AlphaFX is a production-grade, event-driven trading system for live multi-timeframe forex trading. Built around OANDA's v20 REST API, it integrates complex event processing (CEP), robust technical indicator pipelines, advanced signal orchestration, and research-backed risk management strategies.
This system is designed for real-time deployment and also includes offline analysis tools, monitoring dashboards, and performance reporting.
- Asynchronous data fetching from OANDA REST API for M1, M5, M15 timeframes
- Deduplication, caching, and optimized incremental updates
- Pandas-formatted DataFrames with consistent timestamp indexing
- TA-Lib-based indicators: EMA ribbons, MACD, RSI, Bollinger Bands, ADX, ATR, OBV, KAMA, CCI, VWAP
- Custom signal construction features (e.g., candle shapes, price action flags, volume trends)
- Adaptive parameter tuning based on market regime
- Multi-timeframe pattern detection with signal scoring and conflict resolution
- Gaussian Mixture Model-based market regime detection
- Regime-specific feature pipelines and signal thresholds
- Signal consistency filters and stabilization logic
- Kelly Criterion-based position sizing with volatility and confidence adjustments
- ATR-based stop loss and take profit levels
- Maximum correlated risk constraints and drawdown-based shutdown/recovery logic
- Real-time enforcement of max positions per instrument and global exposure
- Terminal-based dashboard for live positions, recent trades, and metrics
- Comprehensive trade logging (CSV + JSON) with full order metadata
- Performance metrics: drawdown, win rate, profit factor, risk level tracking
- Reusable modules for backtesting using stored OHLCV data
- Performance breakdown by instrument, regime, and timeframe
- Monte Carlo simulation and walk-forward analysis support
strategy_tester.pyruns a portfolio backtest with multi-instrument, multi-timeframe data- Fully mirrors the live system’s execution logic and position management
- Includes per-trade tracking, balance curve, drawdown monitoring, and advanced risk metrics
performance_analyzer.pyloads trade logs and performance logs to generate detailed reports- Outputs Sharpe ratio, profit factor, average win/loss, drawdown, risk/reward, and signal statistics
- Includes equity curve plots, drawdown visualizations, and win/loss distribution histograms
- Can generate CLI summaries or write structured reports to disk
AlphaFX has been tested across multiple scenarios using the strategy_tester.py module and live historical OHLC data. These tests validate end-to-end integration of market ingestion, technical indicator generation, signal alignment (CEP), and live-execution logic under realistic constraints.
- Instruments: EUR/USD, AUD/USD, USD/CAD
- Timeframes: M1, M5, M15
- Risk per Trade: 0.8%
- Total Trades: 5
- Duration: 5 days
- Execution Style: High-conviction, low-frequency
- Initial Balance: $15,000
- Final Balance: $21,066
- Net Profit: $6,066
- Return: +40.4%
- Win Rate: 80%
- Profit Factor: 118.1
- Sharpe Ratio: 2.21
- Signal Efficiency: 34.4%
- Instruments: EUR/USD, USD/CAD
- Timeframes: M1, M5, M15
- Risk per Trade: 2.5%
- Total Trades: 11
- Duration: 3 days
- Execution Style: Higher-frequency, aggressive sizing
- Initial Balance: $10,000
- Final Balance: $14,925
- Net Profit: $4,925
- Return: +49.2%
- Win Rate: 100%
- Sharpe Ratio: 4.58
- Signal Efficiency: 96.1%
- The system successfully adapted to both conservative and aggressive configurations, maintaining consistent performance with different trade frequencies and risk levels.
- CEP logic and multi-timeframe alignment demonstrated robustness across both setups.
- Further testing over longer windows and varied market conditions is planned for final validation.
AlphaFX/
├── run_trading_system.py # Entry point for live/practice trading
├── market_data_adapter.py # OANDA candle ingestion and caching
├── technical_indicators.py # TA-Lib and custom feature computation
├── cep_engine.py # Signal generation and regime detection
├── order_routing.py # Order placement, SL/TP, risk sizing
├── trade_logger.py # Trade and performance logging
├── config.py # Configurable settings and thresholds
├── monitor_trades.py # CLI dashboard for live trading
├── strategy_tester.py # Multi-instrument portfolio backtester
├── performance_analyzer.py # Post-trade analyzer and visualizer
├── requirements.txt
└── /output # Logs, trade CSVs, summaries (excluded via .gitignore)
└── /plots # Backtest results
- Python 3.8+
oandapyV20,pandas,numpy,scikit-learnta-lib,matplotlib,seaborn,tabulatepython-dotenv,asyncio
Install with:
pip install -r requirements.txt
Also create a .env file:
OANDA_API_KEY=your_api_key_here
OANDA_ACCOUNT_ID=your_account_id_here
python run_trading_system.py --practice
python monitor_trades.py
python strategy_tester.py --instruments EUR_USD GBP_USD --days 5 --balance 10000 --risk 3.0 --plot
python performance_analyzer.py --plot --report alpha_report.txt
This system was developed independently to demonstrate research-informed signal engineering, real-time execution architecture, and risk-aware trade orchestration. While the core architecture is robust and complete, full profitability requires further parameter tuning, strategy calibration, and live testing under varied market conditions.
Amin Sharifi
Quantitative Researcher | AI/ML Developer | PhD Composer
Email: info@masharifi.com
Website: https://www.masharifi.com
LinkedIn: https://www.linkedin.com/in/amin-sharifi-062a477b
GitHub: https://github.com/amin-sharifi-github