|
no
experience
|
3
products
|
10
demo versions
|
|
0
jobs
|
0
signals
|
0
subscribers
|
My approach prioritizes transparent rules, out-of-sample validation, and clear documentation. I publish tools and trading systems for users seeking verifiable and configurable automation.
Past performance and historical test results do not guarantee future results.
Add friends via their profile or user search and you will be able to see if they are online
I reviewed a Gold Clean Ensemble test on XAUUSD H1 covering 2020-01-01 to 2026-07-23, with a fixed 0.10 lot and an initial deposit of $2,000.
Test snapshot:
• History quality: 98%
• Total trades: 93
• Net profit: $14,028.90
• Profit factor: 3.40
• Sharpe ratio: 2.27
• Maximum equity drawdown: 38.92%
The most useful observation is not the net profit. Profitable trades represented 35.48%, while the average winning trade was $602.00 and the average losing trade was $97.29 — a 6.19× payoff ratio.
In this test, the edge came from payoff asymmetry rather than a high win rate. The drawdown is also material and should not be ignored.
This is a historical research result, not a forecast or guarantee. Broker conditions, spread, execution and contract specifications can materially change the outcome.
Gold Clean Ensemble EA:
https://www.mql5.com/en/market/product/187303
When developing a machine-learning model for MetaTrader 5, using the same historical period for optimization and evaluation can create misleading results.
My current research workflow is:
• Training data: 2022–2024
• Out-of-sample validation: 2025
• Final untouched test: 2026
All input variables must be calculated using only information available at the moment of the signal. Future candles must never influence training features or trading decisions.
This methodology does not guarantee profitability, but it produces a more honest, reproducible, and useful evaluation of an automated trading system.
I will continue sharing practical observations about MQL5, ONNX integration, quantitative research, and real-tick backtesting.
Kumoedge Gold Intelligence Hybrid Ichimoku + XGBoost ONNX EA for XAUUSD on M5 XAUUSD GOLD M5 TIMEFRAME MetaTrader 5 PLATFORM XGBoost ONNX EMBEDDED MODEL Kumoedge Gold Intelligence is a MetaTrader 5 Expert Advisor designed for XAUUSD on M5. It combines Ichimoku Kinko Hyo rules with an XGBoost classification model exported to ONNX. The model is embedded in the EA, remains frozen during trading, and requires no external files or DLLs. 1. Recommended setup Platform: MetaTrader 5. Symbol: XAUUSD
LEVELTIDE XAU Multi-timeframe breakout and liquidity-sweep system for XAUUSD on MetaTrader 5. LevelTide XAU combines two independent modules: a main level-breakout strategy and FlowSweep, designed to identify expansion moves and liquidity sweeps. The EA runs from one XAUUSD M5 chart and autonomously analyzes M1, M5, M15, H1, H4, D1 and W1. Architecture Main module: analyzes support, resistance and adaptive levels, confirming entries with closed candles, trend context and market strength
It is an Expert Advisor designed for XAUUSD on the H1 timeframe, combining Donchian breakout structure, Keltner channel confirmation and MACD-based exit logic.
The EA uses predefined Stop Loss and Take Profit levels, does not use grid or martingale, and includes a free demo version for Strategy Tester evaluation.
Product page:
https://www.mql5.com/es/market/product/187303
Please test the demo version first and evaluate the strategy under your own broker conditions. Past backtest results do not guarantee future performance.
GOLD CLEAN ENSEMBLE EA Trend-breakout system for XAUUSD H1 on MetaTrader 5. Gold Clean Ensemble EA combines Donchian and Keltner channels with MACD logic to seek directional moves through transparent rules and predefined risk management. Strategy architecture Donchian identifies the entry breakout. Keltner confirms sufficient market expansion. MACD supports the regime filter and trend-exit logic. Risk and execution No martingale or grid. Each order includes emergency Stop Loss and Take Profit