TickerQuote: Low-Cost Real-Time Stock Strategy Analyzer

TickerQuote is a cost-efficient, real-time stock strategy platform built using AWS services, React/TypeScript, Python, and GitHub Actions. It provides fundamental screening and real-time price analysis of U.S. stocks, with a strong emphasis on security, automation, and minimal operating cost.

Project Overview

The system is composed of three primary components:

1. Frontend (stock-strategy-app)
   A React + TypeScript web application that:

   • Loads stock fundamentals from S3.
   • Filters stocks based on selected strategies.
   • Provides interactive analysis and data exploration.

2. Lambda Backend Services

   • GetFundamentals: Periodically pulls and stores comprehensive stock fundamentals (e.g., via Finnhub or similar APIs) into S3.
   • EvaluateStockStrategy: Applies logic to screen the current stocks against defined strategies.

3. EC2 WebSocket Price Streamer

   • A lightweight EC2 instance (Amazon Linux 2023, Free Tier) connects to the Twelve Data WebSocket API.
   • Streams live stock prices to clients.
   • Uses systemd and a custom service (streamer.service) to ensure the app is resilient and restartable.

Security Architecture

• Secrets Management

  • All secrets used in the Lambda functions are stored in AWS SSM Parameter Store (SecureString).
  • The EC2 instance also retrieves runtime secrets (like Dynu credentials) from SSM at launch.
  • GitHub Actions workflows use GitHub Actions Secrets for deployment authentication.

• Client Isolation

  • No secrets are ever exposed to the frontend client.
  • The website fetches public JSON files from S3 and receives live pricing from a public WebSocket hosted on the EC2 instance.

GitHub Actions & CI/CD

• GitHub Actions is configured to:

  • Package and deploy the Lambda functions on every commit.
  • Build and upload the static frontend to S3 + CloudFront.
  • SSH into the EC2 instance (or use AWS Session Manager) to deploy or restart the streaming service.

• The deployment is fully automated and triggered by successful pushes to the main branch. This may be customized in the future as development workflows evolve.

Dynamic DNS with Dynu & Zero Fixed IP Cost

• To avoid the cost of a static Elastic IP (EIP), the EC2 instance uses Dynu Dynamic DNS:
  • At boot, a script updates Dynu with the current public IP via a script.
  • A stock-strategy.ddnsfree.com hostname is maintained with Dynu and used for SSH and WebSocket connections.
  • Systemd runs this script before starting the streaming service, ensuring the IP is always current.

Cost Optimization

This project was engineered to run entirely within AWS Free Tier limits:

• Lambda: Fully serverless and only invoked periodically.
• S3 + CloudFront: Used for serving the frontend and stock data, with caching enabled.
• EC2: Runs only during market hours via manual or scheduled automation.
• No Elastic IP: Dynamic DNS removes the need for a static IP.

Strategy Logic

The stock strategy logic is written to be extensible and currently includes:

• Basic screeners using price-to-earnings ratios, earnings growth, and other fundamentals.
• Real-time validation with price feeds to determine if a stock is "in a good place to buy".

Strategy logic runs in both Lambda and browser (via filtered JSON), enabling rapid iteration and transparency.

Directory Structure (Simplified)

tickerquote-main/
├── lambdas/
│   ├── EvaluateStockStrategy/
│   └── GetFundamentals/
├── stock-strategy-app/     # React + TypeScript client
├── ec2/
│   ├── stream_price_data_app.py
│   └── update_dynu_ip.sh
├── .github/workflows/
│   ├── deploy-frontend.yml
│   ├── deploy-lambdas.yml
│   └── deploy-ec2.yml

Getting Started

To clone and deploy your own version:

git clone https://github.com/your-username/tickerquote.git
cd tickerquote

Ensure you configure:

• GitHub Secrets for deployment credentials.
• AWS SSM parameters for runtime secrets (DYNU_USERNAME, DYNU_PASSWORD, etc.).
• Optional: Set up Dynu Dynamic DNS account.

Then simply push changes to main and watch your infrastructure update itself.

Testing

Test Finnhub directly

This works with the free account.

wscat -c 'wss://ws.finnhub.io?token=<API-Token>'

This opens a websocket. Then subscribe to the data you want to stream in the console. Enter:

{"type":"subscribe","symbol":"BINANCE:BTCUSDT"}

Binance is good because it will work on the weekends too.

Test deployed EC2 streaming price data

Run the following:

cd <repo-root>
setopt interactivecomments
npm install --save-dev dotenv-cli # if necessary, install dotenv if not installed
npm run ec2:start
npm run ec2:status # Make sure it starts, rerun until 'running'
npm run test:ec2 # test runs for 10 seconds
npm run ec2:stop # stop it, don't waste money
npm run ec2:status # Make sure it stops, rerun until 'stopped'
unsetopt interactivecomments

Test deployed Lambda OHLCV historic candles

npm run test:ohlcv

Credits

Created and maintained by Barnaby Falls.
Built to maximize transparency, automation, and cost-efficiency.