Welcome to our GitHub repository, where we're innovating the cryptocurrency trading space with our cutting-edge Algorithmic Trading Model focused on the BTC/USDT market. Dive into the world of machine learning, statistical modeling, and advanced programming to unlock new potentials in trading strategies. 🚀
- Problem Statement: Exploring the significance of algorithmic trading models in the BTC/USDT cryptocurrency market.
- Objective: Utilize machine learning, statistical modeling, and programming skills to pioneer ML-based algorithmic trading.
- Tasks: Data acquisition, preprocessing, model design, backtesting, risk management, and optimization with a spotlight on BTC/USDT market dynamics.
- Historical Data: BTC/USDT trading pair data from January 1, 2018, to January 31, 2022.
- Data Sources: Encouragement to utilize public cryptocurrency market data sources, API services, or simulated data for comprehensive analysis.
- 📉 Trend Insights: Examination of BTC/USDT closing prices to decipher cryptocurrency trends.
- 🔄 Lag Plots: Analysis of time series correlation with its lagged values to understand evolving patterns.
- 🔍 Stationarity Assessment: KPSS Test indicates non-stationarity, leading to preprocessing steps like differencing and log transformations for stabilization.
- Time Series Models: Application of moving averages for trend analysis and decision-making.
- Machine Learning Models: Deployment of LSTM networks for capturing complex market dynamics.
- Risk Management: Utilization of GARCH models for volatility assessment and management.
- "Bitcoin Return Volatility Forecasting: A Comparative Study between GARCH and RNN"
- Max Drawdown: 10%
- Sharpe Ratio: 0.0017
- Net Profit: Exceeding benchmark return of 0%
- Risk-Reward Ratio: 1.01
- Max Duration of Single Trade: 0.116 days
Feel free to explore our repository for more details on our models and methodologies. Your contributions and feedback are highly appreciated! 🌟
Happy Trading! 💼