Gold Markovify
- Experts
- Version: 1.1
- Updated: 12 May 2026
- Activations: 10
Gold Markovify is a quantitative trading system designed to identify and exploit structural market regimes using Hidden Markov Models (HMM). Unlike traditional indicators that rely on linear moving averages, this engine utilizes a probabilistic approach to determine the hidden states of price action. It is specifically developed for instruments with high volatility and non-linear distribution, such as Gold and Silver.
Quantitative Core
The system operates by training a Baum-Welch HMM engine on historical price data to categorize the market into discrete hidden states. Each state represents a different market environment, such as trending or mean-reverting. By applying Adaptive Fractional Differentiation, the engine removes market noise while preserving the long-term memory dependencies of the price series.
Volatility and Risk Management
Gold Markovify incorporates a GARCH(1,1) parameter estimation module to forecast conditional volatility. This allows the system to adjust stop-loss and take-profit levels dynamically based on current market risk. The execution pipeline is governed by a multi-gate risk management system that evaluates spread-to-ATR ratios and global exposure limits across multiple instances of the EA.
Metal Hardening Features
The engine includes specialized logic for Gold and Silver trading. The Trap-Guard module analyzes price wicks to identify exhaustion levels and potential stop-hunting behavior before trade execution. Additionally, a divergence filter monitors the correlation between Gold and Silver to confirm institutional commitment to a directional move.
Operational Inputs
The parameters are organized into logical blocks for ease of configuration.
Main Settings
- Threshold: The probability level required to trigger a trade signal.
- HMM States: The number of hidden states the model identifies during training.
- Magic Number: A unique identifier for trades to prevent interference with other robots.
Quantitative Calibration
- Hurst Minimal: The minimum Hurst exponent value required to confirm a trending regime.
- Confidence Min: The minimum Bayesian confidence score needed for execution.
- Viterbi Filter: Enables the Viterbi algorithm for state sequence verification.
Risk Management
- Risk Percent: The percentage of account equity at risk per trade.
- ATR Period: The lookback window for volatility normalization.
- Max Spread: The maximum allowed spread in points before the trade is blocked.
- Global Max Lot: The maximum cumulative volume allowed across all instances of the EA.
Hardening and Safety
- Spike Filter: Blocks trades if the current candle range exceeds a multiplier of the ATR.
- Divergence Sync: Enables correlation checking between Gold and Silver.
- Equity Recovery: Automatically scales down trade volume when the account is in a drawdown.
Requirements and Recommendations
- Symbol: Optimized for XAUUSD and XAGUSD.
- Timeframe: M15 or H1 is recommended for balanced signal frequency.
- Modeling: For the Strategy Tester, please use Every tick based on real ticks to ensure accurate GARCH calculations.
- Account: A hedging account is required for the stealth execution logic to function correctly.
