Backtest Anything

Backtest Anything is a full-stack application that enables users to input plain-English trading strategies and receive Python code that can be backtested on historical Bitcoin data. The app visualizes buy/sell signals and calculates realistic profit and return metrics.

Features

• Translate natural-language strategies into valid Python trading logic using OpenAI GPT-3.5
• Automatically backtest on BTC historical data (2013–2024)
• View realistic trade signals, profit calculations, and return percentages
• No future price look-ahead — ensures realistic strategy evaluation

Tech Stack

• Frontend: Next.js 14+, TypeScript, Tailwind CSS, Recharts
• Backend: FastAPI (Python), Pandas, OpenAI API
• Data: CSV file of Bitcoin historical prices
• Hosting-ready: Structured for deployment on Vercel (frontend) + Render or Railway (backend)

Screenshot

Getting Started

Prerequisites

• Node.js (v18+)
• Python 3.8+
• OpenAI API key

Installation

1. Clone the repository

git clone https://github.com/TobyKThurston/backtest-anything.git
cd backtest-anything

2. Set up the backend

cd backend
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Create a .env file inside the backend folder:

OPENAI_API_KEY=your_openai_api_key_here

Then run the FastAPI server:

uvicorn main:app --reload

3. Set up the frontend

cd ../frontend
npm install
npm run dev

Visit http://localhost:3000 to use the app.

Folder Structure

backtest-anything/
├── backend/             # FastAPI backend with OpenAI integration
│   ├── main.py
│   ├── btc.csv
│   └── .env
├── frontend/            # Next.js frontend app
│   ├── app/
│   └── components/
├── README.md
└── image.png

Motivation

I wanted to build a project that combined AI, financial data, and real-world applications. Backtest Anything allows anyone to prototype and test algorithmic trading strategies with zero coding experience, and helped me deepen my experience in full-stack development and GPT-powered tools.