BTC Paper Trading Bot

A live Bitcoin paper trading bot with Volatility-Conditioned Regime Model and Continuous Online Learning. Features a beautiful Streamlit dashboard with real-time price charts, prediction graphs, and trading metrics.

Features

• ML-Powered Trading: SGDRegressor with L2 regularization for online learning
• Live Dashboard: Real-time Streamlit interface with Plotly charts
• Continuous Learning: Model improves with every new price tick
• Regime Detection: Volatility-conditioned trend/regime model
• Paper Trading: Risk-free simulation with $10,000 starting balance
• Aggressive Mode: No-hold trading for maximum action

Quick Start

Prerequisites

• Python 3.10+
• pip

Installation

1. Clone the repository

   git clone https://github.com/yourusername/btc-live-trading.git
   cd btc-live-trading

2. Create virtual environment

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies

   pip install -r requirements.txt

4. Run the app

   streamlit run app.py

5. Open in browser: Navigate to http://localhost:8501

##Dashboard Preview

The dashboard displays:

• Trading Metrics: Cash, BTC holdings, portfolio value, PnL
• Live Price Chart: Real-time BTC price with prediction overlay
• Signal Indicators: Current action, volatility regime, trend regime
• Trade Log: Recent trade history with timestamps

##Model Architecture

Volatility-Conditioned Regime Model

The bot uses a regime-aware approach to trading:

Regime	Volatility	Strategy
🔥 High Vol	>70%	Trend-following signals
🌡️ Medium Vol	30-70%	Balanced approach
❄️ Low Vol	<30%	Mean-reversion signals

Features (12 total)

1. Volatility regime score
2. Trend regime score
3. Short-term trend (10-period)
4. Long-term trend (EMA crossover)
5. Normalized volatility
6. Volatility change
7. Trend-conditioned signal
8. Reversion-conditioned signal
9. Short-term z-score
10. Long-term z-score
11. Trend acceleration
12. Price/ATR ratio

Online Learning

The model learns from every new price tick:

• Uses SGDRegressor.partial_fit() for incremental updates
• Higher learning rate (0.2) for fast adaptation
• Lower regularization (0.0005) for quick learning

📁 Project Structure

btc-live-trading/
├── app.py                    # Streamlit dashboard
├── src/
│   ├── __init__.py
│   ├── model.py              # ML model & trading logic
│   └── data.py               # Data fetching utilities
├── Bitcoin_Historical_Data.csv
├── requirements.txt
├── .gitignore
└── README.md

⚙️ Configuration

Key parameters in src/model.py:

Parameter	Default	Description
TRADE_FRACTION	0.30	Base trade size (30%)
STRONG_SIGNAL_MULTIPLIER	2.0	2x trades on strong signals
WEAK_SIGNAL_MULTIPLIER	0.5	0.5x trades on weak signals
STRONG_THRESHOLD	0.0003	Strong signal threshold
WEAK_THRESHOLD	0.00005	Weak signal threshold

Terminal Mode

For terminal-based trading (no GUI):

python btc_paper_trader.py

License

MIT License - feel free to use and modify!

Disclaimer

This is a paper trading bot for educational purposes only. It does not involve real money or actual trades. Past performance does not guarantee future results. Always do your own research before trading cryptocurrencies.