AlphaWaveTrader - Machine Learning Crypto Trading Strategy

Overview

AlphaWaveTrader is a machine learning-based trading strategy that executes trades on the ETH/USDT pair. The strategy integrates technical indicators, historical market data, and a pre-trained machine learning model to predict market trends and make automated buy/sell decisions. It aims to maximize returns while managing risk through the use of technical analysis indicators.

This repository contains the necessary code, models, and resources to set up and deploy the trading strategy, along with backtesting results.

Team Members

• codingis4noobs2 (Owner)

Project Setup

Step 1: Create a Virtual Environment

First, create and activate a virtual environment to manage project dependencies:

# Create a virtual environment
python -m venv env

# Activate the virtual environment
# On Windows:
env\Scripts\activate

# On macOS/Linux:
source env/bin/activate

Step 2: Install Dependencies

Once the virtual environment is activated, install all required dependencies using the provided requirements.txt file:

pip install -r requirements.txt

Step 3: Set Up Environment Variables

Run the setup_account.py script to generate the .env file, which will contain the necessary environment variables:

python setup_account.py

This will create a .env file that holds the GCP project ID and the path to the Google Cloud service account credentials.

Step 4: Configure Google Cloud Project

1. Obtain your Google Cloud Project ID.
2. Create a service account(Goto IAM and click Service Accounts) in your GCP project with the role: Basic -> Owner.
3. Make sure you have enabled BigQuery API.
4. Download the service account JSON file and place it in the project directory.
5. Update the .env file with the following fields:

GCP_PROJECT_ID=your_gcp_project_id
GOOGLE_APPLICATION_CREDENTIALS=path_to_your_service_account.json

Make sure the .env file has correct values to ensure successful integration with Google Cloud.

Step 5: Train the Model

To train the machine learning model, open and run the model_training.ipynb Jupyter notebook. This notebook fetches historical data, computes technical indicators, and trains the Random Forest Classifier model. The trained model is then stored in Google Cloud Storage for later use by the trading strategy.

Step 6: Execute the Strategy and Backtest

Run the trading_strategy.ipynb Jupyter notebook to execute the strategy and perform backtesting. This notebook simulates the strategy using historical data, evaluates its effectiveness, and outputs the performance metrics (e.g., returns, Sharpe ratio, drawdown). For 2 weeks testing, Please change the lookback value to 14 and Interval to 30m.

Core Logic of the Strategy

The trading strategy leverages a combination of technical indicators and a machine learning model (Random Forest Classifier) to predict price movements and generate buy/sell signals. Key technical indicators used are:

• Simple Moving Averages (SMA): 10-period and 20-period moving averages.
• RSI (Relative Strength Index): Momentum oscillator to identify overbought/oversold conditions.
• MACD (Moving Average Convergence Divergence): Measures momentum and trend direction.
• Bollinger Bands: Volatility-based indicator to identify overbought/oversold levels.
• ATR (Average True Range): Volatility indicator measuring the range of price movement.
• Momentum (MOM): Measures the rate of price change.
• Rate of Change (ROC): Percentage change between the current price and the price a number of periods ago.

The Random Forest Classifier takes these technical indicators as input features and predicts whether to buy or sell based on historical price action.

Intended Market Behavior

The strategy is designed to capture trends and avoid significant drawdowns. It performs best in trending markets and adjusts its risk based on real-time technical analysis.

Backtesting & Simulation Results

Performance Metrics

Backtesting was performed on the ETH/USDT pair from August 14, 2024 to October 13, 2024 (approximately 60 days). Initial capital was set to $4000, and the results were compared to a simple buy-and-hold strategy.

Key Performance Indicators (KPIs)

• Sharpe Ratio: A measure of risk-adjusted return. A Sharpe ratio above 2 is generally considered excellent.
• Max Drawdown: The maximum observed loss from a peak to a trough in the strategy’s equity curve.
• Profit Factor: The ratio of gross profit to gross loss, with a value greater than 2 indicating high profitability.

The trading strategy outperformed the buy-and-hold approach significantly, achieving a return of 30.74% compared to -10.41% for buy-and-hold. The risk-adjusted metrics (Sharpe ratio and Sortino ratio) also indicate strong performance with relatively low risk.

Data Feeds & Machine Learning Algorithms

Data Feeds

The strategy ingests historical and live OHLCV data (Open, High, Low, Close, Volume) for the ETH/USDT pair, fetched from cryptocurrency exchanges using the ccxt library. This data is used to compute technical indicators and train the machine learning model.

Machine Learning Algorithms

The strategy uses a Random Forest Classifier to predict price movements. The model is trained on historical data and generates buy/sell signals in real-time based on technical indicators.

Tokens Focused on Trading

The primary asset pair traded by this strategy is ETH/USDT. The strategy can be adapted to trade other pairs like BTC/USDT or BNB/USDT by adjusting the data sources and retraining the model with the corresponding asset data.

Notable Points

• Google Cloud Integration: Data is stored in BigQuery, and the trained model is saved in Google Cloud Storage. This allows for scalable storage and remote access.
• Risk Management: The strategy ensures only one position (long or short) is active at any time, reducing exposure to volatile markets.

Future Improvements

• Sentiment Analysis: Future versions could incorporate sentiment data from news sources or social media platforms to improve the model’s predictive power during volatile markets.
• Reinforcement Learning: Exploring reinforcement learning to allow the model to adapt to changing market conditions in real-time.
• Extended Asset Coverage: Expanding the strategy to cover more cryptocurrency pairs like BTC/USDT, BNB/USDT, and other high-liquidity assets.