Financial Analysis Project

A Django-based application for financial data analysis, backtesting, and machine learning predictions. You can run this locally or deploy it to AWS. A demo version is currently hosted and accessible via example API endpoints.

Features

• Fetch and store financial data from Alpha Vantage API
• Perform backtesting with customizable strategies
• Generate predictions using pre-trained machine learning models
• Create performance reports in PDF and JSON formats
• Dockerized setup for easy deployment

Prerequisites

• Python 3.1+
• pip (Python package manager)
• PostgreSQL
• Docker
• Alpha Vantage API key (Alpha Vantage)

Local Setup

1. Clone and install:

git clone https://github.com/yourusername/financial-analysis-project.git
cd financial-analysis-project

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

pip install -r requirements.txt

2. Create .env file:

DATABASE_URL=postgresql://localhost:5432/dbname
ALPHA_VANTAGE_API_KEY=your_alpha_vantage_api_key
DEBUG=True
ALLOWED_HOSTS=localhost,127.0.0.1

3. Setup database:

python manage.py migrate
python manage.py createsuperuser  # Optional: for admin access

Working with Stock Data

Fetching Data

Before running any analysis, you need to fetch stock data:

# Fetch last 30 days of data
python manage.py fetch_stock_data AAPL --days 30

# Fetch 2 years of historical data (recommended for backtesting)
python manage.py fetch_stock_data AAPL --years 2

Important Notes:

• Required before running analysis
• Alpha Vantage API rate limit: 5 requests/minute (free tier)
• Initial fetch may take several minutes

Training Models

For predictions, train a model for each stock symbol:

python manage.py train_ml_model AAPL

Demo API Examples

A demo version is hosted at 3.130.162.114:8000. You can test the functionality using these curl commands:

Backtesting Example

curl -X POST http://3.130.162.114:8000/financial_data/backtest/ \
-H "Content-Type: application/json" \
-d '{
    "symbol": "AAPL",
    "initial_investment": 10000,
    "buy_ma_window": 5,
    "sell_ma_window": 20
}'

Make sure to use the appropriate moving average numbers according to the data fetched

Predictions Example

curl -X POST "http://3.130.162.114:8000/financial_data/predict/" \
-H "Content-Type: application/json" \
-d '{"symbol": "IBM"}'

Report Generation Example

curl -X POST -H "Content-Type: application/json" \
-d '{
    "symbol": "AAPL",
    "start_date": "2024-11-01",
    "end_date": "2024-12-01",
    "initial_investment": 10000,
    "buy_ma_window": 5,
    "sell_ma_window": 20,
    "format": "pdf"
}' \
http://3.130.162.114:8000/financial_data/report/ --output AAPL_report.pdf

This is how it may look:

AWS Deployment (Optional)

AWS RDS Setup

1. Create PostgreSQL RDS instance:

   • Go to AWS RDS Dashboard
   • Click "Create database"
   • Choose PostgreSQL
   • Select Free tier or preferred tier
   • Configure instance details and connectivity

2. Configure Security Group:

   • Add inbound rule for PostgreSQL (port 5432)
   • Source: Your EC2 security group ID

AWS EC2 Setup

1. Launch EC2 instance:

   • Choose Amazon Linux 2 AMI
   • Select instance type
   • Configure security groups (ports 22, 80, 443, 8000)
   • Create/select key pair

2. Install requirements:

sudo yum update -y
sudo yum install -y docker git postgresql15
sudo service docker start
sudo usermod -a -G docker ec2-user

3. Deploy application:

# Clone your repository
git clone your-repository-url
cd your-project-name

# Create production .env
DATABASE_URL=postgresql://username:password@your-rds-endpoint:5432/dbname
ALPHA_VANTAGE_API_KEY=your_alpha_vantage_api_key
DEBUG=False
ALLOWED_HOSTS=your-ec2-public-dns,your-domain.com

# Deploy with Docker
docker build -t financial-analysis .
docker run -d --env-file .env -p 8000:8000 financial-analysis

Post-Deployment Setup

1. SSH into your EC2 instance:

ssh -i your-key.pem ec2-user@your-ec2-ip

2. Access your Docker container:

docker exec -it $(docker ps -q) bash

3. Fetch historical data for each stock you want to analyze:

python manage.py fetch_stock_data AAPL --years 2  # Fetch 2 years of AAPL data
python manage.py fetch_stock_data MSFT --years 2  # Fetch 2 years of MSFT data
# Repeat for other stocks

4. Train the ML model for each stock:

python manage.py train_ml_model AAPL  # Train model for AAPL
python manage.py train_ml_model MSFT  # Train model for MSFT
# Repeat for other stocks

After completing these steps, you can use the same API endpoints shown in the examples above, replacing 3.130.162.114 with your EC2 instance's public IP address.

CI/CD Configuration (Optional)

Required GitHub secrets for automatic deployments:

AWS_ACCESS_KEY_ID
AWS_SECRET_ACCESS_KEY
EC2_HOST
EC2_USERNAME
EC2_SSH_KEY

See .github/workflows/main.yml in the repository for the complete CI/CD configuration.