Autonomous Equity Trading with Python

Overview

A modular, production-grade algorithmic trading system built in Python. This framework automates the lifecycle of a long-only equity rebalancing strategy, integrating real-time market data analysis with state-aware execution across SGX, HKEX, NYSE, and NSE.

Technical Highlights

• Z-Score Based Entry Signals: Implements volatility-adjusted mean reversion logic using standard deviation thresholds rather than static percentages to determine entry points.
• VWAP Momentum Trigger: Features a breakout detection system that initiates trades when price action trends above the Volume-Weighted Average Price on surging volume.
• Portfolio Delta Reconciliation: Custom synchronization engine that calculates the difference between current holdings and target model weights to minimize transaction costs.
• Multi-Market Ticker Normalization: A robust fallback system to handle disparate ticker symbology across international exchanges (e.g., resolving .SI, .HK, .NS).
• Advanced Risk Management: Integrated server-side trailing stops (ATR-based) and a 48-hour earnings blackout filter to mitigate binary event risk.

System Architecture

The engine operates in a five-phase execution loop to ensure safety and precision:

1. Factor-Based Screening: Scans a custom universe to rank assets based on momentum, profitability, and liquidity factors.
2. State Handshake: Establishes a secure session with the Tiger Brokers Open API and retrieves real-time buying power.
3. Diagnostic Health Check: Generates a Target vs. Actual audit table to visualize portfolio drift before any execution occurs.
4. Intraday Signal Scan: Monitors live price action against both mean-reversion (Dip Buy) and momentum (Breakout) windows.
5. Autonomous Cleanup: Automatically identifies and liquidates positions that no longer meet the model's ranking criteria.

Quantitative Analysis & Risk Controls

• Correlation Penalty: Actively reduces the target weight of assets that exhibit high 90-day correlation (>0.80) to higher-conviction positions to prevent correlated drawdown.
• Sector Concentration Caps: Enforces a strict 30% maximum weight per sector to maintain industry diversification.
• Macro Regime Adaptation: Dynamically adjusts volatility and dividend multipliers based on whether the benchmark is trending above or below its 200-day moving average.

Tech Stack

• Language: Python 3.12+
• APIs: Tiger Brokers Open API (Trade/Quote SDK)
• Data Libraries: yfinance, pandas, numpy, vaderSentiment
• Environment Management: python-dotenv for RSA-encrypted key management

Project Structure

• execution/: RSA Authentication, Connection Management, and Order Delta Logic.
• quant/: Alpha Factor generation, Z-score signals, and Earnings Blackout logic.
• analysis/: Risk-parity allocation, correlation penalties, and backtesting modules.
• qual/: Sentiment analysis and NLP-based event classification for news headlines.

Demonstration (Mock Mode)

To view the engine's logic without active API credentials:

1. Clone the repository.
2. Install dependencies: pip install -r requirements.txt.
3. Run python main.py --dry-run to view the Diagnostic Health Check and Signal Scanning in action using sample data.