
      double prev_price = m_l1_buffer[(m_head + L1_SIZE - 1) & (L1_SIZE - 1)].price;
      double vel = (price - prev_price) * m_scale;
      m_l1_buffer[m_head].price     = price;
      m_l1_buffer[m_head].velocity  = vel;
      m_l1_buffer[m_head].curvature = vel * 0.5;
      m_l1_buffer[m_head].entropy   = spread * vol;
      m_l1_buffer[m_head].timestamp = TimeCurrent();
   }
   void EvaluatePhysics()
   {
      double v = m_l1_buffer[m_head].velocity;
      double s = m_l1_buffer[m_head].entropy;
      double raw_cpd = MathAbs(v * (1.0 / (s + 0.0001)));
      double current_cpd = MathMin(raw_cpd, 10.0);
      m_cpd_smooth = 0.95 * m_cpd_smooth + 0.05 * current_cpd;
      m_cpd = m_cpd_smooth;
      m_hamiltonian = (v * v) + (s * 0.5);
      m_shadow_mode = (m_cpd < 0.60);
      m_l1_buffer[m_head].cpd        = m_cpd;
      m_l1_buffer[m_head].hamiltonian= m_hamiltonian;
   }
   MarketState GetCurrentState() const
   {
      return m_l1_buffer[m_head];
   }
   ENUM_ORDER_TYPE GetCausalDirection() const
   {
      return (m_l1_buffer[m_head].velocity > 0) ? ORDER_TYPE_BUY : ORDER_TYPE_SELL;
   }
   bool IsInShadowMode() const
   {
      return m_shadow_mode;
   }
   double GetCPD() const
   {
      return m_cpd;
   }
};
CUMDE_Kernel UMDE;

Neural Advisor

Uses a 5-12-3 neural network architecture that is continuously trained online using previous bar data. The Neural Advisor provides directional confirmation as well as dynamic Take Profit (TP) and Stop Loss (SL) guidance.

Adaptive Risk & Lot Sizing

Supports both Fixed Lot and Risk Percent trading modes. Position size can be calculated automatically based on SL distance, commission, and tick value, helping maintain consistent risk across different market conditions.

ATR Trailing Stop + Anti-Reversal Buffer

Features an ATR-based trailing stop with an optional aggressive mode and dynamic stop widening during extreme volatility. This helps reduce the chance of premature stop-outs when Gold experiences rapid price reversals.

Hard Close Profit — Risk/Reward

Automatically closes positions when the achieved profit reaches the configured Risk/Reward (RR) target, providing a clear and systematic profit-taking mechanism.

//+------------------------------------------------------------------+
//|                                                                  |
//+------------------------------------------------------------------+
void ManageHardCloseProfit()
{
   if(!InpUseRR) return;
   for(int i = PositionsTotal() - 1; i >= 0; i--) {
      ulong ticket = PositionGetTicket(i);
      if(!PositionSelectByTicket(ticket)) continue;
      if(PositionGetString(POSITION_SYMBOL) != _Symbol) continue;
      double entryPrice = PositionGetDouble(POSITION_PRICE_OPEN);
      double currentSL  = PositionGetDouble(POSITION_SL);
      ENUM_POSITION_TYPE posType = (ENUM_POSITION_TYPE)PositionGetInteger(POSITION_TYPE);
      if(currentSL == 0) continue;
      double slPoints = (posType == POSITION_TYPE_BUY) ?
                        (entryPrice - currentSL) / point :
                        (currentSL - entryPrice) / point;
      if(slPoints <= 0) continue;
      double bid = SymbolInfoDouble(_Symbol, SYMBOL_BID);
      double ask = SymbolInfoDouble(_Symbol, SYMBOL_ASK);
      double profitPoints = (posType == POSITION_TYPE_BUY) ?
                            (bid - entryPrice) / point :
                            (entryPrice - ask) / point;
      if(profitPoints >= InpRRatio * slPoints) {
         Print("RR hit: ticket=", ticket, " profit=", DoubleToString(profitPoints,1),
               " pts, SL=", DoubleToString(slPoints,1),
               " RR=", DoubleToString(InpRRatio,1));
         ClosePosition(ticket);
      }
   }
}

Hard Close Cut Loss

Can exit losing positions early when neural confidence falls below the configured threshold, helping protect account equity from prolonged or excessive drawdown.

Strict Gatekeeper

Uses multiple execution filters, including spread control, trade cooldown, and one-position-per-bar protection, designed to reduce unnecessary entries and overtrading.

Target Market

• Instrument: XAUUSD / Gold
  (The EA may also be adapted to other highly volatile instruments.)

• Recommended Timeframes: M15 – H1

• Optimal Timeframe: H1 for more stable signal conditions

• Account Types: Standard, Raw Spread, or ECN accounts with competitive commissions

• Recommended Starting Capital: $500 minimum

• Recommended Capital: $1,000+ when trading 0.01 lots

Competitive Advantages

Adaptive Learning

The neural network continuously updates its weights using new market data, allowing the model to adapt to changing market behavior over time.

Dual-Layer Protection

Combines adaptive ATR trailing protection with hard close cut-loss logic, providing multiple layers of protection against unfavorable market movements.

Transparent Configuration

Major trading, risk, and execution parameters are exposed through the EA inputs, making it easier to optimize the system and adapt it to different trading preferences and market conditions.

Built for Gold's Volatility

AurumNeuro Vanguard is designed for traders looking for an AI-inspired trading system built on a mathematical foundation and disciplined risk management framework.

By combining market dynamics analysis, online neural learning, adaptive volatility management, and strict trade execution controls, AurumNeuro Vanguard is engineered to handle the fast and volatile characteristics of the Gold market with a systematic approach.