🚀 Trading-System

Institutional-Grade Algorithmic Trading & Backtesting Framework

---

📌 Overview

Trading-System is a modular, high-performance algorithmic trading and backtesting framework built with Python and Django.

The system is designed to simulate realistic trading conditions while providing a scalable research environment for developing, testing, and analyzing trading strategies.

It consists of:

• 🧠 Intelligent Trading Agents
• 🏦 High-Fidelity Virtual Broker
• 🚄 Optimized Historical Data Pipeline
• 🌐 Interactive Web Dashboard

The architecture is fully decoupled to allow independent development of strategies, execution logic, and analytics.

---

🏗 System Architecture

The project follows a modular architecture:

---

Component Responsibilities

Component	Responsibility
Agents	Generate trading signals (ML / heuristic logic)
Execution Engine	Coordinate signals and manage trade lifecycle
Virtual Broker	Simulate order execution & account accounting
Database Layer	Efficient historical data storage
Django Dashboard	Visualization, analytics & reporting

---

🧠 Intelligent Trading Engine

The system supports advanced strategy modules including:

▸ Machine Learning Based Agents

• Regime detection
• Directional bias prediction
• Noise-robust classification logic

▸ Multi-Timeframe Pattern Detection

• Swing Failure Pattern (SFP) detection
• Structure-based reversal setups
• Cross-timeframe confirmation logic

Strategies are fully decoupled from execution to allow safe experimentation.

---

🏦 Virtual Broker (High-Fidelity Simulation)

Unlike simple backtesters, the system includes a realistic broker simulation:

✔ Intra-Candle Execution Modeling

Simulates price movement inside OHLC candles:

Ensures accurate SL / TP detection.

✔ Financial Engine

Real-time calculation of:

• Balance
• Equity
• Floating P/L
• Margin
• Free Margin
• Drawdown

✔ Realistic Costs

• Dynamic spreads
• Commission per lot
• Slippage modeling
• Gap handling

This results in near real-world execution simulation.

---

🚄 High-Performance Data Pipeline

Large backtests require efficient storage.

The system uses:

• Zlib compression for historical candle data
• SQLite Binary BLOB storage
• Optimized retrieval pipeline
• Async API serving for large chart datasets

Capable of loading hundreds of thousands of candles efficiently.

---

🌐 Django Web Dashboard

The web interface provides:

• Interactive candlestick charts
• Multi-timeframe switching (1m / 5m / 15m / 1h)
• Execution logs synchronized with chart
• Click-to-zoom trade visualization
• Fast client-side aggregation

Built with:

• Django 5.x
• Bootstrap 5
• TradingView Lightweight Charts

---

🛠 Tech Stack

Backend

• Python 3.10+
• Django 5.x
• NumPy
• Pandas
• SciPy

Frontend

• TradingView Lightweight Charts
• Bootstrap 5

Storage

• SQLite (compressed BLOB storage)
• Optional PostgreSQL for production

---

📥 Installation

1️⃣ Clone Repository

git clone https://github.com/NegativeGravity/Trading-System.git
cd Trading-System
python -m venv .venv
source .venv/bin/activate
.venv\Scripts\activate

cd web_dashboard
python manage.py makemigrations
python manage.py migrate

python manage.py run_backtest_lorentzian --days 60 --tf 5

python manage.py runserver

http://127.0.0.1:8000